2:03 a.m.
After thirteen years as a loyal Red Hat and Fedora user, I reached my
limit today, when an attempt to upgrade one (1) package pitched me
into a four-hour marathon of dependency chasing, at the end of which
an attempt to get around a trivial file conflict rendered my system
unusable.
The proximate causes of this failure were (1) incompetent repository
maintenance, making any nontrivial upgrade certain to founder on a
failed dependency, and (2) the fact that rpm is not statically linked
-- so it's possible to inadvertently remove a shared library it
depends on and be unrecoverably screwed. But the underlying problems
run much deeper.
Over the last five years, I've watched Red Hat/Fedora throw away what
was at one time a near-unassailable lead in technical prowess, market
share and community prestige. The blunders have been legion on both
technical and political levels. They have included, but were not
limited to:
* Chronic governance problems.
* Persistent failure to maintain key repositories in a sane,
consistent state from which upgrades might actually be possible.
* A murky, poorly-documented, over-complex submission process.
* Allowing RPM development to drift and stagnate -- then adding
another layer of complexity, bugs, and wretched performance with yum.
* Effectively abandoning the struggle for desktop market share.
* Failure to address the problem of proprietary multimedia formats with
any attitude other than blank denial.
In retrospect, I should probably have cut my losses years ago. But I
had so much history with Red-Hat/Fedora, and had invested so much effort in
trying to fix the problems, that it was hard to even imagine
breaking away.
If I thought the state of Fedora were actually improving, I might hang
in there. But it isn't. I've been on the fedora-devel list for
years, and the trend is clear. The culture of the project's core
group has become steadily more unhealthy, more inward-looking, more
insistent on narrow "free software" ideological purity, and more
disconnected from the technical and evangelical challenges that must
be met to make Linux a world-changing success that liberates a
majority of computer users.
I have watched Ubuntu rise to these challenges as Fedora fell away
from them. Canonical's recent deal with Linspire, which will give
Linux users legal access to WMF and other key proprietary codecs, is
precisely the sort of thing Red-Hat/Fedora could and should have taken
the lead in. Not having done so bespeaks a failure of vision which I
now believe will condemn Fedora to a shrinking niche in the future.
This afternoon, I installed Edgy Eft on my main development machine --
from one CD, not five. In less than three hours' work I was able to
recreate the key features of my day-to-day toolkit. The
after-installation mass upgrade to current packages, always a
frightening prospect under Fedora, went off without a hitch.
I'm not expecting Ubuntu to be perfect, but I am now certain it will
be enough better to compensate me for the fact that I need to learn
a new set of administration tools.
Fedora, you had every advantage, and you had my loyalty, and you blew it.
And that is a damn, dirty shame.
--
Eric S. Raymond
2:30 a.m.
On Wed, 2007-02-21 at 09:11 +0100, Gianluca Sforna wrote:
On 2/21/07, Eric S. Raymond <esr(a)thyrsus.com> wrote:
> Fedora, you had every advantage, and you had my loyalty, and you blew it.
> And that is a damn, dirty shame.
/me bets this is going to be one of the longest threads in the
fedora-devel history...
*Sigh*
- Gilboa "Thankfully my cable provider doesn't charge me for extra
bandwidth because this is gonna get very ugly, very fast... kinda like
the ESR's previous drive by posting" Davara.
2:17 a.m.
+-------------------+ .:\:\:/:/:.
| PLEASE DO NOT | :.:\:\:/:/:.:
| FEED THE TROLLS | :=.' - - '.=:
| | '=(\ 9 9 /)='
| Thank you, | ( (_) )
| Management | /`-vvv-'\
+-------------------+ / \
| | @@@ / /|,,,,,|\ \
| | @@@ /_// /^\ \\_\
@x@@x@ | | |/ WW( ( ) )WW
\||||/ | | \| __\,,\ /,,/__
\||/ | | | (______Y______)
/\/\/\/\/\/\/\/\//\/\\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\
==================================================================
2:38 a.m.
Regarding package management, I have found apt to work much better
that either SuSE's YAST or Redhat's yum. dpkg and rpm are both
challenging for me, and I have struggled with dependency issues
with each. I tried to track OpenSuSE and found it completely
undoable. Now I track Rawhide and Ubuntu's development packages.
I have had repeated success upgrading installations to Ubuntu's
development tree.
What I appreciate in Fedora is the SELinux integration and all the
patches that get applied to the kernels. The kernel expertise at
Redhat is fantastic.
Although I know that eventually the Gnome folks will whittle all the
beryl options down to a manageable few, with a more spare
configuration UI, compiz lacks any configuration UI. Since Ubuntu
has gone with Beryl and Fedora is more committed to compiz,
I have found it easier to get a pretty solid 3D desktop under Ubuntu.
Currently, I have almost identical desktop environments under
the Fedora and Ubuntu development installations. I contribute
bug reports to both. I find that Fedora bugs tend to get more rapid
fixes that Ubuntu. This is really important to me. I contribute testing
to Linux desktops because I want to see Linux and OSS contribute
to empowering users globally. I have been contributing for about eight
years, and have been very happy with the progress that has been made.
I have recently been tracking the work being done under by folks
associated with the desktop-wireless summits (under the auspices
of OSDL). There is still a lot of work to do with wireless support.
I've been glad to see both Fedora and Ubuntu get on the
NetworkManager integration bandwagon.
We need strong distributions that can put finishing touches on
all the integrated projects. We still have a ways to go with desktop
usability and so forth. I hope that Fedora will flourish and continue
to play a vital role in coordinating, evolving and integrating all the
codebases that feed into it. It's an invaluable role.
2:45 a.m.
Quoting "Eric S. Raymond" <esr(a)thyrsus.com>:
* Persistent failure to maintain key repositories in a sane,
consistent state from which upgrades might actually be possible.
Core and Extras merged on future Fedora 7 so there is only one official Fedora
repository.
* Effectively abandoning the struggle for desktop market share.
* Failure to address the problem of proprietary multimedia formats with
any attitude other than blank denial.
Sorry Eric. It seems you are willing to compromise Fedora philosophies on what
it is not: a distribution with default installed closed sources . Since the
installation of Fedora can be customized with the ability to add an external
repository, the point is completely irrelevant.
I have watched Ubuntu rise to these challenges as Fedora fell away
from them. Canonical's recent deal with Linspire, which will give
Linux users legal access to WMF and other key proprietary codecs, is
precisely the sort of thing Red-Hat/Fedora could and should have taken
the lead in. Not having done so bespeaks a failure of vision which I
now believe will condemn Fedora to a shrinking niche in the future.
I think you threw away the FOSS philosophies that Fedora applied for a benefit
of patented codecs and popularity. IMHO, Linspire/Canonical means Canonical is
seeking a way to earn moneys with the inclusion of patented codecs on their
free distribution. Have you eared the term "victim of its own success"? There
will be a price to pay.
This afternoon, I installed Edgy Eft on my main development machine
--
from one CD, not five. In less than three hours' work I was able to
recreate the key features of my day-to-day toolkit. The
after-installation mass upgrade to current packages, always a
frightening prospect under Fedora, went off without a hitch.
You should know you can use a boot disk to do a network install a system. Five
CD is aiming for people who don't have broadband connection which is still the
case on many countries.
I'm not expecting Ubuntu to be perfect, but I am now certain it
will
be enough better to compensate me for the fact that I need to learn
a new set of administration tools.
Fedora, you had every advantage, and you had my loyalty, and you blew it.
And that is a damn, dirty shame.
--
Eric S. Raymond
--
fedora-devel-list mailing list
fedora-devel-list(a)redhat.com
https://www.redhat.com/mailman/listinfo/fedora-devel-list
--
Luya Tshimbalanga
Fedora Project contributor
http://www.fedoraproject.org/wiki/LuyaTshimbalanga
2:59 a.m.
New subject: Goodbye, /lib/*
Eric S. Raymond wrote:
After thirteen years as a loyal Red Hat and Fedora user, I reached
my
limit today, when an attempt to upgrade one (1) package pitched me
into a four-hour marathon of dependency chasing, at the end of which
an attempt to get around a trivial file conflict rendered my system
unusable.
The proximate causes of this failure were (1) incompetent repository
maintenance, making any nontrivial upgrade certain to founder on a
failed dependency, and (2) the fact that rpm is not statically linked
-- so it's possible to inadvertently remove a shared library it
depends on and be unrecoverably screwed. But the underlying problems
run much deeper.
If I understood this, you deleted something down /lib that rpm depended
on. That's not unrecoverably screwed: you can boot into the rescue CD
and bring over the libs or the rpm the libs came from and use rpm2cpio.
I have done this in the past, and recovered from it, and my assessment
of the proximate cause of the failure was that 1) I was a careless idiot.
RPM being monolithic would help in that situation, but then it's open to
you to delete rpm if you're up for deleting shared libs. Likewise
Ubuntu is going to have things that can be deleted that will render it
equally "unrecoverable". Basically there isn't much to be done to
protect a tired or stupid admin using rm as root on important files: but
the recovery boot is there to get you out of even very bad trouble.
I don't think we should entirely absolve the wielder of the rm machete
from blame even if he didn't ask to be swinging it about.
-Andy
3:15 a.m.
I'll bite.
On Wed, 2007-02-21 at 03:03 -0500, Eric S. Raymond wrote:
After thirteen years as a loyal Red Hat and Fedora user, I reached
my
limit today, when an attempt to upgrade one (1) package pitched me
into a four-hour marathon of dependency chasing, at the end of which
an attempt to get around a trivial file conflict rendered my system
unusable.
I've been using RH/Fedora for about the same amount of time. I've never
broken my system to that level outside of test / rawhide. Not going to
argue that dependencies couldn't be better, because they always can.
Heck, I'd love to build a server with some graphical config tools and
not require printing libraries, but there you go..
* Allowing RPM development to drift and stagnate -- then adding
another layer of complexity, bugs, and wretched performance with yum.
Yeah, RPM did stagnate, but at least the RPM based distros are trying to
do something about it.
* Effectively abandoning the struggle for desktop market share.
* Failure to address the problem of proprietary multimedia formats
with
any attitude other than blank denial.
Nobody ever said it "should" be a major desktop alternative or play
mp3s. The goal of Fedora is "..the rapid progress of free and open
source software and content". Fedora "allows" you to do what you want
with it, it's not everything to everyone and doesn't try to be.
I have watched Ubuntu rise to these challenges as Fedora fell away
from them. Canonical's recent deal with Linspire, which will give
Linux users legal access to WMF and other key proprietary codecs, is
precisely the sort of thing Red-Hat/Fedora could and should have taken
the lead in. Not having done so bespeaks a failure of vision which I
now believe will condemn Fedora to a shrinking niche in the future.
That's all well and nice for Ubuntu which seems to be a desktop centric
OS. But I've always seen Fedora as primarily a server platform (given
it's the basis for RHEL). And that is fine for me because I don't want
to play windows media. I want Apache, Fedora Directory Service and a
fast kernel. Web browsing and email are great for a workstation but
everything else is a luxury.
At the end of the day if I really "need" to see that video of the Chimp
falling out of the tree after smelling his finger then I'll drop mplayer
on from a public repo. If you want that standard then yeah, maybe
Fedora isn't for you.
I can't debate the submission process because I don't submit, but I do
know from the devel list discussions that people are trying to make that
(along with other things) easier with the merge.
This afternoon, I installed Edgy Eft on my main development machine
--
from one CD, not five. In less than three hours' work I was able to
recreate the key features of my day-to-day toolkit. The
after-installation mass upgrade to current packages, always a
frightening prospect under Fedora, went off without a hitch.
I can network install a working fedora box with what I need in 10
minutes. That's not much of a technical argument, especially as fedora
is openly committed to allowing custom spins. So you will be able to
build a box based on whatever you like soon enough. That is moving in
the right direction.
I'm not expecting Ubuntu to be perfect,
That's lucky.
but I am now certain it will
be enough better to compensate me for the fact that I need to learn
a new set of administration tools.
Fedora, you had every advantage, and you had my loyalty, and you blew it.
And that is a damn, dirty shame.
Sleep on it. Fedora is open, and people are committed to it, so it
improves with each release.
3:52 a.m.
On Wed, Feb 21, 2007 at 03:03:50AM -0500, Eric S. Raymond wrote:
[ a vast amount of crap (as usual) deleted]
Don't let the door hit you on the way out.
--
Cheers! (Relax...have a homebrew)
Neil
THEOREM: VI is perfect.
PROOF: VI in roman numerals is 6. The natural numbers < 6 which divide 6 are
1, 2, and 3. 1+2+3 = 6. So 6 is a perfect number. Therefore, VI is perfect.
QED
-- Arthur Tateishi
5:58 a.m.
On Wed, 21 Feb 2007 03:03:50 -0500 (EST), Eric S. Raymond wrote:
After thirteen years as a loyal Red Hat and Fedora user, I reached
my
limit today, when an attempt to upgrade one (1) package pitched me
into a four-hour marathon of dependency chasing, at the end of which
an attempt to get around a trivial file conflict rendered my system
unusable.
Without details it is hardly possible to comment on it. But pointing out
the following should be allowed:
The Fedora Extras 6 repository currently is free of broken dependencies.
The Fedora Extras 5 repository contains a single broken package that
is broken for 131 days (linphone).
Conflicts in Fedora Extras are due to the committee's refusal to publish
related policies. Instead, packagers and reviewers are confronted with
lots of minor policies and bureaucracy.
The proximate causes of this failure were (1) incompetent repository
maintenance,
"Incompetent" in what way? At least on the Fedora Extras side we haven't
had any broken metadata, unsigned rpms or permission problems for a very
very long time. That part of the repository maintenance has been flawless.
In the "development" repositories, broken dependencies could only be
prevented in Core, because Core and Extras are still separate and use
separate build systems, too. When and if they are merged for Fedora 7, it
would be possible to only publish packages (or sets of packages), which
don't break any install-time dependencies. Still, it would not guarantee
that the development side of Fedora is usable and free of problems always.
6:18 a.m.
New subject: Conflicts policy (Was: Re: Goodbye, Fedora)
On 21.02.2007 12:58, Michael Schwendt wrote:
On Wed, 21 Feb 2007 03:03:50 -0500 (EST), Eric S. Raymond wrote:
> After thirteen years as a loyal Red Hat and Fedora user, I reached my
> limit today, when an attempt to upgrade one (1) package pitched me
> into a four-hour marathon of dependency chasing, at the end of which
> an attempt to get around a trivial file conflict rendered my system
> unusable.
Without details it is hardly possible to comment on it. But pointing out
the following should be allowed:
The Fedora Extras 6 repository currently is free of broken dependencies.
The Fedora Extras 5 repository contains a single broken package that
is broken for 131 days (linphone).
Conflicts in Fedora Extras are due to the committee's refusal to publish
related policies. Instead, packagers and reviewers are confronted with
lots of minor policies and bureaucracy. [...]
That reminds me: When will the conflicts policy/guidelines ever get
finished and made effective by the Packaging Committee? I think it was
voted and accepted (not sure), but it's still in the drafts section:
http://www.fedoraproject.org/wiki/PackagingDrafts/Conflicts
CU
thl
7:32 a.m.
New subject: Conflicts policy (Was: Re: Goodbye, Fedora)
On Wed, 21 Feb 2007 13:18:48 +0100, Thorsten Leemhuis wrote:
That reminds me: When will the conflicts policy/guidelines ever get
finished and made effective by the Packaging Committee? I think it was
voted and accepted (not sure), but it's still in the drafts section:
http://www.fedoraproject.org/wiki/PackagingDrafts/Conflicts
* No package for %{dist} must conflict with any other package for the
same %{dist} (multi-lib issues left aside!). Neither implicitly nor
explicitly.
This is a policy FESCo could decide on without having to wait for the
Packaging Comittee. There is no technical knowlegde needed to decide on
that. Does FESCo like surprises during full installations or when future
updates pull in a conflicting dep-chain?
* Exceptions?
Perhaps "lazy packaging" in compat*-devel packages as mentioned in above
document? Though, it should be possible to relocate header files and
modify foo-config scripts appropriately.
Packages with a completely disjunct target group? E.g. "qstat" vs.
"torque"
in Fedora Extras. Still, with the background of torque, it would be
justified if the qstat package (upstream!) renamed its files in order
to resolve that conflict.
7:59 a.m.
New subject: Conflicts policy (Was: Re: Goodbye, Fedora)
On 21.02.2007 14:32, Michael Schwendt wrote:
On Wed, 21 Feb 2007 13:18:48 +0100, Thorsten Leemhuis wrote:
> That reminds me: When will the conflicts policy/guidelines ever get
> finished and made effective by the Packaging Committee? I think it was
> voted and accepted (not sure), but it's still in the drafts section:
> http://www.fedoraproject.org/wiki/PackagingDrafts/Conflicts
* No package for %{dist} must conflict with any other package for the
same %{dist} (multi-lib issues left aside!). Neither implicitly nor
explicitly.
This is a policy FESCo could decide on without having to wait for the
Packaging Comittee.
Agreed, with the new world order (PC below FESCo and the new merged
FESCo for both Core and Extras) that could be done by FESCo -- when I
was still in FESCo and it was still Extras only I tried to avoid such a
move because I didn't want to have such a thing as Extras-specific rule.
There is no technical knowlegde needed to decide on
that.
Well, it will quickly raise questions like "Is 'Conflicts kernel <
2.6.17' okay?"
Does FESCo like surprises during full installations or when future
updates pull in a conflicting dep-chain?
Seems nobody is targeting anymore to make "full installations" possible.
[...]
CU
thl
8:02 a.m.
New subject: Conflicts policy (Was: Re: Goodbye, Fedora)
On Wed, 2007-02-21 at 14:59 +0100, Thorsten Leemhuis wrote:
On 21.02.2007 14:32, Michael Schwendt wrote:
> On Wed, 21 Feb 2007 13:18:48 +0100, Thorsten Leemhuis wrote:
[snip]
> Does FESCo like surprises during full installations or when
future
> updates pull in a conflicting dep-chain?
Seems nobody is targeting anymore to make "full installations" possible.
Is it even theoretically possible?
- Gilboa
8:45 a.m.
New subject: Conflicts policy (Was: Re: Goodbye, Fedora)
On Wed, 21 Feb 2007 16:02:49 +0200, Gilboa Davara wrote:
> Seems nobody is targeting anymore to make "full
installations" possible.
>
Is it even theoretically possible?
If the allocated disk space is large enough and all packages can be
installed without conflicts, it boils down to whether all packages can
coexist, for example whether they use the alternatives system to replace
eachother or whether they fight for resources.
Imagine somebody, who tries to install all -devel packages, because
compiling software from tarballs needs many of them. All sorts of
conflicts that request user input are like poison.
But even without full installs, imagine somebody installs Fedora 7 and
adds a customised selection of packages later. After updating the
installation regularly without problems, suddenly an update leads to a
needed chain of dependencies, which is rejected due to conflicts, because
either a) there are a packaging bug, or b) there are explicit "Conflicts:"
in the packages. Hence it is important that all conflicts inside %dist
package set are examined closely.
8:19 a.m.
New subject: Conflicts policy (Was: Re: Goodbye, Fedora)
On Wed, 21 Feb 2007 14:59:51 +0100, Thorsten Leemhuis wrote:
> There is no technical knowlegde needed to decide on
> that.
Well, it will quickly raise questions like "Is 'Conflicts kernel <
2.6.17' okay?"
No, the question is irrelevant in the context of the %{dist} package space
where the kernel is known to be >= 2.6.17. Conflicts of Fedora 7 with
packages from Red Hat Linux 9 are outside the scope of such a policy. The
primary focus should be on what conflicts can occur if somebody does
a [fresh] install of Fedora 7? And what conflicts are possible a few
months later after hundreds of updates?
> Does FESCo like surprises during full installations or when
future
> updates pull in a conflicting dep-chain?
Seems nobody is targeting anymore to make "full installations" possible.
Possibly not at install-time. But all packages are available after
firstboot.
6:33 a.m.
On Wed, Feb 21, 2007 at 03:03:50AM -0500, Eric S. Raymond wrote:
* Failure to address the problem of proprietary multimedia formats
with
any attitude other than blank denial.
That would be because we believe in Free Software and doing the right thing
(a practice you appear to have given up on). Maybe it is time the term
"open source" also did the decent thing and died out with you.
I'm not expecting Ubuntu to be perfect, but I am now certain it
will
be enough better to compensate me for the fact that I need to learn
a new set of administration tools.
I'm sure they will be delighted to have you
Alan
--
"That was said by Eric Raymond who belongs to another movement"
- Richard Stallman
7:35 a.m.
On Wed, 2007-02-21 at 07:33 -0500, Alan Cox wrote:
On Wed, Feb 21, 2007 at 03:03:50AM -0500, Eric S. Raymond wrote:
> * Failure to address the problem of proprietary multimedia formats with
> any attitude other than blank denial.
That would be because we believe in Free Software and doing the right thing
(a practice you appear to have given up on). Maybe it is time the term
"open source" also did the decent thing and died out with you.
Who the hell made you the arbiter of what's right and wrong? You?
I've about had it up to here with this New Slavery crap, this
pseudo-socialist movement in the Linux world being pushed by self styled
self aggrandizing arrogant ivory tower elitists intent on a fascist
enforcement of the licensing policies of THEIR choice dictated from
inside a closed circle-jerk where everybody pats themselves on the back
and laughs about how stupid the "outsiders" are. The Greg KH's and
Andrew Mortons of the world that are hell bent on telling programmers
how to license their own code at the point of a gun. It's repugnant,
just about as repugnant as this self-important arrogant post of yours.
> I'm not expecting Ubuntu to be perfect, but I am now certain
it will
> be enough better to compensate me for the fact that I need to learn
> a new set of administration tools.
I'm sure they will be delighted to have you
Alan
And it's the toy-hat dictators and the toy-hat fascists that are pushing
him out the door. Socialism was the mother of both fascism and
communism, and it's the same thing that's destroying Fedora, thanks in
part to you. Welcome to the New Slavery. It's the same as the Old
Slavery.
--
"That was said by Eric Raymond who belongs to another movement"
- Richard Stallman
LX
--
°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°
"Anybody else wanna negotiate?" -- Maj. Korben Dallas
Registered Linux User #268899 http://counter.li.org/
°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°
7:51 a.m.
Lyvim Xaphir wrote :
I've about had it up to here with this New Slavery crap, this
pseudo-socialist movement in the Linux world being pushed by self styled
self aggrandizing arrogant ivory tower elitists intent on a fascist
enforcement of the licensing policies of THEIR choice dictated from
inside a closed circle-jerk where everybody pats themselves on the back
and laughs about how stupid the "outsiders" are. The Greg KH's and
Andrew Mortons of the world that are hell bent on telling programmers
how to license their own code at the point of a gun. It's repugnant,
just about as repugnant as this self-important arrogant post of yours.
Yeah, I mean, who are we to enforce our copyrights and licensing
choices!? Even go as far as FORCE freedom upon our users, huh? :P
And it's the toy-hat dictators and the toy-hat fascists that are
pushing
him out the door. Socialism was the mother of both fascism and
communism, and it's the same thing that's destroying Fedora, thanks in
part to you. Welcome to the New Slavery. It's the same as the Old
Slavery.
You've lost me here, as none of the above makes any sense to me. But it
does seem that you haven't really understood the basic goals of the
Fedora Project.
Matthias
--
Clean custom Red Hat Linux rpm packages : http://freshrpms.net/
Fedora Core release 6 (Zod) - Linux kernel 2.6.19-1.2895.fc6
Load : 0.58 0.50 0.56
7:52 a.m.
Lyvim Xaphir wrote:
And it's the toy-hat dictators and the toy-hat fascists that are
pushing
him out the door. Socialism was the mother of both fascism and
communism, and it's the same thing that's destroying Fedora, thanks in
part to you.
Godwin's Law achieved! In a record-breaking 5 hours!
7:58 a.m.
At 14:52 on 21 Feb 2007, Denis Leroy <denis(a)poolshark.org> wrote:
Lyvim Xaphir wrote:
> And it's the toy-hat dictators and the toy-hat fascists that are
> pushing him out the door. Socialism was the mother of both fascism
> and communism, and it's the same thing that's destroying Fedora,
> thanks in part to you.
Godwin's Law achieved! In a record-breaking 5 hours!
5:32, actually... Must try harder.
--
Mark Knoop
8:01 a.m.
On Wed, 2007-02-21 at 14:52 +0100, Denis Leroy wrote:
Lyvim Xaphir wrote:
> And it's the toy-hat dictators and the toy-hat fascists that are pushing
> him out the door. Socialism was the mother of both fascism and
> communism, and it's the same thing that's destroying Fedora, thanks in
> part to you.
Godwin's Law achieved! In a record-breaking 5 hours!
Somehow I doubt that the neither OP nor the troll above have any idea
what Godwin's Law is, nor do they seem to care.
Time to make fedora-devel developers/maintainers/etc only?
- Gilboa
8:45 a.m.
On Wednesday 21 February 2007, Gilboa Davara wrote:
On Wed, 2007-02-21 at 14:52 +0100, Denis Leroy wrote:
> Lyvim Xaphir wrote:
> > And it's the toy-hat dictators and the toy-hat fascists that
> > are pushing him out the door. Socialism was the mother of
> > both fascism and communism, and it's the same thing that's
> > destroying Fedora, thanks in part to you.
>
> Godwin's Law achieved! In a record-breaking 5 hours!
Somehow I doubt that the neither OP nor the troll above have any
idea what Godwin's Law is, nor do they seem to care.
Time to make fedora-devel developers/maintainers/etc only?
- Gilboa
'etc' should include those who actually run fc6-test and/or
f7-rawhide. LX uses neither (fc5-unity).
--
Tom Brinkman Corpus Christi, Texas
4:21 a.m.
On Wed, 2007-02-21 at 08:45 -0600, Tom Brinkman wrote:
On Wednesday 21 February 2007, Gilboa Davara wrote:
> On Wed, 2007-02-21 at 14:52 +0100, Denis Leroy wrote:
> > Lyvim Xaphir wrote:
> > > And it's the toy-hat dictators and the toy-hat fascists that
> > > are pushing him out the door. Socialism was the mother of
> > > both fascism and communism, and it's the same thing that's
> > > destroying Fedora, thanks in part to you.
> >
> > Godwin's Law achieved! In a record-breaking 5 hours!
>
> Somehow I doubt that the neither OP nor the troll above have any
> idea what Godwin's Law is, nor do they seem to care.
>
> Time to make fedora-devel developers/maintainers/etc only?
>
> - Gilboa
'etc' should include those who actually run fc6-test and/or
f7-rawhide. LX uses neither (fc5-unity).
And now you're an LXpert? You're in a Texas trailer park several states
away but yet you know what I have here?
2.6.19-1.2911.fc6 is running on my primary workstation not including my
five other test boxes.
LX
--
°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°
A Kernel Of Socialism
greatunwashed: module license 'great_unwashed' taints kernel.
ich: no version for "unwashed_register_device" found: kernel tainted.
Symbol usb_register_driver is being used by a non-GPL module,
which will not be allowed in the future
Please see the file Documentation/feature-removal-schedule.txt
in the kernel source tree for more details.
°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°
2:03 p.m.
On Friday 23 February 2007, Lyvim Xaphir wrote:
2.6.19-1.2911.fc6 is running on my primary workstation not
including my five other test boxes.
LX
"About fedora-devel-list English (USA)
This list is only for discussion of development issues in the
Fedora Project.
THIS IS NOT A SUPPORT LIST. THIS LIST IS FOR CORE DEVELOPMENT
DISCUSSION ONLY." (their emphasis, not mine)
https://www.redhat.com/mailman/listinfo/fedora-devel-list
Take your moronic B$ back to OT, where it's at least tolerated,
found amusing. As is ESR
I apologize to the rest of y'all
--
Tom Brinkman Corpus Christi, Texas
3:53 a.m.
On Fri, 2007-02-23 at 14:03 -0600, Tom Brinkman wrote:
On Friday 23 February 2007, Lyvim Xaphir wrote:
> 2.6.19-1.2911.fc6 is running on my primary workstation not
> including my five other test boxes.
>
> LX
"About fedora-devel-list English (USA)
This list is only for discussion of development issues in the
Fedora Project.
THIS IS NOT A SUPPORT LIST. THIS LIST IS FOR CORE DEVELOPMENT
DISCUSSION ONLY." (their emphasis, not mine)
https://www.redhat.com/mailman/listinfo/fedora-devel-list
Take your moronic B$ back to OT, where it's at least tolerated,
found amusing. As is ESR
Go see your doctor about your meds.
LX
--
°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°
A Kernel Of Socialism
greatunwashed: module license 'great_unwashed' taints kernel.
ich: no version for "unwashed_register_device" found: kernel tainted.
Symbol usb_register_driver is being used by a non-GPL module,
which will not be allowed in the future
Please see the file Documentation/feature-removal-schedule.txt
in the kernel source tree for more details.
°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°
8 a.m.
On 2/21/07, Lyvim Xaphir <knightmerc(a)yahoo.com> wrote:
I've about had it up to here with this New Slavery crap,
[snip etc]
This is a parody to make ESR look bad... right? (Don't bother
answering, it was a rhetorical question.)
On a more useful note, does anyone have any idea of exactly which
package ESR is claiming caused the complete meltdown of his system?
Is he actually claiming that "yum update <package>" on a FC6 system
caused the problems that he mentions? Or is he actually referring to
how the ItEatsBabiesAndYouWereWarned rawhide ate his babies?
This sort of bug report is not helpful as it's slim on concrete
details and leaves a murky wake of confusion. There's some good
advice on how to avoid that type of thing in these two links:
http://www.catb.org/~esr/faqs/smart-questions.html#symptoms
http://www.catb.org/~esr/faqs/smart-questions.html#id273206
Oisin Feeley
12:28 p.m.
On 2/21/07, Oisin Feeley <oisin.feeley(a)gmail.com> wrote:
On a more useful note, does anyone have any idea of exactly which
package ESR is claiming caused the complete meltdown of his system?
Is he actually claiming that "yum update <package>" on a FC6 system
caused the problems that he mentions? Or is he actually referring to
how the ItEatsBabiesAndYouWereWarned rawhide ate his babies?
This sort of bug report is not helpful as it's slim on concrete
details and leaves a murky wake of confusion. There's some good
advice on how to avoid that type of thing in these two links:
The ekiga update from today has broken dependencies.
Error: Missing Dependency: opal >= 2.2.5 is needed by package ekiga
Error: Missing Dependency: libpt_linux_x86_r.so.1.10.4 is needed by
package ekiga
Error: Missing Dependency: pwlib >= 1.10.4 is needed by package ekiga
It looks like the opal and pwlib versions it needs are in updates-testing.
Excluding ekiga with "yum --exclude=ekiga update" allows the rest of
the packages to update. Enabling updates-testing repository with "yum
--enablerepo=updates-testing update ekiga" allows ekiga to be updated.
- Ian
12:50 p.m.
On Wednesday 21 February 2007 13:28, Ian Burrell wrote:
The ekiga update from today has broken dependencies.
Error: Missing Dependency: opal >= 2.2.5 is needed by package ekiga
Error: Missing Dependency: libpt_linux_x86_r.so.1.10.4 is needed by
package ekiga
Error: Missing Dependency: pwlib >= 1.10.4 is needed by package ekiga
It looks like the opal and pwlib versions it needs are in updates-testing.
Excluding ekiga with "yum --exclude=ekiga update" allows the rest of
the packages to update. Enabling updates-testing repository with "yum
--enablerepo=updates-testing update ekiga" allows ekiga to be updated.
I honestly have no idea how this happened. I did a depcheck before pushing
the update. That said, I'm pushing the missing packages now.
--
Jesse Keating
Release Engineer: Fedora
2:59 p.m.
New subject: FC6 updates broken deps? (was: Re: Goodbye, Fedora)
On Wed, 21 Feb 2007 13:50:50 -0500, Jesse Keating wrote:
On Wednesday 21 February 2007 13:28, Ian Burrell wrote:
> The ekiga update from today has broken dependencies.
>
> Error: Missing Dependency: opal >= 2.2.5 is needed by package ekiga
> Error: Missing Dependency: libpt_linux_x86_r.so.1.10.4 is needed by
> package ekiga
> Error: Missing Dependency: pwlib >= 1.10.4 is needed by package ekiga
>
> It looks like the opal and pwlib versions it needs are in updates-testing.
>
> Excluding ekiga with "yum --exclude=ekiga update" allows the rest of
> the packages to update. Enabling updates-testing repository with "yum
> --enablerepo=updates-testing update ekiga" allows ekiga to be updated.
I honestly have no idea how this happened. I did a depcheck before pushing
the update. That said, I'm pushing the missing packages now.
Does that depcheck also cover multi-lib? Because except for the
first one, I still see these:
source rpm: python-virtinst-0.98.0-1.fc6.src.rpm
package: python-virtinst - 0.98.0-1.fc6.noarch from fedora-core-updates-6-ppc
unresolved deps:
libvirt-python >= 0:0.1.4-4
source rpm: bind-9.3.4-2.fc6.src.rpm
package: bind-devel - 31:9.3.4-2.fc6.i386 from fedora-core-updates-6-x86_64
unresolved deps:
libdns.so.22
libbind9.so.0
libisccc.so.0
liblwres.so.9
libisccfg.so.1
libisc.so.11
source rpm: compiz-0.3.6-2.fc6.src.rpm
package: compiz-devel - 0.3.6-2.fc6.i386 from fedora-core-updates-6-x86_64
unresolved deps:
libdecoration.so.0
source rpm: kdeedu-3.5.6-0.1.fc6.src.rpm
package: kdeedu - 3.5.6-0.1.fc6.i386 from fedora-core-updates-6-x86_64
unresolved deps:
libpython2.4.so.1.0
source rpm: kdeutils-3.5.6-0.1.fc6.src.rpm
package: kdeutils-devel - 6:3.5.6-0.1.fc6.i386 from fedora-core-updates-6-x86_64
unresolved deps:
libkmilo.so.1
libkregexpeditorcommon.so.1
libksimcore.so.1
libkhexeditcommon.so.0
libkcmlaptop.so.0
source rpm: poppler-0.5.4-5.fc6.src.rpm
package: poppler-devel - 0.5.4-5.fc6.i386 from fedora-core-updates-6-x86_64
unresolved deps:
libpoppler-qt.so.1
7:50 p.m.
New subject: FC6 updates broken deps? (was: Re: Goodbye, Fedora)
On Wednesday 21 February 2007 15:59, Michael Schwendt wrote:
Does that depcheck also cover multi-lib? Because except for the
first one, I still see these:
source rpm: python-virtinst-0.98.0-1.fc6.src.rpm
package: python-virtinst - 0.98.0-1.fc6.noarch from
fedora-core-updates-6-ppc unresolved deps:
libvirt-python >= 0:0.1.4-4
source rpm: bind-9.3.4-2.fc6.src.rpm
package: bind-devel - 31:9.3.4-2.fc6.i386 from fedora-core-updates-6-x86_64
unresolved deps:
libdns.so.22
libbind9.so.0
libisccc.so.0
liblwres.so.9
libisccfg.so.1
libisc.so.11
source rpm: compiz-0.3.6-2.fc6.src.rpm
package: compiz-devel - 0.3.6-2.fc6.i386 from fedora-core-updates-6-x86_64
unresolved deps:
libdecoration.so.0
source rpm: kdeedu-3.5.6-0.1.fc6.src.rpm
package: kdeedu - 3.5.6-0.1.fc6.i386 from fedora-core-updates-6-x86_64
unresolved deps:
libpython2.4.so.1.0
source rpm: kdeutils-3.5.6-0.1.fc6.src.rpm
package: kdeutils-devel - 6:3.5.6-0.1.fc6.i386 from
fedora-core-updates-6-x86_64 unresolved deps:
libkmilo.so.1
libkregexpeditorcommon.so.1
libksimcore.so.1
libkhexeditcommon.so.0
libkcmlaptop.so.0
source rpm: poppler-0.5.4-5.fc6.src.rpm
package: poppler-devel - 0.5.4-5.fc6.i386 from fedora-core-updates-6-x86_64
unresolved deps:
libpoppler-qt.so.1
Hrm, I'm not entirely sure, I'll have to defer to Luke Macken on that one.
I'm surprised that these haven't been reported before, people are generally
really quick to notice broken deps in the updates repos.
--
Jesse Keating
Release Engineer: Fedora
3:25 a.m.
New subject: FC6 updates broken deps?
Jesse Keating wrote:
On Wednesday 21 February 2007 15:59, Michael Schwendt wrote:
> Does that depcheck also cover multi-lib? Because except for the
> first one, I still see these:
>
> source rpm: python-virtinst-0.98.0-1.fc6.src.rpm
> package: python-virtinst - 0.98.0-1.fc6.noarch from
> fedora-core-updates-6-ppc unresolved deps:
> libvirt-python >= 0:0.1.4-4
>
> source rpm: bind-9.3.4-2.fc6.src.rpm
> package: bind-devel - 31:9.3.4-2.fc6.i386 from fedora-core-updates-6-x86_64
> unresolved deps:
> libdns.so.22
> libbind9.so.0
> libisccc.so.0
> liblwres.so.9
> libisccfg.so.1
> libisc.so.11
>
> source rpm: compiz-0.3.6-2.fc6.src.rpm
> package: compiz-devel - 0.3.6-2.fc6.i386 from fedora-core-updates-6-x86_64
> unresolved deps:
> libdecoration.so.0
>
> source rpm: kdeedu-3.5.6-0.1.fc6.src.rpm
> package: kdeedu - 3.5.6-0.1.fc6.i386 from fedora-core-updates-6-x86_64
> unresolved deps:
> libpython2.4.so.1.0
>
> source rpm: kdeutils-3.5.6-0.1.fc6.src.rpm
> package: kdeutils-devel - 6:3.5.6-0.1.fc6.i386 from
> fedora-core-updates-6-x86_64 unresolved deps:
> libkmilo.so.1
> libkregexpeditorcommon.so.1
> libksimcore.so.1
> libkhexeditcommon.so.0
> libkcmlaptop.so.0
>
> source rpm: poppler-0.5.4-5.fc6.src.rpm
> package: poppler-devel - 0.5.4-5.fc6.i386 from fedora-core-updates-6-x86_64
> unresolved deps:
> libpoppler-qt.so.1
Hrm, I'm not entirely sure, I'll have to defer to Luke Macken on that one.
I'm surprised that these haven't been reported before, people are generally
really quick to notice broken deps in the updates repos.
Fact of the matter is, that even though people should report such
iregularities, it would be a lot less work for everybody, if yum would
update the largest portion of updates that do not have any dependency
problems. I know we've been over this like a thousand times, but I still
see no valid reason not to make yum do this!
That would cause the 1 or 2 or 3 packages with probles to be held back,
not the rest.
/Thomas
/Thomas
6:39 a.m.
New subject: FC6 updates broken deps?
On Thu, 2007-02-22 at 10:25 +0100, Thomas M Steenholdt wrote:
Jesse Keating wrote:
> On Wednesday 21 February 2007 15:59, Michael Schwendt wrote:
>> Does that depcheck also cover multi-lib? Because except for the
>> first one, I still see these:
>>
>> source rpm: python-virtinst-0.98.0-1.fc6.src.rpm
>> package: python-virtinst - 0.98.0-1.fc6.noarch from
>> fedora-core-updates-6-ppc unresolved deps:
>> libvirt-python >= 0:0.1.4-4
>>
>> source rpm: bind-9.3.4-2.fc6.src.rpm
>> package: bind-devel - 31:9.3.4-2.fc6.i386 from fedora-core-updates-6-x86_64
>> unresolved deps:
>> libdns.so.22
>> libbind9.so.0
>> libisccc.so.0
>> liblwres.so.9
>> libisccfg.so.1
>> libisc.so.11
>>
>> source rpm: compiz-0.3.6-2.fc6.src.rpm
>> package: compiz-devel - 0.3.6-2.fc6.i386 from fedora-core-updates-6-x86_64
>> unresolved deps:
>> libdecoration.so.0
>>
>> source rpm: kdeedu-3.5.6-0.1.fc6.src.rpm
>> package: kdeedu - 3.5.6-0.1.fc6.i386 from fedora-core-updates-6-x86_64
>> unresolved deps:
>> libpython2.4.so.1.0
>>
>> source rpm: kdeutils-3.5.6-0.1.fc6.src.rpm
>> package: kdeutils-devel - 6:3.5.6-0.1.fc6.i386 from
>> fedora-core-updates-6-x86_64 unresolved deps:
>> libkmilo.so.1
>> libkregexpeditorcommon.so.1
>> libksimcore.so.1
>> libkhexeditcommon.so.0
>> libkcmlaptop.so.0
>>
>> source rpm: poppler-0.5.4-5.fc6.src.rpm
>> package: poppler-devel - 0.5.4-5.fc6.i386 from fedora-core-updates-6-x86_64
>> unresolved deps:
>> libpoppler-qt.so.1
>
> Hrm, I'm not entirely sure, I'll have to defer to Luke Macken on that one.
>
> I'm surprised that these haven't been reported before, people are generally
> really quick to notice broken deps in the updates repos.
>
>
Fact of the matter is, that even though people should report such
iregularities, it would be a lot less work for everybody, if yum would
update the largest portion of updates that do not have any dependency
problems. I know we've been over this like a thousand times, but I still
see no valid reason not to make yum do this!
That would cause the 1 or 2 or 3 packages with probles to be held back,
not the rest.
and it would give users very little awareness that something didn't get
patched.
giving them a false sense of security.
-sv
8:15 a.m.
New subject: FC6 updates broken deps?
seth vidal wrote:
On Thu, 2007-02-22 at 10:25 +0100, Thomas M Steenholdt wrote:
> Fact of the matter is, that even though people should report such
> iregularities, it would be a lot less work for everybody, if yum would
> update the largest portion of updates that do not have any dependency
> problems. I know we've been over this like a thousand times, but I still
> see no valid reason not to make yum do this!
>
> That would cause the 1 or 2 or 3 packages with probles to be held back,
> not the rest.
>
>
and it would give users very little awareness that something didn't get
patched.
giving them a false sense of security.
-sv
The burden of a broken repo/updates should not be placed on the end
user. If there is a dependency problem, then the maintainer of the repo
and/or package should be notified and is responsible for fixing the
problem. What good does it do for the end user? If yum, pup, or pirut
notes that packages x,y & z were not updated then no one would get a
false sense of security (no more a false sense than when none of the
packages are updated anyway). Plus this would be a little more
presentable and user friendly.
Demond - beating a dead horse
8:19 a.m.
New subject: FC6 updates broken deps?
On Thu, 2007-02-22 at 09:15 -0500, Demond wrote:
seth vidal wrote:
The burden of a broken repo/updates should not be placed on the end
user. If there is a dependency problem, then the maintainer of the repo
and/or package should be notified and is responsible for fixing the
problem. What good does it do for the end user?
So they can file the bug to let us know. We all make mistakes and we
need lots of people to catch them.
It's not just the end user I'm thinking about here, it's all end users.
-sv
9:06 a.m.
New subject: FC6 updates broken deps?
seth vidal wrote:
On Thu, 2007-02-22 at 09:15 -0500, Demond wrote:
> seth vidal wrote:
> The burden of a broken repo/updates should not be placed on the end
> user. If there is a dependency problem, then the maintainer of the repo
> and/or package should be notified and is responsible for fixing the
> problem. What good does it do for the end user?
>
So they can file the bug to let us know. We all make mistakes and we
need lots of people to catch them.
Also complain in all the outlets available to them that yum is broken.
Perhaps we should look at an automatic/silent notification of dependency
problems.
This problem only really shows up when you have multiple repos that are
dependant on each other or in rawhide. Rawhide is understandable but
there should be someway of sparing the end user of a released Fedora 7
from these update problems. They can then concentrate on higher level bugs.
Demond
8:56 a.m.
New subject: FC6 updates broken deps?
On Thu, 2007-02-22 at 07:39 -0500, seth vidal wrote:
and it would give users very little awareness that something didn't get
patched.
giving them a false sense of security.
And also leave them potentially vulnerable to a larger number of issues.
In some cases, some patching is better than no patching. If httpd
doesn't get updated for a hot zero-day exploit because of a dependency
issue with gimp and my system gets exploited, that seems like a bad
thing.
--
David Hollis <dhollis(a)davehollis.com>
9:07 a.m.
New subject: FC6 updates broken deps?
On Thu, 2007-02-22 at 09:56 -0500, David Hollis wrote:
On Thu, 2007-02-22 at 07:39 -0500, seth vidal wrote:
>
> and it would give users very little awareness that something didn't get
> patched.
>
> giving them a false sense of security.
>
And also leave them potentially vulnerable to a larger number of issues.
In some cases, some patching is better than no patching. If httpd
doesn't get updated for a hot zero-day exploit because of a dependency
issue with gimp and my system gets exploited, that seems like a bad
thing.
and this is why the update system is writing out update data. So, we can
differentiate b/t update for feature and update for security, or
different grades of a security update.
then we can work more intelligently on:
yum update security-critical-only (as an example)
and have it do those only.
-sv
9:26 a.m.
New subject: FC6 updates broken deps?
seth vidal wrote:
and this is why the update system is writing out update data. So, we can
differentiate b/t update for feature and update for security, or
different grades of a security update.
then we can work more intelligently on:
yum update security-critical-only (as an example)
That would be a nice thing for sure, but would really depend on the
inclusion of some sort of --continue-on-failure functionality of yum. It
makes no sense to provide a "security updates only" option, if the 20
security updates in there could potentially be held back because of a
dep problem with one of them.
I've said this a thousand times over and I'll say it again. I would
*much* rather have 1 package with a known security problem (held back
because of a dep issue with that package), than 20.
/Thomas
11:10 a.m.
New subject: FC6 updates broken deps?
Thomas M Steenholdt schrieb:
seth vidal wrote:
>
> and this is why the update system is writing out update data. So, we can
> differentiate b/t update for feature and update for security, or
> different grades of a security update.
>
> then we can work more intelligently on:
> yum update security-critical-only (as an example)
>
That would be a nice thing for sure, but would really depend on the
inclusion of some sort of --continue-on-failure functionality of yum.
It makes no sense to provide a "security updates only" option, if the
20 security updates in there could potentially be held back because of
a dep problem with one of them.
I've said this a thousand times over and I'll say it again. I would
*much* rather have 1 package with a known security problem (held back
because of a dep issue with that package), than 20.
+1 !
there is no reason to hold back an update because of a broken dep thats
not even related to it.
/Thomas
3:27 a.m.
New subject: FC6 updates broken deps?
Jesse Keating wrote:
On Wednesday 21 February 2007 15:59, Michael Schwendt wrote:
> Does that depcheck also cover multi-lib? Because except for the
> first one, I still see these:
>
> source rpm: python-virtinst-0.98.0-1.fc6.src.rpm
> package: python-virtinst - 0.98.0-1.fc6.noarch from
> fedora-core-updates-6-ppc unresolved deps:
> libvirt-python >= 0:0.1.4-4
>
> source rpm: bind-9.3.4-2.fc6.src.rpm
> package: bind-devel - 31:9.3.4-2.fc6.i386 from fedora-core-updates-6-x86_64
> unresolved deps:
> libdns.so.22
> libbind9.so.0
> libisccc.so.0
> liblwres.so.9
> libisccfg.so.1
> libisc.so.11
>
> source rpm: compiz-0.3.6-2.fc6.src.rpm
> package: compiz-devel - 0.3.6-2.fc6.i386 from fedora-core-updates-6-x86_64
> unresolved deps:
> libdecoration.so.0
>
> source rpm: kdeedu-3.5.6-0.1.fc6.src.rpm
> package: kdeedu - 3.5.6-0.1.fc6.i386 from fedora-core-updates-6-x86_64
> unresolved deps:
> libpython2.4.so.1.0
>
> source rpm: kdeutils-3.5.6-0.1.fc6.src.rpm
> package: kdeutils-devel - 6:3.5.6-0.1.fc6.i386 from
> fedora-core-updates-6-x86_64 unresolved deps:
> libkmilo.so.1
> libkregexpeditorcommon.so.1
> libksimcore.so.1
> libkhexeditcommon.so.0
> libkcmlaptop.so.0
>
> source rpm: poppler-0.5.4-5.fc6.src.rpm
> package: poppler-devel - 0.5.4-5.fc6.i386 from fedora-core-updates-6-x86_64
> unresolved deps:
> libpoppler-qt.so.1
Hrm, I'm not entirely sure, I'll have to defer to Luke Macken on that one.
I'm surprised that these haven't been reported before, people are generally
really quick to notice broken deps in the updates repos.
Really - The reason dependency issues are such a big issue for fedora is
that all updates are held back because of a single package missing a
dependency. I'm sure everybody would feel better holding back one
package instead of all updates on this account.
/Thomas
11:35 a.m.
New subject: FC6 updates broken deps?
Thomas M Steenholdt wrote:
Really - The reason dependency issues are such a big issue for fedora
is
that all updates are held back because of a single package missing a
dependency. I'm sure everybody would feel better holding back one
package instead of all updates on this account.
If I'm not mistaken, yum seems fairly unique in this regard.
I mean, heck, look at Microsoft for example. Their update thing applies
as many patches as possible, and those that cannot be applied, well duh,
they don't get applied and the user is notified by big honking red icons
that something failed.
But a single error doesn't nuke the whole process. yum is _really_
boneheaded from this perspective. Your argument that a possibly much
more important security update might get masked by a dependency error
and is not applied is a very serious one and I don't think there can be
any logical rebuttal.
yum needs to be fixed, NOW. At this moment the update process puts
everyone at risk.
--
Florin Andrei
http://florin.myip.org/
1:40 p.m.
New subject: FC6 updates broken deps?
Alan Cox wrote:
On Thu, Feb 22, 2007 at 09:35:41AM -0800, Florin Andrei wrote:
> yum needs to be fixed, NOW. At this moment the update process puts
> everyone at risk.
Send patches.
Well the yum plugins already exist. Certainly a good start would be to
gather the myriad of yum plugins and include them in the main yum
distribution, so at least these things get shipped together (how is an
end-user supposed to learn about yum-skip-broken ?).
2:17 p.m.
New subject: FC6 updates broken deps?
On Thu, Feb 22, 2007 at 08:40:36PM +0100, Denis Leroy wrote:
>>yum needs to be fixed, NOW. At this moment the update process
puts
>>everyone at risk.
>Send patches.
Well the yum plugins already exist. Certainly a good start would be to
gather the myriad of yum plugins and include them in the main yum
distribution, so at least these things get shipped together (how is an
end-user supposed to learn about yum-skip-broken ?).
I don't think lumping them into the yum rpm is a good approach from a system
maintenance perspective.
How is an end-user supposed to learn about any functionality of any package?
--
Matthew Miller mattdm(a)mattdm.org <http://mattdm.org/>
Boston University Linux ------> <http://linux.bu.edu/>
3 p.m.
New subject: FC6 updates broken deps?
Matthew Miller wrote:
On Thu, Feb 22, 2007 at 08:40:36PM +0100, Denis Leroy wrote:
>>> yum needs to be fixed, NOW. At this moment the update process puts
>>> everyone at risk.
>> Send patches.
> Well the yum plugins already exist. Certainly a good start would be to
> gather the myriad of yum plugins and include them in the main yum
> distribution, so at least these things get shipped together (how is an
> end-user supposed to learn about yum-skip-broken ?).
I don't think lumping them into the yum rpm is a good approach from a system
maintenance perspective.
How is an end-user supposed to learn about any functionality of any package?
They don't have to learn much, if the defaults are sensible. Which, of
course, brings us to the discussion about what is "sensible" and what is
not. ;-)
--
Florin Andrei
http://florin.myip.org/
5:09 p.m.
New subject: FC6 updates broken deps?
Matthew Miller wrote:
On Thu, Feb 22, 2007 at 08:40:36PM +0100, Denis Leroy wrote:
>>> yum needs to be fixed, NOW. At this moment the update process puts
>>> everyone at risk.
>> Send patches.
> Well the yum plugins already exist. Certainly a good start would be to
> gather the myriad of yum plugins and include them in the main yum
> distribution, so at least these things get shipped together (how is an
> end-user supposed to learn about yum-skip-broken ?).
I don't think lumping them into the yum rpm is a good approach from a system
maintenance perspective.
I actually meant integrated into yum. For the same reasons we want
kernel RPMs to be pushed upstream. The idea here is to have more people
contribute to yum.
6:51 p.m.
New subject: FC6 updates broken deps?
On Fri, Feb 23, 2007 at 12:09:17AM +0100, Denis Leroy wrote:
>I don't think lumping them into the yum rpm is a good
approach from a
>system maintenance perspective.
I actually meant integrated into yum. For the same reasons we want
kernel RPMs to be pushed upstream. The idea here is to have more people
contribute to yum.
But all these things are already in yum-utils.
--
Matthew Miller mattdm(a)mattdm.org <http://mattdm.org/>
Boston University Linux ------> <http://linux.bu.edu/>
3:54 a.m.
New subject: FC6 updates broken deps?
Matthew Miller wrote:
On Fri, Feb 23, 2007 at 12:09:17AM +0100, Denis Leroy wrote:
>> I don't think lumping them into the yum rpm is a good approach from a
>> system maintenance perspective.
> I actually meant integrated into yum. For the same reasons we want
> kernel RPMs to be pushed upstream. The idea here is to have more people
> contribute to yum.
But all these things are already in yum-utils.
I love yum-utils, but I was talking about all the yum plugins packages
out there :
# yum list | grep ^yum- | wc -l
17
5:56 a.m.
New subject: FC6 updates broken deps?
On Fri, Feb 23, 2007 at 10:54:18AM +0100, Denis Leroy wrote:
>But all these things are already in yum-utils.
I love yum-utils, but I was talking about all the yum plugins packages
out there :
# yum list | grep ^yum- | wc -l
17
$ rpm -q --qf '%{sourcerpm}\n' yum-skip-broken
yum-utils-1.1.1-1.fc7.src.rpm
--
Matthew Miller mattdm(a)mattdm.org <http://mattdm.org/>
Boston University Linux ------> <http://linux.bu.edu/>
7:20 a.m.
New subject: FC6 updates broken deps?
Matthew Miller wrote:
On Fri, Feb 23, 2007 at 10:54:18AM +0100, Denis Leroy wrote:
>> But all these things are already in yum-utils.
> I love yum-utils, but I was talking about all the yum plugins packages
> out there :
> # yum list | grep ^yum- | wc -l
> 17
$ rpm -q --qf '%{sourcerpm}\n' yum-skip-broken
yum-utils-1.1.1-1.fc7.src.rpm
I'm not sure I get your point. Are you saying all the 17 plugins were
written by the same person ? My point is: most of the plugins should be
part of the core yum package and therefore installed by default on all
fedora distros ('installed' not being equal to 'enabled by default'),
and people - community members - should be working on making yum better,
rather than writing plugins to work around the known yum limitations.
11:39 a.m.
New subject: FC6 updates broken deps?
On Fri, Feb 23, 2007 at 02:20:33PM +0100, Denis Leroy wrote:
>>I love yum-utils, but I was talking about all the yum plugins
packages
>>out there :
[...]
>$ rpm -q --qf '%{sourcerpm}\n' yum-skip-broken
>yum-utils-1.1.1-1.fc7.src.rpm
I'm not sure I get your point. Are you saying all the 17 plugins were
written by the same person ? My point is: most of the plugins should be
My point is, these plugins are all collected as part of the yum-utils rpm
already. There's little reason to move them to the yum package.
part of the core yum package and therefore installed by default on
all
fedora distros ('installed' not being equal to 'enabled by default'),
and people - community members - should be working on making yum better,
rather than writing plugins to work around the known yum limitations.
If you want them installed by default, why not just advocate for that?
That's a whole separate issue.
--
Matthew Miller mattdm(a)mattdm.org <http://mattdm.org/>
Boston University Linux ------> <http://linux.bu.edu/>
11:43 a.m.
New subject: FC6 updates broken deps?
Denis Leroy wrote:
Matthew Miller wrote:
> On Fri, Feb 23, 2007 at 10:54:18AM +0100, Denis Leroy wrote:
>>> But all these things are already in yum-utils.
>> I love yum-utils, but I was talking about all the yum plugins
>> packages out there :
>> # yum list | grep ^yum- | wc -l
>> 17
>
> $ rpm -q --qf '%{sourcerpm}\n' yum-skip-broken
> yum-utils-1.1.1-1.fc7.src.rpm
I'm not sure I get your point. Are you saying all the 17 plugins were
written by the same person ? My point is: most of the plugins should be
part of the core yum package and therefore installed by default on all
fedora distros ('installed' not being equal to 'enabled by default'),
and people - community members - should be working on making yum better,
rather than writing plugins to work around the known yum limitations.
Does extensions in Firefox extend the capabilities of Firefox or work
around its limitations?
Rahul
1:38 p.m.
New subject: FC6 updates broken deps?
Hi.
Rahul Sundaram <sundaram(a)fedoraproject.org> wrote:
Does extensions in Firefox extend the capabilities of Firefox or work
around its limitations?
That's a question of definitons.
Both, if you like.
--
TRUST NO ONE.
KEEP YOUR LASER HANDY.
THE COMPUTER IS YOUR FRIEND.
-- "Paranoia," Avalon Hill
11:50 a.m.
New subject: FC6 updates broken deps?
On Thu, 2007-02-22 at 09:35 -0800, Florin Andrei wrote:
Thomas M Steenholdt wrote:
> Really - The reason dependency issues are such a big issue for fedora is
> that all updates are held back because of a single package missing a
> dependency. I'm sure everybody would feel better holding back one
> package instead of all updates on this account.
If I'm not mistaken, yum seems fairly unique in this regard.
I mean, heck, look at Microsoft for example. Their update thing applies
as many patches as possible, and those that cannot be applied, well duh,
they don't get applied and the user is notified by big honking red icons
that something failed.
You don't need to go as far as pointing to a broken update service like
MS's one.
Other Linux distributions (Ubuntu) update as much as possible and warn
the user that some packages can't be updated (it also goes as far as
telling you that you should do a dist-upgrade instead of just an upgrade
in some cases), and they stay in the list of updates as long as the
problem is not fixed.
And it makes perfectly sense.
I had on one of my systems a custom package that kept another one from
being updated and it was OK for me. I would have been pissed off if I
had to manually update or had to remove and reinstall my package each
time to let the update process get through for all the other gazillion
of packages.
But a single error doesn't nuke the whole process. yum is
_really_
boneheaded from this perspective. Your argument that a possibly much
more important security update might get masked by a dependency error
and is not applied is a very serious one and I don't think there can be
any logical rebuttal.
+1
yum needs to be fixed, NOW. At this moment the update process puts
everyone at risk.
+1
Simo.
--
Simo Sorce <ssorce(a)redhat.com>
Sr Software Engineer
Base Operating Systems
Red Hat Inc.
12:46 p.m.
New subject: FC6 updates broken deps?
>>>> "FA" == Florin Andrei
<florin(a)andrei.myip.org> writes:
FA> If I'm not mistaken, yum seems fairly unique in this regard.
FA> I mean, heck, look at Microsoft for example. Their update thing
FA> applies as many patches as possible, and those that cannot be
FA> applied, well duh, they don't get applied and the user is notified
FA> by big honking red icons that something failed.
Yum just doesn't follow the Unix commandline tradition. Unix commands
should succeed quietly, and only show output if it is something the
user asks for. They don't ask whether you really mean it when you tell
them to do something -- heck even cp needs -i to ask whether you want
to clobber a file, and that one is AFAIK a GNUism. They don't block
SIGINT except when absolutely needed, and then only for a short time.
Yum's interface is made for a GUI, where all this interactivity makes
sense. The default should be to do exactly what the user asked, unless
it is impossible. That is, it should take an extra switch to let it
upgrade even when there are broken dependencies, but that extra switch
should not be -y. -y is almost as bad as --force for rpm. There should
probably be a switches to stop yum from upgrading or removing packages
when doing an install.
Add -i and --progress switches to get back the current behaviour, if
people can't live without it. The --progress thing would actually make
sense, if package fetching is slow sometimes.
/Benny
(Sorry for the gripe. Yum is a godsend, it is just the interface which
is a pain)
3:45 p.m.
New subject: FC6 updates broken deps?
Am Donnerstag, den 22.02.2007, 19:46 +0100 schrieb Benny Amorsen:
Yum just doesn't follow the Unix commandline tradition. Unix
commands
should succeed quietly, and only show output if it is something the
user asks for.
Yes you're right. Yum's current mode should be the verbose mode.
A benefit of showing stuff is that it makes the command seem faster.
When should yum be quiet?
# yum install package
Should be quiet unless a dependency needs to be pulled in, otherwise it
should fail.
To counter this:
# yum -r install package
could resolve dependencies.
Shall I file an RFE?
9:58 p.m.
New subject: FC6 updates broken deps?
On Thu, 2007-02-22 at 19:46 +0100, Benny Amorsen wrote:
>>>>> "FA" == Florin Andrei
<florin(a)andrei.myip.org> writes:
FA> If I'm not mistaken, yum seems fairly unique in this regard.
FA> I mean, heck, look at Microsoft for example. Their update thing
FA> applies as many patches as possible, and those that cannot be
FA> applied, well duh, they don't get applied and the user is notified
FA> by big honking red icons that something failed.
Yum just doesn't follow the Unix commandline tradition. Unix commands
should succeed quietly, and only show output if it is something the
user asks for. They don't ask whether you really mean it when you tell
them to do something -- heck even cp needs -i to ask whether you want
to clobber a file, and that one is AFAIK a GNUism. They don't block
SIGINT except when absolutely needed, and then only for a short time.
Aya.
Blocking SIGINT is one of my "favorites".
Having to Ctrl-Z + kill -9 %% every time I need to kill yum is -very-
annoying.
(Ignoring, for a second, that lack of consistency with other command
line application.
Yum's interface is made for a GUI, where all this interactivity makes
sense. The default should be to do exactly what the user asked, unless
it is impossible. That is, it should take an extra switch to let it
upgrade even when there are broken dependencies, but that extra switch
should not be -y. -y is almost as bad as --force for rpm. There should
probably be a switches to stop yum from upgrading or removing packages
when doing an install.
Add -i and --progress switches to get back the current behaviour, if
people can't live without it. The --progress thing would actually make
sense, if package fetching is slow sometimes.
--skip upgrade should be enabled by default.
Yum should show up a list of packages that are about to upgraded, and
more importantly, packages that are skipped. (With an option to see why
package X has been skipped)
Did I mention that Ctrl-C should be kill yum and -not- switch mirrors??!?!??!
/Benny
(Sorry for the gripe. Yum is a godsend, it is just the interface which
is a pain)
2:12 a.m.
New subject: FC6 updates broken deps?
Hi.
On Fri, 23 Feb 2007 05:58:21 +0200, Gilboa Davara wrote:
Did I mention that Ctrl-C should be kill yum and -not- switch
mirrors??!?!??!
That is actually useful. Typing it twice or thrice should kill
yum, however, at least it did some time ago.
3:17 a.m.
New subject: FC6 updates broken deps?
On Fri, 2007-02-23 at 09:12 +0100, Ralf Ertzinger wrote:
Hi.
On Fri, 23 Feb 2007 05:58:21 +0200, Gilboa Davara wrote:
> Did I mention that Ctrl-C should be kill yum and -not- switch
> mirrors??!?!??!
That is actually useful.
I disagree - All programs should immediately terminate
upon "Ctrl-C".
Switching to a different mirror should be done by any arbitrary key.
Ralf
3:24 a.m.
New subject: FC6 updates broken deps?
Hi.
On Fri, 23 Feb 2007 10:17:23 +0100, Ralf Corsepius wrote:
> That is actually useful.
I disagree - All programs should immediately terminate upon "Ctrl-C".
Switching to a different mirror should be done by any arbitrary key.
Oh, you can move that to a different key for all I care (SIGINT?), as
long as the functionality itself is there.
5 a.m.
New subject: FC6 updates broken deps?
On Fri, Feb 23, 2007 at 10:17:23AM +0100, Ralf Corsepius wrote:
I disagree - All programs should immediately terminate upon
"Ctrl-C".
Switching to a different mirror should be done by any arbitrary key.
Someone please fix emacs to terminate immediately on ctrl-C then ;)
Very large numbers of programs override ^C to be an internal interrupt,
including things like ftp. Others specifically ignore it (try using ssh
without that)
7:28 a.m.
New subject: FC6 updates broken deps?
Alan Cox wrote(a):
On Fri, Feb 23, 2007 at 10:17:23AM +0100, Ralf Corsepius wrote:
> I disagree - All programs should immediately terminate upon "Ctrl-C".
> Switching to a different mirror should be done by any arbitrary key.
Someone please fix emacs to terminate immediately on ctrl-C then ;)
Very large numbers of programs override ^C to be an internal interrupt,
including things like ftp. Others specifically ignore it (try using ssh
without that)
I think we can accept such weirdness of some applications.
Most people will understand WHY ^c is overriden in ssh for example.
For me after 10 years of linux experience it was great mistery why a hell
does yum override ^c.
There is no clear reason for it to behave like that.
My suggestion ?
^m and a startup message like:
"If you want to switch from one mirror to another press CTRL+M"
regards
--
Piotr Baranowski - RHCX @ Altkom Akademia S.A.
Dlaczego informatycy mylą Halloween z Bożym Narodzeniem ?
...bo 31oct==25dec :)
7:37 a.m.
New subject: FC6 updates broken deps?
Piotr Baranowski wrote:
Alan Cox wrote(a):
> On Fri, Feb 23, 2007 at 10:17:23AM +0100, Ralf Corsepius wrote:
>> I disagree - All programs should immediately terminate upon "Ctrl-C".
>> Switching to a different mirror should be done by any arbitrary key.
> Someone please fix emacs to terminate immediately on ctrl-C then ;)
>
> Very large numbers of programs override ^C to be an internal interrupt,
> including things like ftp. Others specifically ignore it (try using ssh
> without that)
I think we can accept such weirdness of some applications.
Most people will understand WHY ^c is overriden in ssh for example.
For me after 10 years of linux experience it was great mistery why a hell
does yum override ^c.
There is no clear reason for it to behave like that.
My suggestion ?
^m and a startup message like:
"If you want to switch from one mirror to another press CTRL+M"
regards
That would probably also help the fact, that (last i checked) ^C does
switch mirrors, but it also prevents anything useful from happening at
the end of the download process. I.e. yum -y update would download
packages, ^C cause it to switch mirrors and keep downloading, but the
expected update at the end of the download is skipped.
/Thomas
3:33 a.m.
New subject: FC6 updates broken deps?
On Fri, 2007-02-23 at 14:37 +0100, Thomas M Steenholdt wrote:
Piotr Baranowski wrote:
> Alan Cox wrote(a):
>> On Fri, Feb 23, 2007 at 10:17:23AM +0100, Ralf Corsepius wrote:
>>> I disagree - All programs should immediately terminate upon
"Ctrl-C".
>>> Switching to a different mirror should be done by any arbitrary key.
>> Someone please fix emacs to terminate immediately on ctrl-C then ;)
>>
>> Very large numbers of programs override ^C to be an internal interrupt,
>> including things like ftp. Others specifically ignore it (try using ssh
>> without that)
>
> I think we can accept such weirdness of some applications.
>
> Most people will understand WHY ^c is overriden in ssh for example.
>
> For me after 10 years of linux experience it was great mistery why a hell
> does yum override ^c.
>
> There is no clear reason for it to behave like that.
>
> My suggestion ?
> ^m and a startup message like:
>
> "If you want to switch from one mirror to another press CTRL+M"
>
> regards
>
That would probably also help the fact, that (last i checked) ^C does
switch mirrors, but it also prevents anything useful from happening at
the end of the download process. I.e. yum -y update would download
packages, ^C cause it to switch mirrors and keep downloading, but the
expected update at the end of the download is skipped.
/Thomas
/+1 Switch mirror doesn't always work.
BTW, there's an assortments of bugs about it. [1]
- Gilboa
[1] https://bugzilla.redhat.com/bugzilla/buglist.cgi?product=Fedora
+Core&component=yum&bug_status=NEW&bug_status=VERIFIED&bug_status=ASSIGNED&bug_status=REOPENED&bug_status=CLOSED&bug_status=NEEDINFO&bug_status=MODIFIED&bug_status=ON_DEV&bug_status=ON_QA&bug_status=FAILS_QA&bug_status=RELEASE_PENDING&short_desc_type=allwordssubstr&short_desc=&long_desc_type=allwordssubstr&long_desc=Ctrl-C
7:55 a.m.
New subject: FC6 updates broken deps?
On Friday 23 February 2007 08:28, Piotr Baranowski wrote:
For me after 10 years of linux experience it was great mistery why a
hell
does yum override ^c.
I do believe it has something to do with interaction with rpmlib and rpmlib
trapping and doing nothing with the ^c while it does stuff with the DB. I
could be wrong, but I'm pretty sure that's what is going on.
--
Jesse Keating
Release Engineer: Fedora
8:39 a.m.
New subject: FC6 updates broken deps?
On Fri, 2007-02-23 at 08:55 -0500, Jesse Keating wrote:
On Friday 23 February 2007 08:28, Piotr Baranowski wrote:
> For me after 10 years of linux experience it was great mistery why a hell
> does yum override ^c.
I do believe it has something to do with interaction with rpmlib and rpmlib
trapping and doing nothing with the ^c while it does stuff with the DB. I
could be wrong, but I'm pretty sure that's what is going on.
By a brief look at the rpmlib source it seems so but I didn't study it
thoroughly.
I understand the necessity to block the signals when db is opened but
perhaps rpmlib should block the signal so it can be delivered later when
it is unblocked and not just ignore it.
--
Tomas Mraz
No matter how far down the wrong road you've gone, turn back.
Turkish proverb
8:28 a.m.
New subject: FC6 updates broken deps?
On Fri, Feb 23, 2007 at 02:28:04PM +0100, Piotr Baranowski wrote:
Most people will understand WHY ^c is overriden in ssh for example.
For me after 10 years of linux experience it was great mistery why a hell
does yum override ^c.
It behaves much like ftp in that respect
"If you want to switch from one mirror to another press
CTRL+M"
Not a good idea - ^M is newline 8)
9:20 a.m.
New subject: FC6 updates broken deps?
On Fri, 2007-02-23 at 09:28 -0500, Alan Cox wrote:
> "If you want to switch from one mirror to another press
CTRL+M"
Not a good idea - ^M is newline 8)
So make that Ctrl plus any other character that isn't a control sequence
already. You could also make it Ctrl+Alt+Backspace so that the yum
maintainers finally get their "yum kills X" bug reports they must have
been missing for years ;-P.
Nils
--
Nils Philippsen / Red Hat / nphilipp(a)redhat.com
"Those who would give up Essential Liberty to purchase a little Temporary
Safety, deserve neither Liberty nor Safety." -- B. Franklin, 1759
PGP fingerprint: C4A8 9474 5C4C ADE3 2B8F 656D 47D8 9B65 6951 3011
2:24 p.m.
New subject: yum and Ctl-C (was Re: FC6 updates broken deps?
At 2:28 PM +0100 2/23/07, Piotr Baranowski wrote:
Alan Cox wrote(a):
> On Fri, Feb 23, 2007 at 10:17:23AM +0100, Ralf Corsepius wrote:
>> I disagree - All programs should immediately terminate upon "Ctrl-C".
>> Switching to a different mirror should be done by any arbitrary key.
>
> Someone please fix emacs to terminate immediately on ctrl-C then ;)
>
> Very large numbers of programs override ^C to be an internal interrupt,
> including things like ftp. Others specifically ignore it (try using ssh
> without that)
I think we can accept such weirdness of some applications.
Most people will understand WHY ^c is overriden in ssh for example.
For me after 10 years of linux experience it was great mistery why a hell
does yum override ^c.
...
Yum does not override Ctl-Ç. RPM does. Look at the RPM source, say
rpmsq.c. RPM does not block the signals, but rather saves them and if a
signal has been caught, terminates with extreme prejudice at the end of
whatever it was doing. Thus, signals are not processed immediately and
always terminate RPM's operation.
Not to start any wars here, but I wrote a yum plugin for FC5 (and just
updated it for FC6, laziness on my part) that provides, among other things,
sane Ctl-C handling (during downloading) in a way I believe safe for RPM.
<http://georgeanelson.com/stablemirror.htm>
--
____________________________________________________________________
TonyN.:' <mailto:tonynelson@georgeanelson.com>
' <http://www.georgeanelson.com/>
3:16 p.m.
New subject: yum and Ctl-C (was Re: FC6 updates broken deps?
Am Freitag, den 23.02.2007, 15:24 -0500 schrieb Tony Nelson:
At 2:28 PM +0100 2/23/07, Piotr Baranowski wrote:
>Alan Cox wrote(a):
>> On Fri, Feb 23, 2007 at 10:17:23AM +0100, Ralf Corsepius wrote:
>>> I disagree - All programs should immediately terminate upon
"Ctrl-C".
>>> Switching to a different mirror should be done by any arbitrary key.
>>
>> Someone please fix emacs to terminate immediately on ctrl-C then ;)
>>
>> Very large numbers of programs override ^C to be an internal interrupt,
>> including things like ftp. Others specifically ignore it (try using ssh
>> without that)
>
>I think we can accept such weirdness of some applications.
>
>Most people will understand WHY ^c is overriden in ssh for example.
>
>For me after 10 years of linux experience it was great mistery why a hell
>does yum override ^c.
...
Yum does not override Ctl-Ç. RPM does. Look at the RPM source, say
rpmsq.c. RPM does not block the signals, but rather saves them and if a
signal has been caught, terminates with extreme prejudice at the end of
whatever it was doing. Thus, signals are not processed immediately and
always terminate RPM's operation.
Not to start any wars here, but I wrote a yum plugin for FC5 (and just
updated it for FC6, laziness on my part) that provides, among other things,
sane Ctl-C handling (during downloading) in a way I believe safe for RPM.
<http://georgeanelson.com/stablemirror.htm>
--
____________________________________________________________________
TonyN.:' <mailto:tonynelson@georgeanelson.com>
' <http://www.georgeanelson.com/>
Err.. wasn't this fixed recently, so that ^C exits rather than switches
mirror?
3:38 p.m.
New subject: yum and Ctl-C (was Re: FC6 updates broken deps?
On Fri, 23 Feb 2007 22:16:31 +0100, nodata <lsof(a)nodata.co.uk> wrote:
Err.. wasn't this fixed recently, so that ^C exits rather than
switches
mirror?
No, this wasn't changed. Some people LIKE the "switch mirror"
behaviour. I think the ideal approach would be to emulate what
mpg123 does. If you hit ^C once, it goes to the next track, but
if you hit it twice, it exits. However, implementing this in yum
was seen as unreliable. So, instead, yum was made to exit at
the SIGQUIT (^\ on keyboard).
-- Pete
3:45 p.m.
New subject: yum and Ctl-C (was Re: FC6 updates broken deps?
Am Freitag, den 23.02.2007, 13:38 -0800 schrieb Pete Zaitcev:
On Fri, 23 Feb 2007 22:16:31 +0100, nodata <lsof(a)nodata.co.uk>
wrote:
> Err.. wasn't this fixed recently, so that ^C exits rather than switches
> mirror?
No, this wasn't changed. Some people LIKE the "switch mirror"
behaviour. I think the ideal approach would be to emulate what
mpg123 does. If you hit ^C once, it goes to the next track, but
if you hit it twice, it exits. However, implementing this in yum
was seen as unreliable. So, instead, yum was made to exit at
the SIGQUIT (^\ on keyboard).
-- Pete
I'm confused.
"this should be fixed in 2.9.6 - due out soon."
-- https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=139459
4:39 p.m.
New subject: yum and Ctl-C (was Re: FC6 updates broken deps?
At 1:38 PM -0800 2/23/07, Pete Zaitcev wrote:
On Fri, 23 Feb 2007 22:16:31 +0100, nodata <lsof(a)nodata.co.uk>
wrote:
> Err.. wasn't this fixed recently, so that ^C exits rather than switches
> mirror?
Not on my system, which for months now has been using yum without
stablemirror. Ctl-C sometimes would switch mirrors and sometimes would
cause yum to exit, as others have been complaining. From looking at yum
and the libraries it uses, the behavior seems to me to be accidental.
No, this wasn't changed. Some people LIKE the "switch
mirror"
behaviour. I think the ideal approach would be to emulate what
mpg123 does. If you hit ^C once, it goes to the next track, but
if you hit it twice, it exits. However, implementing this in yum
was seen as unreliable. So, instead, yum was made to exit at
the SIGQUIT (^\ on keyboard).
I like to choose between switching mirrors or quitting yum when I press
Ctl-C. Stablemirror seems to do it reliably, using the
YumRepository.setupGrab.grabfunc.opts.interrupt_callback hook. Yum's
top-level KeyboardInterrupt handler never gets the exception -- it usually
didn't anyway, as RPM had already siezed SIGINT.
--
____________________________________________________________________
TonyN.:' <mailto:tonynelson@georgeanelson.com>
' <http://www.georgeanelson.com/>
4:04 a.m.
New subject: yum and Ctl-C (was Re: FC6 updates broken deps?
Pete Zaitcev wrote:
On Fri, 23 Feb 2007 22:16:31 +0100, nodata <lsof(a)nodata.co.uk>
wrote:
> Err.. wasn't this fixed recently, so that ^C exits rather than switches
> mirror?
No, this wasn't changed. Some people LIKE the "switch mirror"
behaviour. I think the ideal approach would be to emulate what
mpg123 does. If you hit ^C once, it goes to the next track, but
if you hit it twice, it exits. However, implementing this in yum
was seen as unreliable. So, instead, yum was made to exit at
the SIGQUIT (^\ on keyboard).
-- Pete
I wonder if anyone knows how to make that char on a danisk keyboard... \
is AltGr+[ backslash key ] adding a ctrl to that combo just doesnt work.
Perhaps it's doable, perhaps it's not?!? Anyone?
/Thomas
7:02 p.m.
New subject: FC6 updates broken deps?
>>>> "AC" == Alan Cox <alan(a)redhat.com>
writes:
AC> Someone please fix emacs to terminate immediately on ctrl-C then
AC> ;)
AC> Very large numbers of programs override ^C to be an internal
AC> interrupt, including things like ftp. Others specifically ignore
AC> it (try using ssh without that)
emacs, ftp in command mode, and ssh are all interactive. yum has no
reason to be interactive, it does not offer a command prompt or
connect you to another system with a command prompt.
/Benny
7:08 p.m.
New subject: FC6 updates broken deps?
Benny Amorsen wrote:
<snip
It is shell like :)
Cheers
emacs, ftp in command mode, and ssh are all interactive. yum has no
reason to be interactive, it does not offer a command prompt or
connect you to another system with a command prompt.
yum shell:
# yum shell
Loading "protectbase" plugin
Loading "installonlyn" plugin
Loading "fastestmirror" plugin
Setting up Yum Shell
7:34 p.m.
New subject: FC6 updates broken deps?
Mike Kearey napisał(a):
Benny Amorsen wrote:
<snip>
>
> emacs, ftp in command mode, and ssh are all interactive. yum has no
> reason to be interactive, it does not offer a command prompt or
> connect you to another system with a command prompt.
yum shell:
# yum shell
Loading "protectbase" plugin
Loading "installonlyn" plugin
Loading "fastestmirror" plugin
Setting up Yum Shell
Anybody really using that stuff ?
Functionality seems to be cool but any real-life scenarios of yum shell
monkey-super-powers in action ?
--
Piotr Baranowski - RHCX @ Altkom Akademia S.A.
Dlaczego informatycy mylą Halloween z Bożym Narodzeniem ?
...bo 31oct==25dec :)
7:51 p.m.
New subject: FC6 updates broken deps?
Piotr Baranowski wrote:
Anybody really using that stuff ?
Functionality seems to be cool but any real-life scenarios of yum shell
monkey-super-powers in action ?
One scenario I recently used it for was to install an un-signed rpm
package from a trustworthy source. It saved me having to edit
/etc/yum.conf to turn off gpgcheck and then do yum localinstall .ie :
yum shell
...
config gpgcheck 0
locallinstall /path/to/rpm/file.rpm
It really just amounts to 'yet another way to do things' I can imagine
there are more monkey-super-powers in there I have not thought of..
This is not really development discussion however, so this is where my
posting ends on this subject :)
Cheers,
Michael
7:51 p.m.
New subject: FC6 updates broken deps?
>>>> "MK" == Mike Kearey
<mkearey(a)redhat.com> writes:
MK> Benny Amorsen wrote: <snip>
> emacs, ftp in command mode, and ssh are all interactive. yum
has
> no reason to be interactive, it does not offer a command prompt or
> connect you to another system with a command prompt.
MK> yum shell:
MK> # yum shell Loading "protectbase" plugin Loading
"installonlyn"
MK> plugin Loading "fastestmirror" plugin Setting up Yum Shell
>
MK> It is shell like :)
I had no idea that existed. Thanks for the information. yum can
certainly intercept Ctrl-C when it is running in interactive mode. It
should still keep its hands off in command mode.
/Benny
8 p.m.
New subject: FC6 updates broken deps?
On Sun, 2007-02-25 at 02:02 +0100, Benny Amorsen wrote:
[...] yum has no
reason to be interactive, it does not offer a command prompt or
connect you to another system with a command prompt.
Nope. `yum shell` gives you a nice interactive interface. :)
--
Peter Gordon (codergeek42) / FSF Associate Member #5015
GnuPG Public Key ID: 0xFFC19479 / Fingerprint:
DD68 A414 56BD 6368 D957 9666 4268 CB7A FFC1 9479
Blog: http://thecodergeek.com/blog/
About: http://fedoraproject.org/wiki/PeterGordon
5:55 a.m.
New subject: FC6 updates broken deps?
On Fri, 23 Feb 2007 09:12:55 +0100, Ralf Ertzinger wrote:
Hi.
On Fri, 23 Feb 2007 05:58:21 +0200, Gilboa Davara wrote:
> Did I mention that Ctrl-C should be kill yum and -not- switch
> mirrors??!?!??!
That is actually useful. Typing it twice or thrice should kill
yum, however, at least it did some time ago.
A common scenario for newbies, which make you bow your head in shame: You
had thought they would prefer the graphical package tools, you have not
recommended the command-line, but they know Google and have found lots of
documentation about how to install something with Yum. And what happens?
Metadata checksum failures on apparently all mirrors. ^C doesn't help.
Yum switches through all available mirrors. WTF? Some time later, 78
updates are available. Newbie is caught in a loop of download failures,
mirror-switching, ^C attempts, low through-put, and reacts very surprised
and disappointed about "this Linux stuff".
11:20 a.m.
New subject: FC6 updates broken deps?
Michael Schwendt wrote:
A common scenario for newbies, which make you bow your head in shame: You
had thought they would prefer the graphical package tools, you have not
recommended the command-line, but they know Google and have found lots of
documentation about how to install something with Yum. And what happens?
Metadata checksum failures on apparently all mirrors. ^C doesn't help.
Yum switches through all available mirrors. WTF? Some time later, 78
updates are available. Newbie is caught in a loop of download failures,
mirror-switching, ^C attempts, low through-put, and reacts very surprised
and disappointed about "this Linux stuff".
So true. :-(
--
Florin Andrei
http://florin.myip.org/
3:51 p.m.
New subject: FC6 updates broken deps?
Florin Andrei wrote:
Thomas M Steenholdt wrote:
> Really - The reason dependency issues are such a big issue for fedora
> is that all updates are held back because of a single package
> missing a dependency. I'm sure everybody would feel better holding
> back one package instead of all updates on this account.
If I'm not mistaken, yum seems fairly unique in this regard.
I mean, heck, look at Microsoft for example. Their update thing
applies as many patches as possible, and those that cannot be applied,
well duh, they don't get applied and the user is notified by big
honking red icons that something failed.
But a single error doesn't nuke the whole process. yum is _really_
boneheaded from this perspective. Your argument that a possibly much
more important security update might get masked by a dependency error
and is not applied is a very serious one and I don't think there can
be any logical rebuttal.
yum needs to be fixed, NOW. At this moment the update process puts
everyone at risk.
correct.
I think we should agree that yum _is_ broken here and needs to be fixed.
and no including a optional plugin won't help, the default behavior is
broken ( I would even say its a security bug)
12:06 a.m.
New subject: FC6 updates broken deps?
On Fri, 2007-02-23 at 22:51 +0100, dragoran wrote:
I think we should agree that yum _is_ broken here and needs to be
fixed.
and no including a optional plugin won't help, the default behavior is
broken ( I would even say its a security bug)
Does anyone else have an odd feeling of deja-vu?
5 a.m.
New subject: FC6 updates broken deps?
Callum Lerwick wrote:
On Fri, 2007-02-23 at 22:51 +0100, dragoran wrote:
> I think we should agree that yum _is_ broken here and needs to be fixed.
> and no including a optional plugin won't help, the default behavior is
> broken ( I would even say its a security bug)
>
Does anyone else have an odd feeling of deja-vu?
I now that we had this discussion in the past, but the problem is still
there so it will keeps coming up again until its fixed.
3:02 p.m.
New subject: FC6 updates broken deps?
On Sat, 24 Feb 2007, Callum Lerwick wrote:
On Fri, 2007-02-23 at 22:51 +0100, dragoran wrote:
> I think we should agree that yum _is_ broken here and needs to be fixed.
> and no including a optional plugin won't help, the default behavior is
> broken ( I would even say its a security bug)
Does anyone else have an odd feeling of deja-vu?
Yes, and a strong one at that.
This and "but it shouldn't abort when unresolvable dependencies are
present" threads (among various other non-yum related ones) keep recurring
every few months.
Somebody with too much time on their hands, analysis of these threads
might be interesting subject: do these things recur N months before/after
release, only during certain months when it's cold as hell in certain
parts of the world or what :)
- Panu -
8:47 p.m.
New subject: FC6 updates broken deps?
Panu Matilainen wrote:
On Sat, 24 Feb 2007, Callum Lerwick wrote:
> On Fri, 2007-02-23 at 22:51 +0100, dragoran wrote:
>> I think we should agree that yum _is_ broken here and needs to be fixed.
>> and no including a optional plugin won't help, the default behavior is
>> broken ( I would even say its a security bug)
>
> Does anyone else have an odd feeling of deja-vu?
Yes, and a strong one at that.
This and "but it shouldn't abort when unresolvable dependencies are
present" threads (among various other non-yum related ones) keep
recurring every few months.
Perhaps there is a connection with the fact that yum _is_ broken?
I suggest you should study the relation between cause and effect, it's
really illuminating once you learn it.
http://en.wikipedia.org/wiki/Causality
--
Florin Andrei
http://florin.myip.org/
3:37 a.m.
New subject: FC6 updates broken deps?
On Sat, 2007-02-24 at 00:06 -0600, Callum Lerwick wrote:
On Fri, 2007-02-23 at 22:51 +0100, dragoran wrote:
> I think we should agree that yum _is_ broken here and needs to be fixed.
> and no including a optional plugin won't help, the default behavior is
> broken ( I would even say its a security bug)
Does anyone else have an odd feeling of deja-vu?
It's so annoying that I'm thinking about learning python just so I can
submit a patch to fix it... :(
- Gilboa
8:34 a.m.
New subject: FC6 updates broken deps?
Gilboa Davara wrote:
On Sat, 2007-02-24 at 00:06 -0600, Callum Lerwick wrote:
> On Fri, 2007-02-23 at 22:51 +0100, dragoran wrote:
>> I think we should agree that yum _is_ broken here and needs to be fixed.
>> and no including a optional plugin won't help, the default behavior is
>> broken ( I would even say its a security bug)
> Does anyone else have an odd feeling of deja-vu?
It's so annoying that I'm thinking about learning python just so I can
submit a patch to fix it... :(
While learning Python is a great idea, writing yum patches is
unfortunately known to be a complete and utter waste of time :-(
8:38 a.m.
New subject: FC6 updates broken deps?
Denis Leroy wrote:
Gilboa Davara wrote:
> On Sat, 2007-02-24 at 00:06 -0600, Callum Lerwick wrote:
>> On Fri, 2007-02-23 at 22:51 +0100, dragoran wrote:
>>> I think we should agree that yum _is_ broken here and needs to be
>>> fixed.
>>> and no including a optional plugin won't help, the default behavior
>>> is broken ( I would even say its a security bug)
>> Does anyone else have an odd feeling of deja-vu?
>
> It's so annoying that I'm thinking about learning python just so I can
> submit a patch to fix it... :(
While learning Python is a great idea, writing yum patches is
unfortunately known to be a complete and utter waste of time :-(
What patches did you submit?
Rahul
8:38 a.m.
New subject: FC6 updates broken deps?
On Mon, 2007-02-26 at 15:34 +0100, Denis Leroy wrote:
Gilboa Davara wrote:
> On Sat, 2007-02-24 at 00:06 -0600, Callum Lerwick wrote:
>> On Fri, 2007-02-23 at 22:51 +0100, dragoran wrote:
>>> I think we should agree that yum _is_ broken here and needs to be fixed.
>>> and no including a optional plugin won't help, the default behavior is
>>> broken ( I would even say its a security bug)
>> Does anyone else have an odd feeling of deja-vu?
>
> It's so annoying that I'm thinking about learning python just so I can
> submit a patch to fix it... :(
While learning Python is a great idea, writing yum patches is
unfortunately known to be a complete and utter waste of time :-(
It is? Since when?
-sv
8:43 a.m.
New subject: FC6 updates broken deps?
Denis Leroy wrote:
Gilboa Davara wrote:
> On Sat, 2007-02-24 at 00:06 -0600, Callum Lerwick wrote:
>> On Fri, 2007-02-23 at 22:51 +0100, dragoran wrote:
>>> I think we should agree that yum _is_ broken here and needs to be
>>> fixed.
>>> and no including a optional plugin won't help, the default behavior
>>> is broken ( I would even say its a security bug)
>> Does anyone else have an odd feeling of deja-vu?
>
> It's so annoying that I'm thinking about learning python just so I can
> submit a patch to fix it... :(
While learning Python is a great idea, writing yum patches is
unfortunately known to be a complete and utter waste of time :-(
Correcting statement above:
s/patches/patches that add new features to yum/
8:48 a.m.
New subject: FC6 updates broken deps?
On Mon, 2007-02-26 at 15:43 +0100, Denis Leroy wrote:
Denis Leroy wrote:
> Gilboa Davara wrote:
>> On Sat, 2007-02-24 at 00:06 -0600, Callum Lerwick wrote:
>>> On Fri, 2007-02-23 at 22:51 +0100, dragoran wrote:
>>>> I think we should agree that yum _is_ broken here and needs to be
>>>> fixed.
>>>> and no including a optional plugin won't help, the default behavior
>>>> is broken ( I would even say its a security bug)
>>> Does anyone else have an odd feeling of deja-vu?
>>
>> It's so annoying that I'm thinking about learning python just so I can
>> submit a patch to fix it... :(
>
> While learning Python is a great idea, writing yum patches is
> unfortunately known to be a complete and utter waste of time :-(
Correcting statement above:
s/patches/patches that add new features to yum/
?? What're you talking about?
-sv
8:59 a.m.
New subject: FC6 updates broken deps?
seth vidal wrote:
On Mon, 2007-02-26 at 15:43 +0100, Denis Leroy wrote:
> Denis Leroy wrote:
>> Gilboa Davara wrote:
>>> On Sat, 2007-02-24 at 00:06 -0600, Callum Lerwick wrote:
>>>> On Fri, 2007-02-23 at 22:51 +0100, dragoran wrote:
>>>>> I think we should agree that yum _is_ broken here and needs to be
>>>>> fixed.
>>>>> and no including a optional plugin won't help, the default
behavior
>>>>> is broken ( I would even say its a security bug)
>>>> Does anyone else have an odd feeling of deja-vu?
>>> It's so annoying that I'm thinking about learning python just so I
can
>>> submit a patch to fix it... :(
>> While learning Python is a great idea, writing yum patches is
>> unfortunately known to be a complete and utter waste of time :-(
> Correcting statement above:
>
> s/patches/patches that add new features to yum/
?? What're you talking about?
Maybe I'm mistaken. If I submit a patch that integrates --skip-broken
into yum, will you accept it ? I'd be more than happy to write that patch.
9:02 a.m.
New subject: FC6 updates broken deps?
On Mon, 2007-02-26 at 15:59 +0100, Denis Leroy wrote:
Maybe I'm mistaken. If I submit a patch that integrates
--skip-broken
into yum, will you accept it ? I'd be more than happy to write that patch.
It really depends on how the patch is implemented. Though I'm curious
why integrating it is important if the plugin works well enough?
-sv
9:29 a.m.
New subject: FC6 updates broken deps?
seth vidal wrote:
On Mon, 2007-02-26 at 15:59 +0100, Denis Leroy wrote:
> Maybe I'm mistaken. If I submit a patch that integrates --skip-broken
> into yum, will you accept it ? I'd be more than happy to write that patch.
It really depends on how the patch is implemented. Though I'm curious
why integrating it is important if the plugin works well enough?
let me try to answer that.
1. because the plugin is not installed by default, much less enabled by
default. As a new user, I would be confused by a large update failing
because of a single broken dep. My instinctive reaction would be to look
for an option in yum to change that behavior (assuming i even know about
yum, rather than just using the update applet). Having to launch pirut
and search for a plugin seems counter-intuitive to me.
2. I find that strictly relying on plugins to work around issues, change
default behaviors and add new functionality, or more generally accepting
contributions from the community, is a strange way of managing a
project. At least that's the way I understand yum is managed right now,
based on what Tom Callaway said sunday at FOSDEM. But if you tell me
that's not the case, I'll certainly believe you.
9:45 a.m.
New subject: FC6 updates broken deps?
On Mon, 2007-02-26 at 16:29 +0100, Denis Leroy wrote:
let me try to answer that.
1. because the plugin is not installed by default, much less enabled by
default. As a new user, I would be confused by a large update failing
because of a single broken dep. My instinctive reaction would be to look
for an option in yum to change that behavior (assuming i even know about
yum, rather than just using the update applet). Having to launch pirut
and search for a plugin seems counter-intuitive to me.
2. I find that strictly relying on plugins to work around issues,
change
default behaviors and add new functionality, or more generally accepting
contributions from the community, is a strange way of managing a
project. At least that's the way I understand yum is managed right now,
based on what Tom Callaway said sunday at FOSDEM. But if you tell me
that's not the case, I'll certainly believe you.
I don't know what Spot said at fosdem so I can't really comment on that.
Mostly we've been focusing on optimization, fixing bugs and changing a
fair bit of behavior. The plugins are there so we don't have to get
bogged down in feature requests and can let users do a lot of it on
their own.
The other reason the plugins exist is b/c once something makes it into
the base yum code it has to stay there for a while. It has to be
supported and has to be lugged along.
-sv
9:34 a.m.
New subject: FC6 updates broken deps?
seth vidal wrote:
On Mon, 2007-02-26 at 15:59 +0100, Denis Leroy wrote:
> Maybe I'm mistaken. If I submit a patch that integrates --skip-broken
> into yum, will you accept it ? I'd be more than happy to write that patch.
>
It really depends on how the patch is implemented. Though I'm curious
why integrating it is important if the plugin works well enough?
because this way its nothing but a hack that does not solve the main
problem:
yum should _NOT_ fail to update a package because a unreleated package
has a dep problem.
-sv
1:11 p.m.
New subject: FC6 updates broken deps?
dragoran wrote:
because this way its nothing but a hack that does not solve the main
problem:
yum should _NOT_ fail to update a package because a unreleated package
has a dep problem.
Let me re-phrase it:
The light in the kitchen should not fail to function because a bulb in
the basement is broken.
Having a modular architecture is fine and all, but basic functionality
_must_ be included in the core. The skip-broken thing is so obvious, it
boggles the mind there's even a discussion about it.
It would be nice if Red Hat could conduct (more) usability tests on
software that's important for the OS. That would probably de-nerdify a
lot of applications.
--
Florin Andrei
http://florin.myip.org/
1:15 p.m.
New subject: FC6 updates broken deps?
On Mon, 2007-02-26 at 11:11 -0800, Florin Andrei wrote:
dragoran wrote:
> because this way its nothing but a hack that does not solve the main
> problem:
> yum should _NOT_ fail to update a package because a unreleated package
> has a dep problem.
Let me re-phrase it:
The light in the kitchen should not fail to function because a bulb in
the basement is broken.
to extend the analogy:
if the bulb in the basement being out means there could well be a
murderer down there, then maybe the door to the house shouldn't open at
all. :)
Having a modular architecture is fine and all, but basic
functionality
_must_ be included in the core. The skip-broken thing is so obvious, it
boggles the mind there's even a discussion about it.
It would be nice if Red Hat could conduct (more) usability tests on
software that's important for the OS. That would probably de-nerdify a
lot of applications.
To be clear:
If you'd like to help with yum, I'd appreciate it. Yelling the same
thing over and over again isn't really helping.
-sv
5:55 p.m.
New subject: FC6 updates broken deps?
seth vidal wrote:
to extend the analogy:
if the bulb in the basement being out means there could well be a
murderer down there, then maybe the door to the house shouldn't open at
all. :)
That analogy works both ways :-) If you just installed FC-6 from DVD,
you're missing all security updates issued since FC-6 was released. Say
a package maintainer is on vacation and a broken dep is still around.
You have a dark house full of murderers, but you can't turn on the
lights to expose them because a closet light bulb is out...
1:51 p.m.
New subject: FC6 updates broken deps?
seth vidal wrote:
On Mon, 2007-02-26 at 11:11 -0800, Florin Andrei wrote:
> dragoran wrote:
>
>
>> because this way its nothing but a hack that does not solve the main
>> problem:
>> yum should _NOT_ fail to update a package because a unreleated package
>> has a dep problem.
>>
> Let me re-phrase it:
>
> The light in the kitchen should not fail to function because a bulb in
> the basement is broken.
>
to extend the analogy:
if the bulb in the basement being out means there could well be a
murderer down there, then maybe the door to the house shouldn't open at
all. :)
I still don't get this.. do you really think that having no updates is
better than having all non broken updates? sorry but this does not make
any sense to me...
lets say there is a security hole in httpd -> update released
there is a bugfix (non security update) for a game but with a broken dep
(can happen...) -> this blocks the whole update process...
so its better in your opinion to let the system vunerable to attacks
only because a unrelated package has a broken dep? even if package 2 was
a security update too there is no reason to block 1 because this one has
broken deps .. sure there is a problem when suchs things happen but why
let our users systems with unpatched security holes even when the fix is
released?
6:53 p.m.
New subject: FC6 updates broken deps?
dragoran wrote:
I still don't get this.. do you really think that having no
updates is
better than having all non broken updates? sorry but this does not make
any sense to me...
Exactly. I have yet to see a rational argument that justifies the
current update policy implemented by yum. All the replies just ignore
this valid and important question.
--
Florin Andrei
http://florin.myip.org/
1:19 p.m.
New subject: FC6 updates broken deps?
Florin Andrei wrote:
It would be nice if Red Hat could conduct (more) usability tests on
software that's important for the OS. That would probably de-nerdify a
lot of applications.
How about you dont depend on one entity to do that? Usability can be a
community and volunteer driven process or atleast we should attempt to
do that. Recording user actions on the desktop is trivial to do these
days. There is http://fedoraproject.org/wiki/Usability. We could
probably get a place to upload those desktop recordings. What else do
you need?
Rahul
2:19 a.m.
New subject: FC6 updates broken deps?
seth vidal wrote:
On Mon, 2007-02-26 at 15:59 +0100, Denis Leroy wrote:
> Maybe I'm mistaken. If I submit a patch that integrates --skip-broken
> into yum, will you accept it ? I'd be more than happy to write that patch.
>
It really depends on how the patch is implemented. Though I'm curious
why integrating it is important if the plugin works well enough?
-sv
I wrote the skip-broken plugin because Seth asked me too, we agreed that
it was best to make it as a plugin because the check
skip-broken does take extra time, so you don't want to do it every time.
Integration of the skip-broken option into yum cli base, will not solve
the issue for the new users because they use pup to update the system,
so the have to be build some support for skip-broken into pup.
In yumex i am planning to ask the user if they want to remove packages
with broken depencies, if dependency error is found in the transaction.
I think that the plugin is fine for yum-cli, but extra stuff can be
build into the gui's like pup,pirut & yumex.
Tim
Tim
8:51 a.m.
Whoa! The big gun AC-130 coming out and hang with RMS with that footnote
quote from RMS. I like it.
It's ironic that a freedom loving guy ESR would selectively deviate on
issues of the contrary.
I secretly hope someday the ESR ship will turn around. I really do. I'll
wait.
Personally, I always actively run different distros so that I can acquire
the ability to approximately compare apples to apples. Three times now
Fedora managed to win me back. I hope my experience means this opinion is
balanced. I do see problems of not including codecs and how it'd affect
desktop adoption. I think that's a problem for Marketing. Oh yeah, I use
Fedora for desktop. Actually a powerful latop as desktop. I just don't
like noise.
"That was said by Eric Raymond who belongs to another movement"
- Richard Stallman
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
This post completed with hijacking of RMS's quote. :)
On 2/21/07, Alan Cox <alan(a)redhat.com> wrote:
On Wed, Feb 21, 2007 at 03:03:50AM -0500, Eric S. Raymond wrote:
> * Failure to address the problem of proprietary multimedia formats with
> any attitude other than blank denial.
That would be because we believe in Free Software and doing the right
thing
(a practice you appear to have given up on). Maybe it is time the term
"open source" also did the decent thing and died out with you.
> I'm not expecting Ubuntu to be perfect, but I am now certain it will
> be enough better to compensate me for the fact that I need to learn
> a new set of administration tools.
I'm sure they will be delighted to have you
Alan
--
"That was said by Eric Raymond who belongs to another movement"
- Richard Stallman
--
fedora-devel-list mailing list
fedora-devel-list(a)redhat.com
https://www.redhat.com/mailman/listinfo/fedora-devel-list
1:13 a.m.
On Wednesday 21 February 2007, Alan Cox wrote:
I'm sure they will be delighted to have you
...maybe at first, but not for long.
--
Patrick "The N-Man" Barnes
nman64(a)n-man.com
http://n-man.com/
LinkedIn:
http://linkedin.com/in/nman64
Have I been helpful? Rate my assistance!
http://rate.affero.net/nman64/
--
6:51 a.m.
(cutting media cc's)
I told myself I wasn't going to respond, but couldn't resist seeing as I
publicly blogged about switching to Ubuntu 6 months ago, but came back
to Fedora.
* Persistent failure to maintain key repositories in a sane,
consistent state from which upgrades might actually be possible.
You mean rawhide? Rawhide is *meant* to eat babies.
* A murky, poorly-documented, over-complex submission process.
Getting better - the core+extras merge was never going to be pretty but
I think it's going well and will evolve over a short amount of time to
be very sane.
* Allowing RPM development to drift and stagnate -- then adding
another layer of complexity, bugs, and wretched performance with
yum.
Aye, but as other replies have said, RPM is going to be fixed.
Yum on the other hand I agree with. pup, pirut and yum-updatesd all
fight for control for me, and none of them are easy to use or HIG
compliant.
I'm using synaptic - which seems to be much faster and easier to use
then pirut. The ubuntu software install and update tools are very slick
compared with Fedora.
* Effectively abandoning the struggle for desktop market share.
Really? Have you hung out in IRC or read the mailing lists recently?
Have you seen the really cool ConsoleKit, PolicyKit and HAL work that's
being done to make fast user switching work?
FC7 will be the first distro supporting ACL's on device objects and
allowing users to switch desktops with stuff like pulseaudio and
gnome-power-manager JUST WORKING.
That's real innovation, and really not abandoning the struggle for the
desktop. Please stop making unquantifiable statements.
* Failure to address the problem of proprietary multimedia formats
with any attitude other than blank denial.
Fedoras stance on this is stated. Sure, I watch DVD's and such with the
help of a really easy to use repo beginning with l.
Not shipping the non-free bits makes zero difference to me.
So, please stop using your position and trying to get a headline on
slashdot.
Many thanks.
Richard.
7:12 a.m.
Eric
I've followed your comments on Fedora and multimedia with interest. I
recall thinking you'd be better off with Ubuntu.
I'd be genuinely interested if you plan to write up your Ubuntu
experiences, good and bad. I searched for a blog but what I found was
mainly political (Islam etc.). Do you plan to write about your
experiences with Ubuntu somewhere?
-Cam
--
camilo(a)mesias.co.uk <--
7:26 a.m.
At 13:12 on 21 Feb 2007, Cam <camilo(a)mesias.co.uk> wrote:
Eric
I've followed your comments on Fedora and multimedia with interest. I
recall thinking you'd be better off with Ubuntu.
I'd be genuinely interested if you plan to write up your Ubuntu
experiences, good and bad. I searched for a blog but what I found was
mainly political (Islam etc.). Do you plan to write about your
experiences with Ubuntu somewhere?
I just made the mistake of looking for ESR's blog. Unfortunately, I
found it - http://esr.ibiblio.org/ - what a truly hideous individual he
appears to be.
Good riddance.
--
Mark Knoop
8:59 a.m.
New subject: Goodbye, Eric!
Mark Knoop wrote:
At 13:12 on 21 Feb 2007, Cam <camilo(a)mesias.co.uk> wrote:
I just made the mistake of looking for ESR's blog. Unfortunately, I
found it - http://esr.ibiblio.org/ - what a truly hideous individual he
appears to be.
Good riddance.
Let's say it with his own words:
"Well, screw you and the camel you rode in. I will not be silenced, and
we will not be silenced." - Eric S. Raymond [1]
Sorry, but I could not resist...
[1] http://esr.ibiblio.org/?p=290
--
Christian Nolte
mailto:ch.nolte@noltec.org
key : http://www.noltec.org/christian-nolte.asc
or : www.keyserver.net
GPG-fingerprint:
1088 6C2D 1496 0A34 D159 1108 08D8 C0D2 77E1 5BBC
----------------------------------------------------------------------
For more than 4 generations the IT Professionals were the guardians
of quality and stability in software. Before the dark times.
Before Microsoft...
----------------------------------------------------------------------
10:19 a.m.
Cam <camilo(a)mesias.co.uk>:
I'd be genuinely interested if you plan to write up your Ubuntu
experiences, good and bad. I searched for a blog but what I found was
mainly political (Islam etc.). Do you plan to write about your
experiences with Ubuntu somewhere?
It's likely.
--
<a href="http://www.catb.org/~esr/">Eric S. Raymond</a>
9:18 a.m.
Dnia 21-02-2007, śro o godzinie 03:03 -0500, Eric S. Raymond napisał(a):
[..]
I have watched Ubuntu rise to these challenges as Fedora fell away
from them. Canonical's recent deal with Linspire, which will give
Linux users legal access to WMF and other key proprietary codecs, is
precisely the sort of thing Red-Hat/Fedora could and should have taken
the lead in. Not having done so bespeaks a failure of vision which I
now believe will condemn Fedora to a shrinking niche in the future.
But .. WMF is *fully supported in Fedora* 8->
Look at libwmf package wchich provides library for handling this format.
This package provides library for handling WMF files and also gtk+ DSO
loader which enalbes using WMF files for examle in all GTK+/GNOME
applications.
You wrote "WMF" but probably in this case you though about "WMV".
Packages with codecs for this and other proprietar formats as sam as in
in Ubunt case are maintained _outside_ core packages set (leagal
reasons) .. so there is no *big/any* diffrence beetween Fedora and
Ubuntu on this area.
Please decide Raymond .. are you have technical problems with Fedora
"packages dependencies" (why are you not describe all your problems on
this list ?) or with legion of "political problems" ? (own problems ?).
kloczek
9:31 a.m.
On Wed, 2007-02-21 at 03:03 -0500, Eric S. Raymond wrote:
After thirteen years as a loyal Red Hat and Fedora user, I reached
my
limit today
We got this email last year, Eric, but it had slightly different
content. I think. Maybe. It's hard to read your messages. They are
SO BORING. They read like anything in the USA Today.
https://www.redhat.com/archives/fedora-devel-list/2006-March/msg01286.html
From your post last year, you were "threatening to jump very publicly to
another distribution and orphan several Fedora-related HOWTOs in the
process" and it looks like [from the content of last year's email], we
fixed a lot of things that bothered you about FC3 and FC4 and we have
been able to keep you as a loyal user for 13 years now. Clearly
threatening us gets us to fix things that bother people.
Last year you noted that our infrastructure by FC5 was great. Yum was
doing upgrades well and we no longer suffered from "egregious fuckups."
But now you say it's a huge problem again in the current email by noting
our "incompetent repository maintenance" and our non-statically linked
rpm. So what is the problem? Is it user incompetence or are you
actually aware of the infrastructure changes that take place or are
taking place? Reading your message, my guess is you are just
misinformed and that frustrates you.
It also looks like you failed on this issue from last year's message
from you:
...it's clear from the devel-list traffic that the
submission pathway for new packages has got most if not all of the
process issues shaken out of it. I had meant to get more directly
involved in that, there are four or five things I want to package
and submit myself, but higher-priority tasks ate my bandwidth. One
of my resolutions for 2006 is to make time for those submissions.
But then in the new message, you say:
* A murky, poorly-documented, over-complex submission process.
So, did you get your four or five packages done and submitted?
And that's about all of the message I can take.
Looking forward to next year's post. Consider an InfoGraphic for
next year's post. Imagine what you can say with a chart!
--
David Cantrell <dcantrell(a)redhat.com>
Red Hat / Westford, MA
10:43 a.m.
Eric S. Raymond wrote:
The proximate causes of this failure were (1) incompetent repository
maintenance, making any nontrivial upgrade certain to founder on a
failed dependency, and (2) the fact that rpm is not statically linked
-- so it's possible to inadvertently remove a shared library it
depends on and be unrecoverably screwed. But the underlying problems
run much deeper.
We only have your analysis of the problem. Not a description of the
actual problem. That would be more helpful. Perhaps a bug report.
* Chronic governance problems.
More details required. I have a special interest in this.
* Persistent failure to maintain key repositories in a sane,
consistent state from which upgrades might actually be possible.
* A murky, poorly-documented, over-complex submission process.
Last time you claimed this, I requested you to point out specific issues
which you agreed to. That was never delivered. Your actual submission
was struck after the reviewer pointed out some things to fix
https://www.redhat.com/archives/fedora-devel-list/2006-December/msg00538....
https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=220736
* Allowing RPM development to drift and stagnate -- then adding
another layer of complexity, bugs, and wretched performance with yum.
RPM indeed drifted but I dont think it actually stagnated. Anyway
http://rpm.org for more details. RPM doesnt do automatic dependency
resolving. Yum does. I did read somewhere your claim that introducing
yum was a big change that put Fedora in a position of advantage or some
such thing.
* Effectively abandoning the struggle for desktop market share.
Unless this is just about proprietary codecs, Red Hat continues to do a
lot of work on the desktop
http://fedoraproject.org/wiki/RedHatContributions
* Failure to address the problem of proprietary multimedia formats
with
any attitude other than blank denial.
We spend a lot of time even recently on FUDCon Boston 2007 discussing
this. See http://fedoraproject.org/wiki/Releases/FeatureCodecBuddy
Rahul
10:51 a.m.
On Wed, Feb 21, 2007 at 10:13:14PM +0530, Rahul Sundaram wrote:
>The proximate causes of this failure were (1) incompetent
repository
>maintenance, making any nontrivial upgrade certain to founder on a
>failed dependency, and (2) the fact that rpm is not statically linked
>-- so it's possible to inadvertently remove a shared library it
>depends on and be unrecoverably screwed. But the underlying problems
>run much deeper.
We only have your analysis of the problem. Not a description of the
actual problem. That would be more helpful. Perhaps a bug report.
Yes, excellent advice. Eric, here's a document which may help you get better
answers rather than the storm of flames:
<http://www.catb.org/~esr/faqs/smart-questions.html>;.
--
Matthew Miller mattdm(a)mattdm.org <http://mattdm.org/>
Boston University Linux ------> <http://linux.bu.edu/>
10:57 a.m.
Eric S. Raymond wrote:
* Persistent failure to maintain key repositories in a sane,
consistent state from which upgrades might actually be possible.
Ah. Forgot to add. F7 update system should fix the major problems
proactively by adding checks to ensure that dependencies are resolving
and there is a proper update path for every package being pushed out
but repositories doesnt seem to be in a bad state with regards to
upgrades as per the EVR check reports in fedora-maintainers list.
http://fedoraproject.org/wiki/Releases/FeatureUpdateSystem
Rahul
12:25 p.m.
Rahul Sundaram <sundaram(a)fedoraproject.org>:
We only have your analysis of the problem. Not a description of the
actual problem. That would be more helpful. Perhaps a bug report.
You have half of it. rpm should be statically linked to avoid this sort
of cul-de-sac. It's not like multople instances of it running are
going to be a frequent occurrence.
I can tell you the library in question was libcom_err, and I think I
deleted it when removing e2fsprogs-libs to get around a file conflict.
But with rpm not working I couldn't reinstall the library. Boot failed,
ssh/sshd failed -- I had to kluge with netcat and tar just to back up
my files. It was horrible.
Alas, I no longer have the logs, as I lost them during installation.
I don't recall what the file was, but the conflict was completely
avoidable and would never have occurred if the repo and RPMs had been
in a sane state.
>* Chronic governance problems.
More details required. I have a special interest in this.
I was thinking of the the endless internal wrangling over what to do
about the core/extras split, and the years the project has spent
failing to engage with third-party developers and repo maintainers.
That IRC-session parody of Fedora politics somebody wrote back around
2002 remains as painfully on-point today as it was then.
The Fedora project has never resolved the tension between Red Hat's
business need for it to be an adjunct of RHEL development and its core
group's stated aim to be community-facing. The result is a sociology
that has increasingly combined the least useful features of a corporate
project with the least useful snakepit-like features of 'community'
politics -- as perfectly illustrated by the list response to my goodbye
note.
>* A murky, poorly-documented, over-complex submission process.
Last time you claimed this, I requested you to point out specific issues
which you agreed to. That was never delivered. Your actual submission
was struck after the reviewer pointed out some things to fix
I was doing what I told Jesse Keating I'd do -- using that submission
as a probe of the process, and planning to write up a narrative of the
difficulties and some recommendations once it was done. (I'll note
that I thought the reviewer did a good job of critique, so the early
indications were somewhat positive.)
The things I could address in that particular submission were trivial
and are fixed, as I noted in a comment attached to the bug. I was
waiting on resubmitting until Mike Smith shipped a new stylesheet RPM,
as that was critical to fixing the man-page glitches.
Now it's no longer my problem.
>* Allowing RPM development to drift and stagnate -- then adding
> another layer of complexity, bugs, and wretched performance with yum.
RPM indeed drifted but I dont think it actually stagnated. Anyway
http://rpm.org for more details. RPM doesnt do automatic dependency
resolving. Yum does. I did read somewhere your claim that introducing
yum was a big change that put Fedora in a position of advantage or some
such thing.
Yes, I thought so at the time. But yum failed to live up to what I thought
was considerable early promise. It remained painfully slow and buggy.
Improvements in dependency resolution that seemed conceptually obvious
to me were never made.
The most obvious of these would have been a clique analysis of mass
updates to fail the smallest possible subset hanging on a missing
dependency, rather than failing the entire update. I used to be a
mathematician, I have the graph-theory chops to do something about
this, and I tried to contribute some refactoring patches that would
have led in this direction. They were rejected, and I had other
things to do.
>* Effectively abandoning the struggle for desktop market share.
Unless this is just about proprietary codecs, Red Hat continues to do a
lot of work on the desktop
To no visible result. You had the developers, the first-mover
advantage, and the corporate backing; you lacked the vision and the
will to follow through. Ubuntu shows what could have been done -- but
if you guys had executed correctly, Ubuntu's market- and mindshare
would be at best statistical noise (if it existed at all).
>* Failure to address the problem of proprietary multimedia
formats with
> any attitude other than blank denial.
We spend a lot of time even recently on FUDCon Boston 2007 discussing
this. See http://fedoraproject.org/wiki/Releases/FeatureCodecBuddy
Again, to no visible result. And with zealots screaming their heads off
against it whenever the topic came up on fedora-devel. I wound up
expecting that any Fedora 'solution' would be half-hearted and ineffective
and involve far too much sneering at users' actual needs and too little
effort spent on actually meeting them.
The wiki page does nothing to disabuse me of this expectation by using
terms like "brainwashing". The contempt for mere users this exhibits is
very revealing, and is a significant part of the reason I've decamped.
Meanwhile, Ubuntu and Linspire are actually addressing the problem at
its root. They might get it solved in time to meet the 64-bit
deadline Rob Landley and I have seen coming, but Fedora clearly will
not. So much for Fedora.
--
<a href="http://www.catb.org/~esr/">Eric S. Raymond</a>
12:50 p.m.
On 2/21/07, Eric S. Raymond <esr(a)thyrsus.com> wrote:
I was thinking of the the endless internal wrangling over what to do
about the core/extras split, and the years the project has spent
failing to engage with third-party developers and repo maintainers.
That IRC-session parody of Fedora politics somebody wrote back around
2002 remains as painfully on-point today as it was then.
No, it's not. Every point I've raised in that mockalogue has been
largely addressed. Red Hat is not even conflicted by Fedora any more,
as it seems to have finally gotten through to the sales drones that
people who are using Fedora aren't likely to pay for a RHEL license
anyway, so offering Fedora does not translate into lost sales.
Lots of problems remain, especially now that the Core/Extras merger is
making everyone's life miserable due to the fact that packaging
procedures have been changing daily (or so it seems) since early
February, but that's a hurdle that is being actively worked on -- once
the dust settles, things will clear up and new docs will be written.
Everything is far from being "broken" as you imply, and you will find
that grass may not necessarily be greener on the other side -- just
different flavours of the same weed.
--
Konstantin Ryabitsev
Montréal, Québec
12:55 p.m.
New subject: The Long Goodbye
Eric S. Raymond wrote:
I can tell you the library in question was libcom_err, and I think I
deleted it when removing e2fsprogs-libs to get around a file conflict.
But with rpm not working I couldn't reinstall the library. Boot failed,
ssh/sshd failed -- I had to kluge with netcat and tar just to back up
my files. It was horrible.
Alas, I no longer have the logs, as I lost them during installation.
I don't recall what the file was, but the conflict was completely
avoidable and would never have occurred if the repo and RPMs had been
in a sane state.
Why does "booted from the install media in recovery mode" not feature in
this anywhere? Then you could have wgetted
http://download.fedora.redhat.com/pub/fedora/linux/core/6/i386/os/Fedora/...
from inside the recovery boot and installed it after a chroot or with
--root= on rpm.
Yesterday I rendered this laptop unbootable by broken code I added to
the iwlwifi driver sources: I was back up in five minutes thanks to the
recovery boot, deleting the driver and rebooting into the kernel so I
could fix it and go on.
That experience, like each time I use the recovery boot, left me feeling
grateful for the feature and empowered, that even with a half trashed
system I can force it back to life. Whereas without it, you were left
feeling very different.
To no visible result. You had the developers, the first-mover
advantage, and the corporate backing; you lacked the vision and the
will to follow through. Ubuntu shows what could have been done -- but
if you guys had executed correctly, Ubuntu's market- and mindshare
would be at best statistical noise (if it existed at all).
...
its root. They might get it solved in time to meet the 64-bit
deadline Rob Landley and I have seen coming, but Fedora clearly will
not. So much for Fedora.
Rob sent me a link to your thesis some weeks ago, but as I told him then
it seems completely bogus to me. Some geeky metric like the sizeof(int)
going to control whether the man in the street switches OS or not.
Crushing Ubuntu isn't what the other OSes are about. If you look at it
from, say, a Gnome or kernel perspective where RHAT are active, is it a
loser globally for RHAT if Gnome or the kernel are gaining in popularity
in another distro? So it is a lot more nuanced, Fedora is not a
commercial entity where there is per-user revenue and retarding the
competition advances yourself. While there are hundreds of thousands of
users it has a lot of meaning, especially seen as a Fedora / RHEL /
CentOS axis where the OS semantics are all compatible. I paid to get my
*RH*CE, a Canonical or Linspire version won't do.
-Andy
2:18 p.m.
New subject: The Long Goodbye
Andy Green wrote:
Eric S. Raymond wrote:
Why does "booted from the install media in recovery mode" not feature
in this anywhere? Then you could have wgetted
http://download.fedora.redhat.com/pub/fedora/linux/core/6/i386/os/Fedora/...
from inside the recovery boot and installed it after a chroot or with
--root= on rpm.
Yesterday I rendered this laptop unbootable by broken code I added to
the iwlwifi driver sources: I was back up in five minutes thanks to
the recovery boot, deleting the driver and rebooting into the kernel
so I could fix it and go on.
That experience, like each time I use the recovery boot, left me
feeling grateful for the feature and empowered, that even with a half
trashed system I can force it back to life. Whereas without it, you
were left feeling very different.
same here wanted to install d80211 and deleted all my kernel modules by
acciedent (only noticed after reboot; only one kernel installed) I
reinstalled the kernel in rescue mode and the system was up and running
again in less then 10 min.
12:58 p.m.
On Wed, Feb 21, 2007 at 01:25:01PM -0500, Eric S. Raymond wrote:
Rahul Sundaram <sundaram(a)fedoraproject.org>:
> We only have your analysis of the problem. Not a description of the
> actual problem. That would be more helpful. Perhaps a bug report.
You have half of it. rpm should be statically linked to avoid this sort
of cul-de-sac. It's not like multople instances of it running are
going to be a frequent occurrence.
Then why did you not propose this till now? If this is something that is such a
big issue for you, why not bring it up until this rather noisy uneeded exit? If
you had before hand then I apologize and please show me where I can read what
happened, but if not, isn't the wonderful thing about open source building a
better produce through collaboration?
I can tell you the library in question was libcom_err, and I think I
deleted it when removing e2fsprogs-libs to get around a file conflict.
But with rpm not working I couldn't reinstall the library. Boot failed,
ssh/sshd failed -- I had to kluge with netcat and tar just to back up
my files. It was horrible.
Alas, I no longer have the logs, as I lost them during installation.
I don't recall what the file was, but the conflict was completely
avoidable and would never have occurred if the repo and RPMs had been
in a sane state.
What repo? What was the conflict? Did you file a bugzilla before you went
hacking away? If there is a problem it needs to be tracked and fixed. Again, I
thought this was how Open Source software was supposed to work.
> >* Chronic governance problems.
>
> More details required. I have a special interest in this.
I was thinking of the the endless internal wrangling over what to do
about the core/extras split, and the years the project has spent
failing to engage with third-party developers and repo maintainers.
That IRC-session parody of Fedora politics somebody wrote back around
2002 remains as painfully on-point today as it was then.
AFAIK core/extras is no longer split, I'm not sure my head has been in the sand
for the last 4 or 5 months. What 3rd party developers and repo maintainers
specifically are you talking about? Have you offered up a solution to this
so-called problem?
The Fedora project has never resolved the tension between Red
Hat's
business need for it to be an adjunct of RHEL development and its core
group's stated aim to be community-facing. The result is a sociology
that has increasingly combined the least useful features of a corporate
project with the least useful snakepit-like features of 'community'
politics -- as perfectly illustrated by the list response to my goodbye
note.
Welcome to human nature. This community is not what the problem is, it is
simple human nature to take your angry tirade marking all of the things we "do
wrong" without any suggestions of what to do right and to lash out against it,
as its a personal offense to everybody who has worked on this project in order
to make it better. This is a wonderful and helpful community, and if you write
things as you do to be offensive to those humans in the community, then expect
some/many of them to react as any other human would, with anger.
> >* Allowing RPM development to drift and stagnate -- then
adding
> > another layer of complexity, bugs, and wretched performance with yum.
>
> RPM indeed drifted but I dont think it actually stagnated. Anyway
> http://rpm.org for more details. RPM doesnt do automatic dependency
> resolving. Yum does. I did read somewhere your claim that introducing
> yum was a big change that put Fedora in a position of advantage or some
> such thing.
Yes, I thought so at the time. But yum failed to live up to what I thought
was considerable early promise. It remained painfully slow and buggy.
Improvements in dependency resolution that seemed conceptually obvious
to me were never made.
Then fix it. If the problems are so obvious to you why did you not submit
patches? Yum works fine for me, so I'm not going to try and fix it, but if its
not working for you, where are the bugzillas detailing the problems you have,
and the patches to fix them? I know you are a capable programmer, if these
things are so obvious and so much of a pain, why did you not fix them yourself
and submit them upstream? Isn't that the open source way?
The most obvious of these would have been a clique analysis of mass
updates to fail the smallest possible subset hanging on a missing
dependency, rather than failing the entire update. I used to be a
mathematician, I have the graph-theory chops to do something about
this, and I tried to contribute some refactoring patches that would
have led in this direction. They were rejected, and I had other
things to do.
Ok so instead of working with the developers in order to get these things
resolved, you just dropped them to do other things. So who's fault is this
again? It works fine for me, and apparently other people as well, we aren't
just going to read your mind or pick up after your work, if you feel that
something needs to be changed then why not try your hardest until some sort of
solution is gained?
All I see through this little tirade of yours is that you have no
interest in actually fixing problems, just complaining about them, and then not
even complaining about them in a manner which gets them fixed, just in
inflammatory ways in order to get attention. As I said before, the way things
are done now work fine for me, so I'm not going to change them, and this goes
for every individual, if it does what they need it to, they aren't going to
change it. So if it doesn't do what you need it to do, change it, or make sure
somebody knows it doesn't do what you need it to and try to work towards a
solution.
Josef
12:59 p.m.
On Wed, 2007-02-21 at 13:25 -0500, Eric S. Raymond wrote:
I can tell you the library in question was libcom_err, and I think I
deleted it when removing e2fsprogs-libs to get around a file conflict.
But with rpm not working I couldn't reinstall the library.
So you manually removed a critical system library and wondered why
things broke? Have you ever tried randomly deleting .dll files in
Windows XP?
Richard.
4:29 p.m.
Am Mittwoch, den 21.02.2007, 18:59 +0000 schrieb Richard Hughes:
On Wed, 2007-02-21 at 13:25 -0500, Eric S. Raymond wrote:
> I can tell you the library in question was libcom_err, and I think I
> deleted it when removing e2fsprogs-libs to get around a file conflict.
> But with rpm not working I couldn't reinstall the library.
So you manually removed a critical system library and wondered why
things broke? Have you ever tried randomly deleting .dll files in
Windows XP?
Richard.
XP has locking :)
3:58 a.m.
On Wed, 2007-02-21 at 23:29 +0100, nodata wrote:
Am Mittwoch, den 21.02.2007, 18:59 +0000 schrieb Richard Hughes:
> On Wed, 2007-02-21 at 13:25 -0500, Eric S. Raymond wrote:
> > I can tell you the library in question was libcom_err, and I think I
> > deleted it when removing e2fsprogs-libs to get around a file conflict.
> > But with rpm not working I couldn't reinstall the library.
>
> So you manually removed a critical system library and wondered why
> things broke? Have you ever tried randomly deleting .dll files in
> Windows XP?
>
> Richard.
>
XP has locking :)
With huge holes in it.
Try deleting random registry entries from HKEY_LOCAL_MACHINE and you'll
see what I mean... :)
- Gilboa
3:29 a.m.
Richard Hughes wrote:
On Wed, 2007-02-21 at 13:25 -0500, Eric S. Raymond wrote:
> I can tell you the library in question was libcom_err, and I think I
> deleted it when removing e2fsprogs-libs to get around a file conflict.
> But with rpm not working I couldn't reinstall the library.
So you manually removed a critical system library and wondered why
things broke? Have you ever tried randomly deleting .dll files in
Windows XP?
Richard.
Actually, this is probably a silly example, since XP has the
silly-admin-avoidance-system (or whatever it's called) that would
automatically restore vital .dll files upon deletion.
/Thomas
1:21 p.m.
Eric S. Raymond wrote:
I can tell you the library in question was libcom_err, and I think I
deleted it when removing e2fsprogs-libs to get around a file conflict.
But with rpm not working I couldn't reinstall the library. Boot failed,
ssh/sshd failed -- I had to kluge with netcat and tar just to back up
my files. It was horrible.
Well, deleting a critical library in a random fashion to avoid conflicts
isnt exactly a bright idea.
Alas, I no longer have the logs, as I lost them during installation.
I don't recall what the file was, but the conflict was completely
avoidable and would never have occurred if the repo and RPMs had been
in a sane state.
Perhaps but a proper bug report on the conflict and RFE's on whatever
suggestions you had on improving it would have been better. We cant
verify or deny your problem now.
I was thinking of the the endless internal wrangling over what to do
about the core/extras split, and the years the project has spent
failing to engage with third-party developers and repo maintainers.
I dont think we had any internal wrangling about that. Mostly it was
clear we had to bring them together as much as possible and the idea
that we are going to combine them in some way has been floating for a
while. The merge is very disruptive but it is a very significant step
forward. Possibly the best step we have every taken since the project
launch.
Third party repo maintainers like Mathias and Axel continue to be
heavily involved in Fedora now. I would say we did engage them.
That IRC-session parody of Fedora politics somebody wrote back
around
2002 remains as painfully on-point today as it was then.
Except it appears that the person who wrote it doesnt now agree with you
on that.
I was doing what I told Jesse Keating I'd do -- using that
submission
as a probe of the process, and planning to write up a narrative of the
difficulties and some recommendations once it was done. (I'll note
that I thought the reviewer did a good job of critique, so the early
indications were somewhat positive.)
I dont see how you could continue claiming that it is poorly documented
if you hadnt gone through the process completely. Please either specify
the problem or drop your claim. Spending a lot of time on something only
to get vague criticism repeatedly is very frustrating.
>> * Effectively abandoning the struggle for desktop market
share.
> Unless this is just about proprietary codecs, Red Hat continues to do a
> lot of work on the desktop
To no visible result. You had the developers, the first-mover
advantage, and the corporate backing; you lacked the vision and the
will to follow through. Ubuntu shows what could have been done -- but
if you guys had executed correctly, Ubuntu's market- and mindshare
would be at best statistical noise (if it existed at all).
I wouldnt describe the work on fundamental pieces of desktop such as
HAL, DBus, Network Manager, AIGLX and so on as having no visible result.
> We spend a lot of time even recently on FUDCon Boston 2007
discussing
> this. See http://fedoraproject.org/wiki/Releases/FeatureCodecBuddy
Again, to no visible result. And with zealots screaming their heads off
against it whenever the topic came up on fedora-devel. I wound up
expecting that any Fedora 'solution' would be half-hearted and ineffective
and involve far too much sneering at users' actual needs and too little
effort spent on actually meeting them.
You are just too early to see any results of this work. Judging the list
discussion to be reflective on the software development assumes that the
list discussions is actually from the developers.
The wiki page does nothing to disabuse me of this expectation by
using
terms like "brainwashing". The contempt for mere users this exhibits is
very revealing, and is a significant part of the reason I've decamped.
You just lose your sense of humor there.
Rahul
1:59 p.m.
Rahul Sundaram <sundaram(a)fedoraproject.org>:
Perhaps but a proper bug report on the conflict and RFE's on
whatever
suggestions you had on improving it would have been better. We cant
verify or deny your problem now.
I would have been happy to give you a detailed bug report -- if my system
had remained *usable*. As it is, I was too busy trying to figure out a way
not to lose all of my personal files in the reinstall.
>The wiki page does nothing to disabuse me of this expectation by
using
>terms like "brainwashing". The contempt for mere users this exhibits is
>very revealing, and is a significant part of the reason I've decamped.
You just lose your sense of humor there.
Oh, yes, it's very funny that Greg deKoenigsberg -- who is supposed to be
some kind of responsible adult supervision -- implies that anybody who
wants to watch YouTube or listen to a podcast is a glassy-eyed zombie.
Ha ha. See me laugh. Yes, that is *exactly* the kind of attitude that
will get us above 4% market share. Not.
--
<a href="http://www.catb.org/~esr/">Eric S. Raymond</a>
2:10 p.m.
On Wednesday 21 February 2007 14:59, Eric S. Raymond wrote:
Yes, that is *exactly* the kind of attitude that
will get us above 4% market share. Not.
Market share means nothing if we sacrifice our core values for a few more
percentage points. Why do we care if people use Linux? Well we want them to
use opensource software and contribute. We want to free them from
proprietary lock in. We want to give them freedom. Handing them a stack of
closed source software is not the way to accomplish those goals. All it does
is muddy the line between us in the %4, and the other guys in the
remaining %.
--
Jesse Keating
Release Engineer: Fedora
2:30 p.m.
New subject: Core values
Jesse Keating <jkeating(a)redhat.com>:
Market share means nothing if we sacrifice our core values for a few
more
percentage points.
This is so out of contact with reality that it makes me grind my teeth.
Want to fix or scrap the damned patent system? Want to get the damned
DMCA repealed? Do you want even as little a thing as 3D graphics and
wireless cards with open specs?
To get these things, we certainly need market share over 30% and probably
need market share over 50%. We need politicians and vendors to fear
our wrath.
It wasn't 'sacrificing our core values' to ship a binary Netscape
blob until Mozilla was ready. It won't be sacrificing our core values
to ship proprietary codecs until we are positioned to crush them out
of existence.
The Ubuntu crowd gets this. The Fedora crowd doesn't. That difference is
nowhere near the only reason I'm gone, but it's a big one.
--
<a href="http://www.catb.org/~esr/">Eric S. Raymond</a>
3:09 p.m.
New subject: Core values
On 2/21/07, Eric S. Raymond <esr(a)thyrsus.com> wrote:
[snip]
Want to fix or scrap the damned patent system? Want to get the
damned
DMCA repealed? Do you want even as little a thing as 3D graphics and
wireless cards with open specs?
To get these things, we certainly need market share over 30% and probably
need market share over 50%. We need politicians and vendors to fear
our wrath.
Why not make them irrelevant or obsolete by providing alternatives,
educating people, and making progress a step at a time?
[snip]
The Ubuntu crowd gets this. The Fedora crowd doesn't. That
difference is
nowhere near the only reason I'm gone, but it's a big one.
You have your own positions and your own goals. You are entitled to have them.
Other people sometimes have differing goals, and differing positions.
This doesn't make people wrong.
If I wanted an OS stuffed with proprietary software I could run MacOS X.
After many years, a desktop of completely free software finally
satisfies my needs with virtually no compromise. This is not true for
everyone, but it is true for some people. I've been in a situation of
having a mixed system, and it stinks. I don't want to go back. Not an
*inch*.
Be thankful that there exists a Linux distribution which, apparently,
shares your value of short term ease over other factors. This saves
you the trouble of stuffing your free operating system full of
proprietary stuff on your own.
You preference is already provided by someone else, so please don't
demand that I lose the ability to choose a reasonable mainstream
distribution which contains Free Software.
If you are right than surely your solution will win at the end of the
day. So if you are convinced that you are right, why are you wasting
your time ranting about this?
3:30 p.m.
New subject: Core values
On Wed, Feb 21, 2007 at 03:30:08PM -0500, Eric S. Raymond wrote:
The Ubuntu crowd gets this. The Fedora crowd doesn't. That
difference is
nowhere near the only reason I'm gone, but it's a big one.
Didn't you leave already?
Dave
--
http://www.codemonkey.org.uk
7:24 a.m.
New subject: Core values
Verily I say unto thee, that Dave Jones spake thusly:
On Wed, Feb 21, 2007 at 03:30:08PM -0500, Eric S. Raymond wrote:
> The Ubuntu crowd gets this. The Fedora crowd doesn't. That difference is
> nowhere near the only reason I'm gone, but it's a big one.
Didn't you leave already?
I don't think he's generated enough publicity for Ubuntu/Linspire yet,
to justify his paycheck.
--
K.
http://slated.org - Slated, Rated & Blogged
.----
| "We shall pay any price, bear any burden, meet any hardship, support
| any friend, oppose any foe, in order to assure the survival and the
| success of liberty." - John F. Kennedy
`----
Fedora Core release 5 (Bordeaux) on sky, running kernel 2.6.19-1.2288.fc5
13:21:46 up 3 days, 47 min, 2 users, load average: 0.06, 0.04, 0.00
9:41 a.m.
New subject: Core values
Keith G. Robertson-Turner <fedora-gmane.00003(a)genesis-x.nildram.co.uk>:
I don't think he's generated enough publicity for
Ubuntu/Linspire yet,
to justify his paycheck.
Linspire hasn't paid me anything for being on the *Freespire* advisory
board, nor will they.
--
<a href="http://www.catb.org/~esr/">Eric S. Raymond</a>
11:29 a.m.
New subject: Core values
Keith G. Robertson-Turner wrote:
Verily I say unto thee, that Dave Jones spake thusly:
> On Wed, Feb 21, 2007 at 03:30:08PM -0500, Eric S. Raymond wrote:
>
> > The Ubuntu crowd gets this. The Fedora crowd doesn't. That difference is
> > nowhere near the only reason I'm gone, but it's a big one.
>
> Didn't you leave already?
I don't think he's generated enough publicity for Ubuntu/Linspire yet,
to justify his paycheck.
Color me old-fashioned, but isn't now a good time to stop the mud
slinging contest?!? I really doubt anything constructive will ever come
from that.
/Thomas
11:37 a.m.
New subject: Core values
On 2/22/07, Thomas M Steenholdt <tmus(a)tmus.dk> wrote:
Keith G. Robertson-Turner wrote:
> Verily I say unto thee, that Dave Jones spake thusly:
>> On Wed, Feb 21, 2007 at 03:30:08PM -0500, Eric S. Raymond wrote:
>>
>> > The Ubuntu crowd gets this. The Fedora crowd doesn't. That
difference is
>> > nowhere near the only reason I'm gone, but it's a big one.
>>
>> Didn't you leave already?
>
> I don't think he's generated enough publicity for Ubuntu/Linspire yet,
> to justify his paycheck.
>
Color me old-fashioned, but isn't now a good time to stop the mud
slinging contest?!? I really doubt anything constructive will ever come
from that.
/Thomas
To which mudslinging do you refer to exactly?
--
Fedora Core 6 and proud
12:04 p.m.
New subject: Core values
Arthur Pemberton wrote:
On 2/22/07, Thomas M Steenholdt <tmus(a)tmus.dk> wrote:
> Keith G. Robertson-Turner wrote:
> > Verily I say unto thee, that Dave Jones spake thusly:
> >> On Wed, Feb 21, 2007 at 03:30:08PM -0500, Eric S. Raymond wrote:
> >>
> >> > The Ubuntu crowd gets this. The Fedora crowd doesn't. That
> difference is
> >> > nowhere near the only reason I'm gone, but it's a big one.
> >>
> >> Didn't you leave already?
> >
> > I don't think he's generated enough publicity for Ubuntu/Linspire yet,
> > to justify his paycheck.
> >
>
> Color me old-fashioned, but isn't now a good time to stop the mud
> slinging contest?!? I really doubt anything constructive will ever come
> from that.
>
> /Thomas
To which mudslinging do you refer to exactly?
Mudslinging, name-calling and negatively worded posts, without a real
value to the debate, in general...
Who cares who's paid for what and by whom. Some possible areas of
improvement was brought up. Perhaps we can use some that as valuable
input to improve Fedora, perhaps not. So we should all filter the input
to end up with something we can use constructively, rather than to keep
"feeding the fire".
I'm all for the debate, but some of the posts are just not constructive
at all. So lets try to focus on what we can use the input for, rather
than spending a lot of time getting all worked up.
/Thomas
12:45 p.m.
New subject: Core values
Thomas M Steenholdt wrote:
Arthur Pemberton wrote:
> On 2/22/07, Thomas M Steenholdt <tmus(a)tmus.dk> wrote:
>> Keith G. Robertson-Turner wrote:
>> > Verily I say unto thee, that Dave Jones spake thusly:
>> >> On Wed, Feb 21, 2007 at 03:30:08PM -0500, Eric S. Raymond wrote:
>> >>
>> >> > The Ubuntu crowd gets this. The Fedora crowd doesn't. That
>> difference is
>> >> > nowhere near the only reason I'm gone, but it's a big
one.
>> >>
>> >> Didn't you leave already?
>> >
>> > I don't think he's generated enough publicity for Ubuntu/Linspire
>> yet,
>> > to justify his paycheck.
>> >
>>
>> Color me old-fashioned, but isn't now a good time to stop the mud
>> slinging contest?!? I really doubt anything constructive will ever come
>> from that.
>>
>> /Thomas
>
> To which mudslinging do you refer to exactly?
Mudslinging, name-calling and negatively worded posts, without a real
value to the debate, in general...
Who cares who's paid for what and by whom. Some possible areas of
improvement was brought up. Perhaps we can use some that as valuable
input to improve Fedora, perhaps not. So we should all filter the
input to end up with something we can use constructively, rather than
to keep "feeding the fire".
I'm all for the debate, but some of the posts are just not
constructive at all. So lets try to focus on what we can use the input
for, rather than spending a lot of time getting all worked up.
You are to be
praised for this post. In a thread filled with way too
much venom, way too much arrogance, there have been too few who, like
you, seem to possess reason and sense.
Maybe Eric does have some points. Maybe Fedora can be better because of
his rant. But most important of all, he has every right to speak his
mind, as do all the naysayers.
I have been there. Fedora in general and Rawhide in particular drives
me nuts. I am not as technically astute as many on this list. But I'd
sure like to be. So I hang in there. And I'd like to help make Fedora
better than it is. Because of what you said, I will look for an
opportunity to help you.
Thank you, Thomas.
-pmr
/Thomas
4:38 p.m.
New subject: Core values
On Wed, Feb 21, 2007 at 03:30:08PM -0500, Eric S. Raymond wrote:
The Ubuntu crowd gets this. The Fedora crowd doesn't. That
difference is
nowhere near the only reason I'm gone, but it's a big one.
For really bizarrely limited definitions of "The Ubuntu crowd", perhaps.
(Speaking as a member of the Ubuntu technical board)
--
Matthew Garrett | mjg59(a)srcf.ucam.org
4:52 p.m.
New subject: Core values
Matthew Garrett <mjg59(a)srcf.ucam.org>:
> The Ubuntu crowd gets this. The Fedora crowd doesn't. That
difference is
> nowhere near the only reason I'm gone, but it's a big one.
For really bizarrely limited definitions of "The Ubuntu crowd", perhaps.
(Speaking as a member of the Ubuntu technical board)
I'm gathering that it wasn't you who did the deal with Linspire to get
Ubuntu's hands on their codecs, then.
--
<a href="http://www.catb.org/~esr/">Eric S. Raymond</a>
4:57 p.m.
New subject: Core values
On Wed, Feb 21, 2007 at 05:52:42PM -0500, Eric S. Raymond wrote:
I'm gathering that it wasn't you who did the deal with
Linspire to get
Ubuntu's hands on their codecs, then.
The fact that I'm not employed by Canonical does mean I don't generally
get involved in business deals, yes. The Linspire/Ubuntu deal is a
commercial issue, not one that's been in any way shaped by the opinions
of the Ubuntu community or governance structure.
--
Matthew Garrett | mjg59(a)srcf.ucam.org
5:13 p.m.
New subject: Core values
Matthew Garrett <mjg59(a)srcf.ucam.org>:
> I'm gathering that it wasn't you who did the deal with
Linspire to get
> Ubuntu's hands on their codecs, then.
The fact that I'm not employed by Canonical does mean I don't generally
get involved in business deals, yes. The Linspire/Ubuntu deal is a
commercial issue, not one that's been in any way shaped by the opinions
of the Ubuntu community or governance structure.
OK. Then I think I may safely deduce that Mark Shuttleworth gets it, but
that the extent to which his enlightenment is shared by the 'community'
part of the project is not established.
Thank you for the correction.
--
<a href="http://www.catb.org/~esr/">Eric S. Raymond</a>
5:25 p.m.
New subject: Core values
On Wed, Feb 21, 2007 at 06:13:56PM -0500, Eric S. Raymond wrote:
OK. Then I think I may safely deduce that Mark Shuttleworth gets it,
but
that the extent to which his enlightenment is shared by the 'community'
part of the project is not established.
I should clarify that despite being elected by the community, I am in no
direct position to represent their opinions. I reply merely to make it
clear that it's not possible to claim that the Ubuntu community "get"
anything without asking them directly. To that extent, I agree with you.
My own opinions are certainly incompatible with yours in this respect,
and I suspect that the same is true of many others, but I have
absolutely no ability to provide you with any judgement over the
"community opinion".
--
Matthew Garrett | mjg59(a)srcf.ucam.org
5:27 p.m.
New subject: Core values
On Wednesday 21 February 2007 05:13:56 pm Eric S. Raymond wrote:
Matthew Garrett <mjg59(a)srcf.ucam.org>:
> > I'm gathering that it wasn't you who did the deal with Linspire to get
> > Ubuntu's hands on their codecs, then.
>
> The fact that I'm not employed by Canonical does mean I don't generally
> get involved in business deals, yes. The Linspire/Ubuntu deal is a
> commercial issue, not one that's been in any way shaped by the opinions
> of the Ubuntu community or governance structure.
OK. Then I think I may safely deduce that Mark Shuttleworth gets it, but
that the extent to which his enlightenment is shared by the 'community'
part of the project is not established.
Thank you for the correction.
--
<a href="http://www.catb.org/~esr/">Eric S. Raymond</a>
I
thought you were gone already.
If you can't participate in a manner that is not troll like please don't
participate at all. Fedora has a clear set of goals an objectives. and
proprietary anything has no place in Fedora. of course you are free to
produce your own distro and call it something else with the bits added. we
even provide tools for you to do it. you just cant call it Fedora.
--
Dennis Gilmore, RHCE
5:31 p.m.
New subject: Core values
On Wed, Feb 21, 2007 at 03:30:08PM -0500, Eric S. Raymond wrote:
This is so out of contact with reality that it makes me grind my
teeth.
Eric I think you lost the plot, actually I don't think you ever got the plot
in the first place but that is another story
Want to fix or scrap the damned patent system? Want to get the
damned
DMCA repealed? Do you want even as little a thing as 3D graphics and
wireless cards with open specs?
I spend material amounts of time with people like politicians. They don't
care about 30% market share, they care (on the whole) about doing the right
thing and where there isn't an easy answer picking a path that doesn't result
in chaos.
need market share over 50%. We need politicians and vendors to fear
our wrath.
Ah yes its the gun/power thing again. "Me got big gun, me scary", which usually
results in "me dead". You don't win hearts and minds by being scary and
violent
(see Iraq for a worked example).
It wasn't 'sacrificing our core values' to ship a binary
Netscape
blob until Mozilla was ready. It won't be sacrificing our core values
The real free distributions didn't ship binary netscape blobs.
to ship proprietary codecs until we are positioned to crush them out
of existence.
There is a much easier way to work with these people. You support those
who open source stuff (eg intel video), you ignore those who won't (eg AMD
bugs get a lot less time nowdays and will continue to do so until ATI grow up)
For proprietary codecs its up to end users if they want to install them, yum
provides a mechanism for anyone on the planet to distribute new repositories
not just livna and friends. You've got a nice level playing field and a
business opportunity. The moment Fedora includes non-free stuff it becomes
a problem for all the people who redistribute and respin it, and it becomes
unfair in the proprietary world in the eyes of everyone who didn't get
included.
Alan
7:25 p.m.
New subject: Core values
On 2/22/07, Eric S. Raymond <esr(a)thyrsus.com> wrote:
Jesse Keating <jkeating(a)redhat.com>:
> Market share means nothing if we sacrifice our core values for a few more
> percentage points.
This is so out of contact with reality that it makes me grind my teeth.
Want to fix or scrap the damned patent system? Want to get the damned
DMCA repealed? Do you want even as little a thing as 3D graphics and
wireless cards with open specs?
To get these things, we certainly need market share over 30% and probably
need market share over 50%. We need politicians and vendors to fear
our wrath.
It wasn't 'sacrificing our core values' to ship a binary Netscape
blob until Mozilla was ready. It won't be sacrificing our core values
to ship proprietary codecs until we are positioned to crush them out
of existence.
The Ubuntu crowd gets this. The Fedora crowd doesn't. That difference is
nowhere near the only reason I'm gone, but it's a big one.
Seem slike Ubuntu people tend to be people just to cheap to buy a Mac.
What it *sounds* like you need/want is a Mac. They are really quite
nice - I was on one the other day - all the multimedia that you can
possibly shale a stick at.
I kinda like Fedora though, and I'll be here for awhile, and when I
get a contracts for projects, i'll be sure to suggest CentOS or RHEL.
But yah...head over to an apple store. Macs are quite nice.
--
Fedora Core 6 and proud
4:59 a.m.
New subject: Core values
Eric S. Raymond wrote :
[...]
To get these things, we certainly need market share over 30% and
probably
need market share over 50%. We need politicians and vendors to fear
our wrath.
This starts to sound a bit too much like your views on terrorism. Ouch.
The Ubuntu crowd gets this. The Fedora crowd doesn't. That
difference is
nowhere near the only reason I'm gone, but it's a big one.
You're gone? Oh, but you didn't want to miss the fun troll-fest!!! :-)
Matthias
--
Clean custom Red Hat Linux rpm packages : http://freshrpms.net/
Fedora Core release 6 (Zod) - Linux kernel 2.6.19-1.2895.fc6
Load : 0.25 0.46 0.56
5:26 a.m.
New subject: Core values
On 2/23/07, Matthias Saou
<thias(a)spam.spam.spam.spam.spam.spam.spam.egg.and.spam.freshrpms.net>
wrote:
Eric S. Raymond wrote :
[...]
> To get these things, we certainly need market share over 30% and probably
> need market share over 50%. We need politicians and vendors to fear
> our wrath.
This starts to sound a bit too much like your views on terrorism. Ouch.
I wouldn't have understood this comment had I not read this guy's
(ESR) blog, which was frankly kinda scary and troubling
> The Ubuntu crowd gets this. The Fedora crowd doesn't. That
difference is
> nowhere near the only reason I'm gone, but it's a big one.
You're gone? Oh, but you didn't want to miss the fun troll-fest!!! :-)
Matthias
I hear good things about Ubuntu, and I don't really believe in distro
"wars", but just know that people like him use, and champion Ubuntu
gives me a sick feeling about it. I some times want to recommend
Ubuntu to newbies interested in Linux because I hear that it is so
user friendly, but it always seems like a bad idea after I think it
through.
--
Fedora Core 6 and proud
2:16 p.m.
Eric S. Raymond wrote:
Oh, yes, it's very funny that Greg deKoenigsberg -- who is supposed to be
some kind of responsible adult supervision -- implies that anybody who
wants to watch YouTube or listen to a podcast is a glassy-eyed zombie.
Ha ha. See me laugh. Yes, that is *exactly* the kind of attitude that
will get us above 4% market share. Not.
Just a quite note. Bill Nottingham wrote that. Not Greg Dek. Just so
that you can blame the right people for some self depreciating humor ..
or not.
Rahul
3:24 p.m.
Eric S. Raymond wrote:
Rahul Sundaram <sundaram(a)fedoraproject.org>:
>> The wiki page does nothing to disabuse me of this expectation by using
>> terms like "brainwashing". The contempt for mere users this exhibits
is
>> very revealing, and is a significant part of the reason I've decamped.
>>
> You just lose your sense of humor there.
>
Oh, yes, it's very funny that Greg deKoenigsberg -- who is supposed to be
some kind of responsible adult supervision -- implies that anybody who
wants to watch YouTube or listen to a podcast is a glassy-eyed zombie.
Ha ha. See me laugh. Yes, that is *exactly* the kind of attitude that
will get us above 4% market share. Not.
Um, the brainwashing mentioned is on the part of the Fedora community -
a jest (admittedly weak) about "indoctrinating" users into the free
software way of thinking.
Probably not as brilliant as showing up at a protest dressed as Obi Wan
Kenobi, but we can't all be Important, Serious People.
1:35 p.m.
On Wed, Feb 21, 2007 at 01:25:01PM -0500, Eric S. Raymond wrote:
Rahul Sundaram <sundaram(a)fedoraproject.org>:
> We only have your analysis of the problem. Not a description of the
> actual problem. That would be more helpful. Perhaps a bug report.
You have half of it. rpm should be statically linked to avoid this sort
of cul-de-sac. It's not like multople instances of it running are
going to be a frequent occurrence.
No, that's not really an issue, see below for the kernel analogon.
I can tell you the library in question was libcom_err, and I think I
deleted it when removing e2fsprogs-libs to get around a file
conflict.
Whatever the conflict may have been, in order to rip out a library
needed by other bits you need to use force (--nodeps) or an equivalent
step in a higher level tool. E.g. this does not happen accidentally,
you had to remove the safety pin.
It's like removing ext3.ko from the kernel and rebooting. If you can
access your data anymore would you conclude that a monolithic kernel
would had been better? In fact the package system employed in Fedora
(rpm) did have the dependencies in place and warned you when you tried
to remove the package w/o using --nodeps. There is no such safety
mechanism for the kernel, so would you recommend even stringer to have
monolithic kernels?
And even if rpm was build statically, you (or another user) may have
accidentially used rpm -e --nodeps rpm or even simpler rm -f
/bin/rpm and be in the same situation.
--
Axel.Thimm at ATrpms.net
4:40 p.m.
Eric S. Raymond <esr <at> thyrsus.com> writes:
I had to kluge with netcat and tar just to back up
my files. It was horrible.
And here we see the real problem - there was no proper backup of the system
BEFORE any of this was attempted. I'm sure that's Fedora's fault...
PS. I also hit the e2fsprogs-libs/libcom_err problem (while upgrading ot Rawhide
at some point in the past) and somehow managed NOT to screw up my system. But
even if I did, it wouldn't have been a problem - I had backup.
--
Bojan
4:49 p.m.
On 2/21/07, Bojan Smojver <bojan(a)rexursive.com> wrote:
And here we see the real problem - there was no proper backup of the
system
BEFORE any of this was attempted. I'm sure that's Fedora's fault...
speaking of backups......
I'm looking to submit a set of safekeep packages into the fedora
package universe.
If you are running rdiff-backup based backups now, you might want to
take these packages for a spin and see if you like safekeep's
features.
http://jspaleta.thecodergeek.com/Fedora%20SRPMS/safekeep/FC6/
-jef
7:29 a.m.
Verily I say unto thee, that Jeff Spaleta spake thusly:
I'm looking to submit a set of safekeep packages into the fedora
package universe.
If you are running rdiff-backup based backups now, you might want to
take these packages for a spin and see if you like safekeep's
features.
http://jspaleta.thecodergeek.com/Fedora%20SRPMS/safekeep/FC6/
Oh, that's new.
Hmm, will give it a spin. I'm a backup junkie.
--
K.
http://slated.org - Slated, Rated & Blogged
.----
| "Future archaeologists will be able to identify a 'Vista Upgrade
| Layer' when they go through our landfill sites" - Sian Berry, the
| Green Party.
`----
Fedora Core release 5 (Bordeaux) on sky, running kernel 2.6.19-1.2288.fc5
13:27:46 up 3 days, 53 min, 2 users, load average: 0.02, 0.04, 0.00
4:49 p.m.
Bojan Smojver <bojan(a)rexursive.com>:
And here we see the real problem - there was no proper backup of the
system
BEFORE any of this was attempted. I'm sure that's Fedora's fault...
Actually, there was. But it was a couple of days old. I didn't expect
a *single-package update* to be so dangerous that I needed to do
a full backup first. How foolishly optimistic of me.
--
<a href="http://www.catb.org/~esr/">Eric S. Raymond</a>
5:06 p.m.
Eric S. Raymond <esr <at> thyrsus.com> writes:
Actually, there was. But it was a couple of days old. I didn't
expect
a *single-package update* to be so dangerous that I needed to do
a full backup first. How foolishly optimistic of me.
You mean "the single package update that involved manual removal, by YOU, of a
crucial component from the system".
And you didn't have to do a full backup. You could just do a quick incremental
instead. Unless you had terabytes of new stuff, it wouldn't have taken that long.
Even after you created that problem for yourself by removing a vital component
from your system, you had options. Instead, as usual, you've chosen to throw a
tantrum on the list. Not cool.
--
Bojan
3:40 a.m.
Bojan Smojver wrote:
Eric S. Raymond <esr <at> thyrsus.com> writes:
> Actually, there was. But it was a couple of days old. I didn't expect
> a *single-package update* to be so dangerous that I needed to do
> a full backup first. How foolishly optimistic of me.
You mean "the single package update that involved manual removal, by YOU, of a
crucial component from the system".
And you didn't have to do a full backup. You could just do a quick incremental
instead. Unless you had terabytes of new stuff, it wouldn't have taken that long.
Even after you created that problem for yourself by removing a vital component
from your system, you had options. Instead, as usual, you've chosen to throw a
tantrum on the list. Not cool.
--
Bojan
In all fairness, I wouldn't have performed that backup neither, as I
expect is true for a lot of even semi-knowledgable sysadmins and generic
linux powerusers. Not sor something as simple as this. However - There
is still the rescue mode that's been mentioned a few times already,
which could have fixed the issue in a matter of minutes. Without the backup!
/Thomas
4:53 a.m.
Thomas M Steenholdt <tmus <at> tmus.dk> writes:
In all fairness, I wouldn't have performed that backup neither,
as I
expect is true for a lot of even semi-knowledgable sysadmins and generic
linux powerusers.
Well, been there, done that. Experienced cold sweat etc.
OK, next time I blow up one of my own systems, I'll write one of those letters.
Actually, hang on - just a few days ago I forgot to run grub install on a disk I
replaced and the system wouldn't boot after that. Consider it done - I'm
co-signing ESR's letter - it's all Fedora's fault! Never mind that the system
wasn't actually running Fedora... ;-)
--
Bojan
5:16 p.m.
On Wed, 21 Feb 2007 17:49:18 -0500, Eric S. Raymond wrote:
Bojan Smojver:
> And here we see the real problem - there was no proper backup of the system
> BEFORE any of this was attempted. I'm sure that's Fedora's fault...
Actually, there was. But it was a couple of days old. I didn't expect
a *single-package update* to be so dangerous that I needed to do
a full backup first. How foolishly optimistic of me.
You still haven't filled in the missing details. What repositories?
Rawhide? FC6? What conflict? I don't want to know why you choose to
broke something by using the --nodeps or --force options. I am just
interested in what conflict or broken dependency caused you to take
a try at fixing it without asking somebody for help.
8:47 a.m.
On Wed, 2007-02-21 at 17:49 -0500, Eric S. Raymond wrote:
Bojan Smojver <bojan(a)rexursive.com>:
> And here we see the real problem - there was no proper backup of the system
> BEFORE any of this was attempted. I'm sure that's Fedora's fault...
Actually, there was. But it was a couple of days old. I didn't expect
a *single-package update* to be so dangerous that I needed to do
a full backup first. How foolishly optimistic of me.
According to what I've read out of the scarce info about your problem,
/usr/lib/rpm/rpm2cpio.sh could easily have helped you (unless you
managed to damage other vital system stuff as well). No need to resort
to drastic methods like static linking ;-).
Nils
--
Nils Philippsen / Red Hat / nphilipp(a)redhat.com
"Those who would give up Essential Liberty to purchase a little Temporary
Safety, deserve neither Liberty nor Safety." -- B. Franklin, 1759
PGP fingerprint: C4A8 9474 5C4C ADE3 2B8F 656D 47D8 9B65 6951 3011
5:19 p.m.
On Wed, Feb 21, 2007 at 01:25:01PM -0500, Eric S. Raymond wrote:
You have half of it. rpm should be statically linked to avoid this
sort
of cul-de-sac. It's not like multople instances of it running are
going to be a frequent occurrence.
If you've lost a critical library then you may well have other problems
if you just ran a static rpm - such as the scripts. Normal users on being
told there is a package file collision believe the package manager and don't
randomly delete stuff breaking their system. Experts go and check the
dependancy tree and then decide not to do it.
But with rpm not working I couldn't reinstall the library. Boot
failed,
ssh/sshd failed -- I had to kluge with netcat and tar just to back up
my files. It was horrible.
You boot the rescue CD and use rpm --root to do the reinstall of the package.
If you aren't an expert you ask on the user list and get told how to do that.
Asking would have received a fairly prompt answer. Folks who run rawhide all
the time are pretty familiar with scenarios like "this glibc is bad" and
getting out of them.
Alan
12:49 a.m.
>>>> "AC" == Alan Cox <alan(a)redhat.com>
writes:
AC> You boot the rescue CD and use rpm --root to do the reinstall of
AC> the package. If you aren't an expert you ask on the user list and
AC> get told how to do that.
Normally you can rescue a deleted file by finding a process holding
the file open and doing a cp from /proc/123/fd/. Unfortunately
libraries tend to only be mmapped, with the fd long gone. Is there a
way to rescue a file which is only mmapped?
/Benny
5:36 a.m.
On Wed, Feb 21, 2007 at 01:25:01PM -0500, Eric S. Raymond wrote:
Rahul Sundaram <sundaram(a)fedoraproject.org>:
> We only have your analysis of the problem. Not a description of the
> actual problem. That would be more helpful. Perhaps a bug report.
You have half of it. rpm should be statically linked to avoid this sort
of cul-de-sac.
That doesn't help. We (SUSE) had a statically linked rpm for years
and switched to a dynamically linked one. Statically linked doesn't
work because rpm needs to lookup passwd entries when installing
packages and the nss code in the passwd lookup uses dynamically
loaded plugins. You get lots of unpleasant effects if the
statically linked glibc code can't work with plugins from a
glibc update.
Cheers,
Michael.
--
Michael Schroeder mls(a)suse.de
main(_){while(_=~getchar())putchar(~_-1/(~(_|32)/13*2-11)*13);}
1:43 p.m.
On Wednesday, 21 February 2007 at 17:43, Rahul Sundaram wrote:
Eric S. Raymond wrote:
[...]
>* Allowing RPM development to drift and stagnate -- then adding
> another layer of complexity, bugs, and wretched performance with yum.
RPM indeed drifted but I dont think it actually stagnated. Anyway
http://rpm.org for more details.
RPM doesnt do automatic dependency resolving.
Actually, that sentence is not exactly true. Given a rpmdb with all available
packages, rpm can suggest packages to satisfy missing dependencies.
Regards,
R.
--
Fedora Extras contributor http://fedoraproject.org/wiki/DominikMierzejewski
Livna contributor http://rpm.livna.org MPlayer developer http://mplayerhq.hu
"Faith manages."
-- Delenn to Lennier in Babylon 5:"Confessions and Lamentations"
1:47 p.m.
Dominik 'Rathann' Mierzejewski wrote:
On Wednesday, 21 February 2007 at 17:43, Rahul Sundaram wrote:
> Eric S. Raymond wrote:
[...]
>> * Allowing RPM development to drift and stagnate -- then adding
>> another layer of complexity, bugs, and wretched performance with yum.
> RPM indeed drifted but I dont think it actually stagnated. Anyway
> http://rpm.org for more details.
> RPM doesnt do automatic dependency resolving.
Actually, that sentence is not exactly true. Given a rpmdb with all available
packages, rpm can suggest packages to satisfy missing dependencies.
Pointing out libraries or packages is different from automatically
downloading and actually resolving them.
Rahul
2:05 p.m.
On Friday, 23 February 2007 at 20:47, Rahul Sundaram wrote:
Dominik 'Rathann' Mierzejewski wrote:
>On Wednesday, 21 February 2007 at 17:43, Rahul Sundaram wrote:
>>Eric S. Raymond wrote:
>[...]
>>>* Allowing RPM development to drift and stagnate -- then adding
>>> another layer of complexity, bugs, and wretched performance with yum.
>>RPM indeed drifted but I dont think it actually stagnated. Anyway
>>http://rpm.org for more details.
>>RPM doesnt do automatic dependency resolving.
>
>Actually, that sentence is not exactly true. Given a rpmdb with all
>available
>packages, rpm can suggest packages to satisfy missing dependencies.
>
Pointing out libraries or packages is different from automatically
downloading and actually resolving them.
Not very much. It actually resolves them (after all, rpm operates on rpmdb),
it just can't download them.
Regards,
R.
--
Fedora Extras contributor http://fedoraproject.org/wiki/DominikMierzejewski
Livna contributor http://rpm.livna.org MPlayer developer http://mplayerhq.hu
"Faith manages."
-- Delenn to Lennier in Babylon 5:"Confessions and Lamentations"
12:33 p.m.
[snip]
I figured, OK, since the morning is shot anyway, I might as well go to
the diner for some breakfast.
It’s one of those classic New York diners, like the one on Seinfeld.
There’s a thirty page menu and a kitchen the size of a phone booth. It
doesn’t make sense. They must have Star Trek technology to get all those
ingredients into such a small space. Maybe they rearrange atoms on the
spot.
I was sitting by the cash register.
An older woman came up to pay her check. As she was paying, she said to
the owner, “you know, I’ve been coming here for years and years, and
that waiter was really rather rude to me.”
The owner was furious.
“What do you mean? No he wasn’t! He’s a good waiter! I never had a
complaint!’
The customer couldn’t believe it. Here she was, a loyal customer, and
she wanted to help out the owner by letting him know that one of his
waiters needed a little bit of help in the manners department, but the
owner was arguing with her!
“Well, that’s fine, but I’ve been coming here for years, and everybody
is always very nice to me, but that guy was rude to me,” she explained,
patiently.
“I don’t care if you’ve been coming here forever. My waiters are not
rude.” The owner proceeded to yell at her. “I never had no problems. Why
are you making problems?”
“Look, if you’re going to treat me this way I won’t come back.”
“I don’t care!” said the owner. One of the great things about owning a
diner in New York is that there are so many people in the city that you
can offend every single customer who ever comes into your diner and
you’ll still have a lot of customers. “Don’t come back! I don’t want you
as a customer!”
Good for you, I thought. Here’s a 60-something year old man, owner of a
diner, and you won some big moral victory against a little old lady. Are
you proud of yourself? How macho do you have to be? Does the moral
victory make you feel better? Did you really have to lose a repeat
customer?
Would it have made you feel totally emasculated to say, “I’m so sorry.
I’ll have a word with him?”
It’s easy to get caught up in the emotional heat of the moment when
someone is complaining.
-- http://www.joelonsoftware.com/articles/customerservice.html
3:30 p.m.
On 2/21/07, nodata <lsof(a)nodata.co.uk> wrote:
It's easy to get caught up in the emotional heat of the moment
when
someone is complaining.
-- http://www.joelonsoftware.com/articles/customerservice.html
+1. Good criticism leads to healthy, if sometimes painful,
self-examination, debate, and improvement.
-Chris
--
Chris Weyl
Ex astris, scientia
6:24 a.m.
Verily I say unto thee, that Eric S. Raymond spake thusly:
This afternoon, I installed Edgy Eft on my main development machine
--
Did Kevin Carmony approve your switch to Ubuntu, or are you just testing
the package base for inclusion in Linspire, whilst generating some
valuable publicity.
Fedora, you had every advantage
Except a financial one, apparently.
--
K.
http://slated.org - Slated, Rated & Blogged
.----
| "We shall pay any price, bear any burden, meet any hardship, support
| any friend, oppose any foe, in order to assure the survival and the
| success of liberty." - John F. Kennedy
`----
Fedora Core release 5 (Bordeaux) on sky, running kernel 2.6.19-1.2288.fc5
12:22:44 up 2 days, 23:48, 2 users, load average: 0.03, 0.06, 0.04
11:07 a.m.
Keith G. Robertson-Turner wrote:
> Fedora, you had every advantage
Except a financial one, apparently.
I'm somewhat confused by your statement - what lack of financial
do you see in Fedora? Red Hat is one of the primary financial sponsors
of the Fedora Project; even so far as to have many of its [Fedora's]
primary developers on its [Red Hat's] payroll.
Please clarify this for me. Thanks.
--
Peter Gordon (codergeek42)
This message was sent through a webmail
interface, and thus not signed.
11:11 a.m.
On 2/22/07, Peter Gordon <peter(a)thecodergeek.com> wrote:
Keith G. Robertson-Turner wrote:
>> Fedora, you had every advantage
>
> Except a financial one, apparently.
>
I'm somewhat confused by your statement - what lack of financial
do you see in Fedora? Red Hat is one of the primary financial sponsors
of the Fedora Project; even so far as to have many of its [Fedora's]
primary developers on its [Red Hat's] payroll.
Please clarify this for me. Thanks.
I could be just in a bad mood, or simply wrong....but I believe he
meant financial advantage for the OP / complainer.
--
Fedora Core 6 and proud
11:39 a.m.
Arthur Pemberton wrote:
> Please clarify this for me. Thanks.
I could be just in a bad mood, or simply wrong....but I believe he
meant financial advantage for the OP / complainer.
That would make a lot of sense given the context of his post.
Thanks for the hopeful explanation.
--
Peter Gordon (codergeek42)
This message was sent through a webmail
interface, and thus not signed.
11:42 a.m.
Verily I say unto thee, that Arthur Pemberton spake thusly:
On 2/22/07, Peter Gordon <peter(a)thecodergeek.com> wrote:
> Keith G. Robertson-Turner wrote:
> >> Fedora, you had every advantage
> >
> > Except a financial one, apparently.
> >
>
> I'm somewhat confused by your statement - what lack of financial
> do you see in Fedora? Red Hat is one of the primary financial sponsors
> of the Fedora Project; even so far as to have many of its [Fedora's]
> primary developers on its [Red Hat's] payroll.
>
> Please clarify this for me. Thanks.
I could be just in a bad mood, or simply wrong....but I believe he
meant financial advantage for the OP / complainer.
Correct.
Oh and I'm probably wrong too, and in a bad mood.
--
K.
http://slated.org - Slated, Rated & Blogged
.----
| "Future archaeologists will be able to identify a 'Vista Upgrade
| Layer' when they go through our landfill sites" - Sian Berry, the
| Green Party.
`----
Fedora Core release 5 (Bordeaux) on sky, running kernel 2.6.19-1.2288.fc5
17:39:59 up 3 days, 5:05, 2 users, load average: 0.00, 0.00, 0.00
10:07 p.m.
On Wed, 2007-02-21 at 03:03 -0500, Eric S. Raymond wrote:
[snip, snip, snip]
Am I the only one that finds ESR's post even more amusing now that
Microsoft lost a 1.52B$ patent case against Alcatel-Lucent concerning
the use of MP3 technology in WMP? [1]
- Gilboa
[1] http://www.dailytech.com/article.aspx?newsid=6198
4:16 a.m.
Gilboa Davara wrote:
On Wed, 2007-02-21 at 03:03 -0500, Eric S. Raymond wrote:
[snip, snip, snip]
Am I the only one that finds ESR's post even more amusing now that
Microsoft lost a 1.52B$ patent case against Alcatel-Lucent concerning
the use of MP3 technology in WMP? [1]
- Gilboa
[1] http://www.dailytech.com/article.aspx?newsid=6198
Amusing or not, it does confirm that our current policy of keeping all
the proprietary stuff away from Fedora, is really for the best. Even if
users have to pull those packages from 3rd party repos.
/Thomas
6:15 a.m.
New subject: keeping all he proprietary stuff away from Fedora (Was: Re: Goodbye, Fedora)
On 23.02.2007 11:16, Thomas M Steenholdt wrote:
Gilboa Davara wrote:
> On Wed, 2007-02-21 at 03:03 -0500, Eric S. Raymond wrote:
> [snip, snip, snip]
> Am I the only one that finds ESR's post even more amusing now that
> Microsoft lost a 1.52B$ patent case against Alcatel-Lucent concerning
> the use of MP3 technology in WMP? [1]
> - Gilboa
> [1] http://www.dailytech.com/article.aspx?newsid=6198
Amusing or not, it does confirm that our current policy of keeping all
the proprietary stuff away from Fedora, is really for the best. Even if
users have to pull those packages from 3rd party repos.
Agreed. Nevertheless it IMHO would be helpful,
- if we had one major 3rd party repo instead of multiple ones that are
conflicting with each other sometimes
- if that 3rd party repo would have be slitted in non-free stuff and
potential harmful stuff, so users and contributors from the US and some
other country's can safely work with the non-free repo without getting
involved with the more problematic stuff
- if we had some kind backing for a non-free repo from Red Hat or some
other "Big Company" when it comes to knocking of doors to ask questions
like "Hi Adobe, may we include a Adobe Reader RPM in the non-free
community repo foo? They would do some some minor adjustments to the RPM
to make it perfectly work with Fedora; that best for both sides"
Just my 2 cent.
CU
thl
6:47 a.m.
New subject: keeping all he proprietary stuff away from Fedora (Was: Re: Goodbye, Fedora)
On Fri, 23 Feb 2007 13:15:54 +0100, Thorsten Leemhuis wrote:
- if we had some kind backing for a non-free repo from Red Hat or
some
other "Big Company" when it comes to knocking of doors to ask questions
like "Hi Adobe, may we include a Adobe Reader RPM in the non-free
community repo foo? They would do some some minor adjustments to the RPM
to make it perfectly work with Fedora; that best for both sides"
?? Contradictory. It's closed-source. Basically, you don't want to
include it in that repo at all.
6:57 a.m.
New subject: keeping all he proprietary stuff away from Fedora (Was: Re: Goodbye, Fedora)
On 23.02.2007 13:47, Michael Schwendt wrote:
On Fri, 23 Feb 2007 13:15:54 +0100, Thorsten Leemhuis wrote:
> - if we had some kind backing for a non-free repo from Red Hat or some
> other "Big Company" when it comes to knocking of doors to ask questions
> like "Hi Adobe, may we include a Adobe Reader RPM in the non-free
> community repo foo? They would do some some minor adjustments to the RPM
> to make it perfectly work with Fedora; that best for both sides"
?? Contradictory. It's closed-source. Basically, you don't want to
include it in that repo at all.
Well, or yet another repo (but that might be to confusing) -- that
depends on the definition of "non-free" in this case.
CU
thl
1:16 a.m.
New subject: keeping all he proprietary stuff away from Fedora
On Feb 23, 2007, Thorsten Leemhuis <fedora(a)leemhuis.info> wrote:
- if that 3rd party repo would have be slitted in non-free stuff and
potential harmful stuff, so users and contributors from the US and
some other country's can safely work with the non-free repo without
getting involved with the more problematic stuff
I for one would love to be able to enable third-party repositories
containing only Free Software, even if encumbered by patents (that are
worth nothing in Brazil, where I live), without the risk of pulling in
some non-Free Software by accident.
--
Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member http://www.fsfla.org/
Red Hat Compiler Engineer aoliva(a){redhat.com, gcc.gnu.org}
Free Software Evangelist oliva(a){lsd.ic.unicamp.br, gnu.org}
1:07 a.m.
On Feb 23, 2007, Thomas M Steenholdt <tmus(a)tmus.dk> wrote:
Amusing or not, it does confirm that our current policy of keeping
all
the proprietary stuff away from Fedora, is really for the best.
Just a small clarification. Patent-encumbered doesn't mean
proprietary. Many of the multi-media packages that are kept out of
Fedora *are* Free Software, but they can't be shipped from the US
because of software patents.
Fedora actually helps weaken the software patents system by rejecting
even such Free Software packages, and promoting Free Software
implementations of Free Standards instead. A wise use of the fifth
freedom: http://www.fsfla.org/?q=en/node/139#1
But if it weren't for legal, patent-related reasons, Fedora could
include such packages without breaking its vow to not ship non-Free
Software (other than policy allowing for non-Free firmware).
--
Alexandre Oliva http://www.lsd.ic.unicamp.br/~oliva/
FSF Latin America Board Member http://www.fsfla.org/
Red Hat Compiler Engineer aoliva(a){redhat.com, gcc.gnu.org}
Free Software Evangelist oliva(a){lsd.ic.unicamp.br, gnu.org}
7:28 a.m.
Alexandre Oliva wrote:
On Feb 23, 2007, Thomas M Steenholdt <tmus(a)tmus.dk> wrote:
> Amusing or not, it does confirm that our current policy of keeping all
> the proprietary stuff away from Fedora, is really for the best.
Just a small clarification. Patent-encumbered doesn't mean
proprietary. Many of the multi-media packages that are kept out of
Fedora *are* Free Software, but they can't be shipped from the US
because of software patents.
Fedora actually helps weaken the software patents system by rejecting
even such Free Software packages, and promoting Free Software
implementations of Free Standards instead. A wise use of the fifth
freedom: http://www.fsfla.org/?q=en/node/139#1
But if it weren't for legal, patent-related reasons, Fedora could
include such packages without breaking its vow to not ship non-Free
Software (other than policy allowing for non-Free firmware).
Absolutely!
/Thomas
7:23 a.m.
* Gilboa Davara [23/02/2007 07:28]:
Am I the only one that finds ESR's post even more amusing now that
Microsoft lost a 1.52B$ patent case against Alcatel-Lucent concerning
the use of MP3 technology in WMP? [1]
Nope. I also find it amusing that the two companies that he claims are
"succesful" in marketing Linux are both bleeding money like there's no
tomorrow.
Emmanuel
6260
days inactive
6267
days old
186 comments
69 participants
participants (69)
-
Alan Cox -
Alexandre Oliva -
Andy Green -
Arthur Pemberton -
Axel Thimm -
Benny Amorsen -
Bojan Smojver -
Callum Lerwick -
Camilo Mesias -
Chris Weyl -
Christian Iseli -
Christian Nolte -
Dave Jones -
David Cantrell -
David Hollis -
Demond -
Denis Leroy -
Dennis Gilmore -
Dominik 'Rathann' Mierzejewski -
drago01 -
Emmanuel Seyman -
Eric S. Raymond -
Florian La Roche -
Florin Andrei -
Gianluca Sforna -
Gilboa Davara -
Gregory Maxwell -
Ian Burrell -
Jef Spaleta -
Jesse Keating -
Josef Whiter -
Keith G. Robertson-Turner -
Kip Thomas -
Konstantin Ryabitsev -
Kłoczko Tomasz -
Luya Tshimbalanga -
Lyvim Xaphir -
Mark Knoop -
Matthew Garrett -
Matthew Miller -
Matthias Saou -
Michael Knepher -
Michael Schroeder -
Michael Schwendt -
Mike Kearey -
Miles Lane -
Naoki -
Neil Thompson -
Nils Philippsen -
nodata -
Oisin Feeley -
Panu Matilainen -
Patrick W. Barnes -
Paul Michael Reilly -
Pete Zaitcev -
Peter Gordon -
Piotr Baranowski -
Rahul Sundaram -
Ralf Corsepius -
Ralf Ertzinger -
Richard Hughes -
seth vidal -
Simo Sorce -
Thomas M Steenholdt -
Thorsten Leemhuis -
Tim Lauridsen -
Tom -
Tomas Mraz -
Tony Nelson