PK Fix isn't working
by Eric Christensen
I just tried the fix (should have done this hours ago) and it doesn't
work.
My recommendation is to take the fix out of the Release Notes and the
Security Guide and say that more information is forthcoming.
Thoughts?
--Eric
14 years, 5 months
Docs Meeting 2009-11-18 Summary
by Eric Christensen
===================================================================================================
#fedora-meeting: Docs Project Meeting - Agenda: http://fedoraproject.org/wiki/Docs_Project_meetings
===================================================================================================
Meeting started by Sparks at 00:02:20 UTC. The full logs are available
at
http://meetbot.fedoraproject.org/fedora-meeting/2009-11-19/fedora-meeting...
.
Meeting summary
---------------
* Roll Call (Sparks, 00:02:29)
* Docs Project Leader Elections (Sparks, 00:08:50)
* ACTION: Sparks to send out email on nominations (Sparks, 00:15:35)
* Release Notes (Sparks, 00:16:19)
* LINK: http://fedora.is-sixsigma.com/F12beta/ - regenerated every
night, from t12zeroday even tho it says beta (jjmcd, 00:29:46)
* Guide Status (Sparks, 00:49:38)
* New Guides (Sparks, 00:52:28)
* Status on CC license rollout. (Sparks, 00:53:03)
* All other business (Sparks, 00:54:14)
* LINK: https://bugzilla.redhat.com/show_bug.cgi?id=534047 ? :)
(radsy, 00:55:10)
Meeting ended at 01:01:19 UTC.
Action Items
------------
* Sparks to send out email on nominations
Action Items, by person
-----------------------
* Sparks
* Sparks to send out email on nominations
* **UNASSIGNED**
* (none)
People Present (lines said)
---------------------------
* Sparks (76)
* jjmcd (70)
* rudi (44)
* radsy (13)
* zodbot (3)
* bcotton (2)
* buggbot (1)
Generated by `MeetBot`_ 0.1.4
.. _`MeetBot`: http://wiki.debian.org/MeetBot
14 years, 5 months
Docs Meeting 2009-11-18 IRC log
by Eric Christensen
00:02:20 <Sparks> #startmeeting Docs Project Meeting - Agenda: http://fedoraproject.org/wiki/Docs_Project_meetings
00:02:20 <zodbot> Meeting started Thu Nov 19 00:02:20 2009 UTC. The chair is Sparks. Information about MeetBot at http://wiki.debian.org/MeetBot.
00:02:20 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
00:02:29 <Sparks> #topic Roll Call
00:02:30 * Sparks
00:02:32 * radsy is here
00:02:35 * rudi is here
00:02:38 * bcotton is here
00:02:48 * jjmcd .
00:04:41 * Sparks gives everyone a few more minutes
00:08:21 <Sparks> Okay, let's get started
00:08:41 <Sparks> #chair jjmcd
00:08:41 <zodbot> Current chairs: Sparks jjmcd
00:08:50 <Sparks> #topic Docs Project Leader Elections
00:09:18 <Sparks> When I took this job back in January(?) I wanted to make sure that we had elections for a new leader after F12 got released.
00:09:26 <Sparks> So here is my suggestion...
00:09:47 <Sparks> We take nominations starting tonight and vote next week.
00:09:51 <Sparks> Thoughts?
00:09:59 <jjmcd> wikipage
00:10:22 <Sparks> That's a good place to start
00:10:41 <rudi> sounds like a plan; and the wiki is the natural place for noms to go
00:10:55 <jjmcd> Although, we could just agree to re-elect you by acclimation
00:10:55 <Sparks> rudi: Works for me.
00:11:12 <Sparks> I won't take the job unless someone else runs
00:11:13 <Sparks> :)
00:11:25 * Sparks votes for ---> that guy
00:11:41 <jjmcd> And I won't run unless I can find a sucker ... errr ... leader for RNs
00:12:04 <Sparks> We should probably figure out if we want to do this yearly or releasely or...
00:12:32 <jjmcd> Personally, I could argue for either
00:12:47 <jjmcd> Yearly gives somebody a chance to get established
00:12:51 <Sparks> yes
00:13:04 <jjmcd> releasely gives is a quicker way to get rid of a turkey
00:14:00 <bcotton> i'd think releasely would be a bit hectic
00:14:10 <Sparks> jjmcd: I think you should be able to get rid of me... err... a turkey at any point
00:14:20 * Sparks establishes https://fedoraproject.org/wiki/Docs_Project_Leader_Elections
00:14:28 <jjmcd> So we need an impeachment process?
00:14:45 <Sparks> WE DON'T LIKE YOU ANYMORE... GET OFF THE STAGE!
00:15:14 * jjmcd prefers burning torches
00:15:22 <Sparks> Okay, I'll send out an email after the meeting
00:15:35 <Sparks> #action Sparks to send out email on nominations
00:15:52 <Sparks> Okay, anything else on this topic?
00:16:19 <Sparks> #topic Release Notes
00:16:39 <jjmcd> You saw the notes to the list
00:16:42 <Sparks> First I'd like to thank jjmcd and rudi and everyone else who helped in getting the Release Notes completed.
00:17:10 <Sparks> You guys did an awesome job!
00:17:18 <Sparks> Now if our technology could keep up... :)
00:17:25 <Sparks> jjmcd: What are the translation issues?
00:17:27 <rudi> Ta -- there are still some bugs in the process, but we're getting better :)
00:17:27 <jjmcd> They do look nice on the new F12 background - but that's all on mo
00:17:46 <jjmcd> rudi is tracking them down, I think most are explainable
00:18:00 <jjmcd> We do need to decide when to publish
00:18:20 <jjmcd> and I like rudi's idea of moving in a direction of translators publishing, or at least deciding when
00:18:43 <jjmcd> And OBTW, as translations are committed, the built docs ARE available
00:18:53 <Sparks> jjmcd: I'd like it if the translators would join us on these meetings... especially around release time
00:19:01 <jjmcd> But without building themselves, translators have no view into the errors
00:19:20 <jjmcd> Yes, noriko at least was joining us for a while
00:19:25 <Sparks> jjmcd: So we need to do a classroom on how to use Publican?
00:19:32 <jjmcd> but they are spread out across timezones
00:19:37 <rudi> Sparks -- no, the problem is more deep-seated
00:19:42 <jjmcd> Well, it is a major PITA until the new TFX
00:19:56 <Sparks> jjmcd: True. Maybe the lead translator for each lang could join us... TZs make it tough
00:20:02 <rudi> yeah; Tx 0.7 is the blocker
00:20:12 <Sparks> rudi: more deep-seated?
00:20:26 <rudi> Until then, translators not only need to know how to use publican, but how to wrangle the merging and splitting as well
00:20:45 <jjmcd> And that isn't always obvious
00:20:47 <Sparks> yes
00:20:55 <rudi> and the "how to use Publican" is the least of their worries.
00:21:00 <Sparks> it isn't obvious to me!
00:21:11 <rudi> Even after all this time, *we* still bungle the merging and splitting
00:21:13 <jjmcd> When everything is hunky-dory, we could script it. But things are rarely perfect
00:21:34 <rudi> (which is the root of most of the English strings we're still seeing cropping up in the translations)
00:21:45 <jjmcd> Yep, that was my suspicion
00:22:05 <rudi> jjmcd -- yeah, confirmed. With yours truly being the chief bungler :)
00:22:43 <jjmcd> I will remind docs and trans later where to find the nightly builds of f12zeroday-tx
00:23:11 <rudi> But there are also some genuinely inexplicable ones, where strings we can see are definitely translated somehow don't get carried across the split...
00:23:11 <Sparks> jjmcd: When will the next batch ready to commit?
00:23:39 <rudi> Sparks, jjmcd -- we need to make a decision about when and whether to refresh the POTs
00:23:49 <jjmcd> rudi: yep. Sparks: Schedule says tomorrow, but I am poking L10N to ask if that makes sense
00:23:49 <rudi> And we probably need some input from L10N on that
00:24:11 <jjmcd> And with the latest flamebug, we may want to do that again real soon now
00:24:32 <Sparks> ya
00:24:39 <rudi> The refresh would pull in the latest fixes (from f12-temp), plus the flamebug, plus assorted small glitches that are breaking translations
00:25:14 <rudi> jjmcd -- did you want to do this before or after packaging the RN?
00:25:24 <jjmcd> I would also like to get something authoratative on processors. We have an outstanding bug that disagrees with other experts, and I don't know who to believe
00:25:45 <jjmcd> rudi, we have a package in the DVD, and a more recent package on updates
00:26:09 <jjmcd> We will want to do another probably for the install thing first en then maybe all langs
00:26:16 <rudi> jjmcd -- Ah OK! so any update is not going to interfere with thath
00:26:23 <rudi> * that
00:26:25 <jjmcd> I don't think we want to wait that one for the translations
00:26:42 <jjmcd> yeah, Sparks has been great on getting these things to the repos
00:26:45 <rudi> OK -- so regenerate today?
00:26:56 <jjmcd> sounds good to me
00:27:25 <jjmcd> There is more manual than I would like in the packaging, but it isn't some huge deal
00:28:08 <jjmcd> The real problem is waiting hours and hours to build all the langs
00:28:16 <jjmcd> Well, mostly to merge
00:28:28 <jjmcd> BTW, I saw you committed some merged pos
00:28:44 <jjmcd> If we can keep up with that, it could be a real help
00:29:16 <rudi> NP -- I'm trying to make sure that the master branch stays current and ready to build
00:29:24 <jjmcd> good, I like that
00:29:46 <jjmcd> http://fedora.is-sixsigma.com/F12beta/ - regenerated every night, from t12zeroday even tho it says beta
00:29:58 <rudi> So -- the other question is -- do we wait till we have an answer on the arches before rebuilding?
00:30:05 <rudi> *regenerating
00:30:30 <jjmcd> I think we need to deal with the install policy immediately
00:31:15 <rudi> We have text on the policy change ready to insert?
00:31:19 <jjmcd> As much as I would like an answer on the other, I think all those guys are burning their cycles on BZ534047
00:31:24 <jjmcd> No
00:31:28 <jjmcd> Well, there is text
00:31:35 <jjmcd> and widespread agreement it is wron
00:32:09 <rudi> OK -- well, we need something definitive in the f12-temp branch ASAP
00:32:19 <rudi> radsy -- are you in a position to write it up?
00:32:24 <jjmcd> So, I will craft some weasel words that try to avoid the emotion, but we still need a definitive workaround
00:32:43 <rudi> Yeah -- whether the change is good, bad, or ugly is not relevant to the RN
00:32:58 <jjmcd> The fix posted by Seth isn't even available to most people
00:33:03 <rudi> Just its effects and what to do if you don't like it for whatever reason
00:33:20 <jjmcd> You gotta be a policykit developer to even have the tool
00:33:27 <jjmcd> It is the what to do that is the problem
00:34:03 <jjmcd> Not too hard to describe the change
00:34:44 <rudi> OK
00:34:51 <jjmcd> Does that become 5.2.1, or 4.1?
00:34:55 <rudi> So where are we going to get that from?
00:35:16 <jjmcd> Well, the bug thread hasn't slowed yet
00:35:41 <rudi> I'd say 5.2.1, with a pointer in 4.1
00:35:59 <jjmcd> Oh maybe it has. I couldn't get a word in edgewise earlier, but now it's been an hour since someone commented
00:36:19 <jjmcd> On the bug, I'm sure the mailing list is still lively
00:36:39 <Sparks> jjmcd: Would now be a good time to ask for a definitive answer on what needs to go in the RNs?
00:36:48 <jjmcd> Or everyone got frustrated with the in-flight collisions on bugzilla
00:37:00 <jjmcd> Ask where?
00:37:13 <Sparks> jjmcd: The more I think about it the more I think that PK should just be uninstalled on systems where users shouldn't be installing software... almost makes sense.
00:37:16 <Sparks> jjmcd: On the bug.
00:37:24 <jjmcd> Seems like everyone has a horse in this race
00:37:25 <Sparks> jjmcd: you/we still need an answer
00:37:30 <jjmcd> yep
00:37:44 * Sparks got hit twice on the same post with a collision
00:37:52 <jjmcd> And we need to test it
00:38:09 <jjmcd> Much of what has been talked about doesn't work for ordinary mortals
00:38:49 <Sparks> jjmcd: I agree
00:39:10 <jjmcd> What I think we are saying, rudi, is that we aren't ready for a full rebuild
00:39:30 <Sparks> jjmcd: Well... work up some text saying what PK does and that a solution is in the works
00:39:31 <jjmcd> I'll post en-US as soon as we get an answer, and we can do the big thing tomorrow
00:39:38 <rudi> yeah, that's what I'm hearing.
00:39:44 <Sparks> jjmcd: If you get the solution before you crash for the night then put it in...
00:39:53 <Sparks> jjmcd: otherwise build it and let's do a push
00:40:02 <Sparks> ok
00:40:02 <jjmcd> Yeah, at least on docs.fp.o I can put a warning
00:40:10 <Sparks> WORKSFORME
00:42:26 <Sparks> Okay, anything else on the Release Notes?
00:42:42 <jjmcd> rudi, any clue why the rev history is missing from the trans?
00:43:11 <rudi> jjmcd -- yeah; I turned them off
00:43:16 <jjmcd> Oh, OK
00:43:18 <rudi> when I was building
00:43:38 <rudi> the changes that you had to make for yelp caused them to blow up when I was building the translated versions
00:43:47 <jjmcd> I had no theory about that, as long as there's an answer and it isn't a mystery
00:43:57 <rudi> And, strapped for time, that seemed like the best solution
00:44:06 <jjmcd> oooohhhh, I thought I kept the yelp changes out of git except for the table length
00:44:10 <rudi> Yeah; no mystery
00:44:38 <rudi> Hmmm.
00:45:01 <rudi> So the "xyz changes (cont'd)" shouldn't be in there either?
00:45:09 <jjmcd> I cut the max table size down some in the hopes it might help - didn't
00:45:20 <jjmcd> Oh yeah, just the tables were shorter
00:45:24 <rudi> Oh OK
00:45:36 <jjmcd> I think I was limiting them to like 250 entries or something, shortened to 100
00:46:05 <jjmcd> didn't help,.. but I also hoped they might help you build, too
00:46:20 <rudi> Well, *something* helped me build
00:46:40 <rudi> I was finally able to build them on my machine without having to SSH into ryanlerch's :)
00:46:42 <jjmcd> Don't you love a good mystery
00:49:29 <Sparks> Okay... Moving on...
00:49:38 <Sparks> #topic Guide Status
00:49:53 <Sparks> Anyone have any problems with their guide or want to talk about their guide?
00:50:37 <radsy> i did lots of updates to the selinux user-guide yesterday, thanks to a BZ that slipped through the cracks of my memory
00:50:46 <radsy> got it up in CVS overnight, it's fine now
00:50:55 <Sparks> radsy: Cool!
00:51:01 <Sparks> Anyone else?
00:51:10 <radsy> basically the removal of setroubleshootd as a permanent daemon
00:52:18 <Sparks> :)
00:52:28 <Sparks> #topic New Guides
00:52:33 <Sparks> Any new guides in the works?
00:53:03 <Sparks> #topic Status on CC license rollout.
00:53:11 <Sparks> For those of you that didn't know...
00:53:28 <Sparks> We have officially changed over to the CC-BY-SA license!
00:53:34 <Sparks> All of these tasks are complete!
00:53:56 <Sparks> Thanks to everyone who made this possible... it was truly a team effort.
00:54:14 <Sparks> #topic All other business
00:54:28 <Sparks> And with six minutes left... does anyone have anything else they'd like to discuss?
00:55:10 <radsy> https://bugzilla.redhat.com/show_bug.cgi?id=534047 ? :)
00:55:11 <buggbot> Bug 534047: high, low, ---, richard, ASSIGNED, All users get to install software on a machine they do not have the root password to
00:55:41 <radsy> i'm starting to think we might want to wait for the dust to settle
00:56:04 <radsy> security response team are trying to find out if it will be reverted
00:56:21 <radsy> nobody knows it seems...
00:56:35 <Sparks> radsy: IMHO we need to get the word out that right now the vulnerability exists.
00:56:53 <radsy> your call
00:57:02 <Sparks> if admins/owners don't know that there is a change then they are assuming that they are protected.
00:57:20 <Sparks> but, of course, that is my opinion...
00:57:31 <radsy> we do have a lot of hits on the RN right now, being so close after release
00:57:54 <Sparks> Yeah
00:58:17 <radsy> i guess it can't hurt, quick enough to pull if needed
00:58:27 <Sparks> anyone have any thoughts on how Docs should handle the PK ordeal within the next 24-hours?
00:58:42 <Sparks> I'm assuming that within 72-hours we'll know more and can stand down more.
00:58:55 <Sparks> radsy: ya... we can get text up and down very quickly
00:58:56 <rudi> Well, we could add a note about the *existence* of the issue right now
00:59:08 <Sparks> rudi: Yeah, that's basically what I'm thinking right now
00:59:13 <rudi> Even if we don't have a suggested fix
00:59:35 <rudi> I kinda think that if there were a fix out there that most people could implement, we would have heard about it by now...
00:59:42 <radsy> and mention it's still in contention
00:59:50 <rudi> radsy +1
01:00:05 <Sparks> yep
01:00:33 <Sparks> Okay... it is the end of our hour. Anything else to discuss before we close?
01:00:55 <radsy> nice timing :)
01:01:16 <Sparks> Thanks everyone for coming tonight!
01:01:19 <Sparks> #endmeeting
14 years, 5 months
More details on Issues with Release Notes translations
by John J. McDonough
Going through all the languages on docs.fp.o I find some interesting things.
pt-BR is marked draft and is, in fact, older than the others. It is missing
the sections that are not translated in the other languages. I think the
discoonect here is that on the first push to docs.fp.o, I pushed all
languages that were at greater than 90%. Since then we have only pushed
100%. pt-BR is still only at 92%.
We probably should bring up on the trans list whether we want to wait for
100%. The French team in particular objected to pushing partly translated
documents. I suspect some of the others may want to see mostly translated
documents published.
In all the other languages, 4.1.6, 5.2.2, 5.6.1, and 5.7.1 are not
translated
All are missing the revision history
Titles for 7.5 and 8.9 are untranslated
Title for 4.4 is untranslated in all except Chinese and Russian
Title for 4.1 is untranslated in Dutch
On the copies I ran last night (http://fedora.is-sixsigma.com/F12beta/), I
only checked Dutch, Portuguese, Polish and Spanish. On those, all of the
sections untranslated on docs.fp.o were translated. In Dutch, Portuguese
and Polish, the title for 8.9 was translated.
OK, so now I know less than I did before!
--McD
14 years, 5 months
