Re: rpms/fence-agents/F-11 fence-agents.spec,1.13,1.14
by Michael Schwendt
On Tue, 27 Oct 2009 10:51:14 +0000 (UTC), Fabio wrote:
> Author: fabbione
>
> Update of /cvs/pkgs/rpms/fence-agents/F-11
> In directory cvs1.fedora.phx.redhat.com:/tmp/cvs-serv15209
>
> Modified Files:
> fence-agents.spec
> Log Message:
> Fix Requires: on libvirt/libvirt-client
> +%if 0%{?fedora} >= 12
> +Requires: libvirt-client
> +%else
> +Requires: libvirt
> +%endif
> +
What is this explicit dependency on a package name supposed to achieve?
There is the automatic arch-specific dependency on the libvirt SONAME
already, and it is tons better than a non-arch-specific and version-less
dependency on a package name.
https://fedoraproject.org/wiki/Packaging:Guidelines#Explicit_Requires
14 years, 5 months
ld ignoring parts of LD_LIBRARY_PATH?
by Christoph Höger
Hi,
this is slightly off topic, but may be relevant for fedora developing,
too:
I have set up a small test bed for an application I am converting to a
autotools build system. In this testbed I have:
./lib/libfoo.so
./lib/bar/libbar.so
When I add the lib/ entry to LD_LIBRARY_PATH it is recognized and
scanned for libraries. The same for the lib/bar entry. But when I add
both, the lib/ entry is ignored completely. Is this normal?
regards
Christoph
14 years, 5 months
Re: Fwd: Request to update ATi OSS driver for Fedora 12
by Adam Williamson
On Fri, 2009-10-23 at 09:39 +0200, Jerome Glisse wrote:
> > > I've installed F-12 beta on my new laptop with ati radeon hd 4570 graphic
> > > card, I was going to file new bug. With kms enabled, everything is really
> > > sloooow, with 'nomodeset' it's much faster. I can't say exactly "how" slow it
> > > is, is there anything I can use for measuring?
> >
> > cairo-perf-trace is, afaik, the best we have:
> > http://cworth.org/intel/performance_measurement/
> >
> > I'm not sure if bug reports on this are useful. To some extent, slow
> > performance with KMS on some chips is a known issue. Jerome, do you want
> > bug reports from people who have this problem, or would more reports not
> > provide any new information?
> I think we already have bug open for slowness with KMS, r600/r700 is
> missing few pieces. We will work on performance latter, likely not in
> time for F12. We are mostly focusing on fixing bugs which makes GPU
> unusable first.
Thanks, Jerome. So, Michal H, there's no need to file a bug for this.
Thanks for reporting it, though.
--
Adam Williamson
Fedora QA Community Monkey
IRC: adamw | Fedora Talk: adamwill AT fedoraproject DOT org
http://www.happyassassin.net
14 years, 5 months
Test please disregard
by Adam Miller
Setup my fp.org email recently and now testing sending to the list from my
myTouch.
-Adam (From Android)
14 years, 5 months
Re: Fedora with Universal Binaries?
by Ikem Krueger
> I just saw this article about an effort to create Universal binary style ELF
> binaries for Linux, and I thought that this would be something to watch, so
> that Fedora could integrate both x86-32 and x86-64 into single DVD sets.
I don't suggest to do that. As already mentioned, that would double
the size of the distro/iso. I would use this technic only, if
neccessary.
About "fat-elf" in general: As long as it is optional, I am fine with
it. May it at compile time or after compiling by stripping binaries.
(I'd like to see both options.)
14 years, 5 months
Unreadable binaries
by Richard W.M. Jones
$ ll /usr/libexec/pt_chown
-rws--x--x 1 root root 28418 2009-09-28 13:42 /usr/libexec/pt_chown
$ ll /usr/bin/chsh
-rws--x--x 1 root root 18072 2009-10-05 16:28 /usr/bin/chsh
What is the purpose of making binaries like these unreadable?
Originally I thought it was something to do with them being setuid,
but there are counterexamples:
$ ll /usr/bin/passwd
-rwsr-xr-x 1 root root 25336 2009-09-14 13:14 /usr/bin/passwd
Surely there is no possible secret in those binaries, since an
attacker could just as easily download the binary RPMs on another
machine in order to find out what is inside them.
There's a genuine reason for me asking about this. When we build the
libguestfs supermin appliance[1] we would like to be able to read
these binaries as non-root.
Rich.
[1] http://libguestfs.org/README.txt section "Supermin appliance"
--
Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones
virt-df lists disk usage of guests without needing to install any
software inside the virtual machine. Supports Linux and Windows.
http://et.redhat.com/~rjones/virt-df/
14 years, 5 months
Re: Action Tags concept
by Martin Bacovsky
On Friday 23 October 2009 20:17:12 Toshio Kuratomi wrote:
> On Fri, Oct 23, 2009 at 08:03:02PM +0200, Martin Bacovsky wrote:
> > On Friday 23 October 2009 17:51:16 you wrote:
> > > On Fri, Oct 23, 2009 at 04:46:36PM +0200, Martin Bacovsky wrote:
> > > > On Thursday 22 October 2009 16:33:06 you wrote:
> > > > > I like this concept. How does it relate to tagging?
> > > > > * As a replacement for tagging
> > > > > * As a separate feature from tagging
> > > > > * In addition to tagging where some output utilizes both tags and
> > > > > actions
> > > >
> > > > I was thinking of replacing current tags, because I'm affraid users
> > > > will be confused by two kinds of tags. On the other hand I would keep
> > > > categories imported from .desktop files (readonly/searchonly).
> > >
> > > So I think we might be committed to having freeform tags for its use as
> > > a comps replacement and grouping mechanism. I'm not 100% sure though.
> > > There is overlap::
> > >
> > > python-openssl
> >
> > As python-openssl is not an application, this is rather hypothetical, but
> > I get your point.
>
> Well, that could be part of the issue -- actions may be a great mapping for
> applications. But tags are also for non-applications.
>
> > > Also, how do we get the users to only enter actions(verbs) and not
> > > nouns when supplying new actions for a package?
> >
> > I'll start from here. The form for action tag input should look like
> > this: "I use this application to ________________ and give it *****
> > stars." I hope it makes majority of users to fill it properly. I would
> > also expect users to fill actions like "secure connection" or "encrypt
> > data".
> >
> > Whole Application Database is targeted mainly to common users. This
> > concept was designed with that in mind. So Action is something I use my
> > computer (application) for i.e. write document, edit photos, play music
> > etc. Task is the same thing with the difference that I need more steps to
> > get it done. In other words Task is just container for Actions and is not
> > connected with application.
> >
> > The other way around, if I had such Task "Write programs in python that
> > can communicate with network services over SSL" I would it rather expect
> > to contain actions such as "edit sourcecode", "run commandline
> > application" and so on.
> >
> > Also if didn't choose the right names for entities I'm talking about
> > (Action, Task, ActionTag,..), feel free to suggest better ones.
> >
> > Anyway, the example with tags would apply for any application too, and it
> > seems we would loose usefull information by removing freeform tags. The
> > question is how to effectively combine these two. Don't you think it may
> > be confusing/contarproductive to have two types of tags?
>
> Yeah I agree we might be better off having both and figuring out UI that
> makes it less confusing.
>
> I think that tagging is a superset of actions in some ways. If I operate
> on tags, I can't see a time where I wouldn't want to include actions when
> figuring out the results. OTOH, when I'm looking for actions, I probably
> wouldn't be interested in the non-action tags.
>
> Do actions and tasks seem like a mostly browse functionality? Or a mostly
> search functionality to you?
>
> -Toshio
>
I would like to have actions visible on application page. Tasks are rather search functionality.
I'll try to update the mockup and let you know.
Martin
14 years, 5 months
