From Fedora Project Wiki

< QA‎ | Meetings

Revision as of 20:50, 1 February 2010 by Jlaska (talk | contribs) (Updated notes)

The printable version is no longer supported and may have rendering errors. Please update your browser bookmarks and please use the default browser print function instead.

Attendees

People present (lines said)

  • jlaska (163)
  • kparal (48)
  • adamw (45)
  • wwoods (31)
  • Oxf13 (23)
  • maxamillion (7)
  • nirik (4)
  • zodbot (3)
  • skvidal (2)
  • poelcat (1)

Regrets:

Agenda

Previous meeting follow-up

  1. jlaska to add bugzilla shared query links to Common_Bugs wiki page
  2. wwoods and kparal to talk about design ideas
    • No discussion took place yet, but kparal noted he'd like to have a small group meet and discuss requirements. Wwoods agreed and suggested outlining some requirements and goals before we start looking at tools/implementations
  3. adamw update on security policy

Release validation wiki updates

rhe and awilliam coordinated to improve the wiki presence of several release validation efforts, including installation test and desktop sanity test. Adamw gave a quick update noting that so far we've added the installation validation testing wiki page (QA:Installation_validation_testing), a a desktop validation testing wiki page, which is still full of duct tape (QA:Desktop_validation_testing), and modified the QA/Join page to mention this testing (QA/Join#Release_validation).

Jlaska asked if we should change the install-specific nature of the QA schedule to reference the more all-encompassing release validation efforts. Adam agreed.

Adam concluded noting there will be other types of validation we could do in future; there's already a couple of criteria we can test that don't really come under installation or desktop groups.

Rawhide Acceptance Testing

Last week, 2010-01-28 was the second pre-Alpha rawhide acceptance test milestone. Jkeating and dlehman coordinated to do a second compose (see rel-eng ticket#3292). Install images were built and posted on alt.fedoraproject.org. Since AutoQA doesn't yet monitor this location for new install images, tests were initiated manually.

Test Summary:

Next Steps:

  • Fix AutoQA post-tree-compose hook to monitor for RAT droppings (see ticket#115)
  • 2010-02-04 test run

Questions/Concerns:

  • Jkeating asked what does 'last known good' (LKG) mean? - Does it mean that the install images and package source are good? Jlaska felt that for QA, it means here's we tested a particular install image and package set, and it was good. Adamw offered to help document this on the wiki, but asked for clarification/guidance once the mechanics have settled down.
  • Jlaska asked how do we know we're done with the 'last known good' (LKG) project?, Jkeating offered ...
   <Oxf13> in my head there is a simple webpage that is maintained by QA
   <Oxf13> this webpage describes Last Known Good and the fact that rawhide no
   longer has images of it's own
   <Oxf13> onthis page would be links to the install image files, and to the tree
   of packages it was found to be good with
   <Oxf13> with instructions on how to combine the two into an install
   <Oxf13> and finally notes that without picking a specific repo of packages, the
   images will attempt to install the latest repo of packages on the mirror system
   <Oxf13> which has not been validated, and may not work

AutoQA project update

Deps/conflicts prevention

Last week

This week

  • Build a testsuite to validate depchecks functionality (possibly using rpmfluff)

rpmguard and autoqa results collection

Last week

  • Address ticket#113 - skip checks against same package NVR
  • Address ticket#114 - locate latest previous package from stable and updates repos. Makes it easier to run the test manually at some future date.

This week

  • Initial brainstorming on results db ideas

install automation

Last week

This week

  • Continue discussion around dvd-install test case and identify how to proceed

packaging/deployment

Last week

This week

  • Seek guidance from upstream GWT for ...
    1. Cases where multiple versions of a JAR are bundled
    2. Need for remaining bundled JARs? (see User:Jlaska/gwt#Status_uncertain)
    3. Begin packaging JPackage JAR files for Fedora

Open discussion - <Your topic here>

Upcoming QA events

Action items

  1. request zodbot tracking for updated F-13 QA RSS feeds (blocker bugs, common_bugs? etc...
  2. jlaska to organize a gobby meeting with wwoods+kparal to begin requirements/goal definition for a results db
  3. jlaska to discuss updating QA schedule with poelstra to reflect new 'release validation' focus
  4. jlaska and adamw to work on drafting a 'last known good' wiki page

IRC transcript

jlaska #startmeeting Fedora QA Meeting 16:01
zodbot Meeting started Mon Feb 1 16:01:41 2010 UTC. The chair is jlaska. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:01
jlaska #meetingname qa 16:01
zodbot Useful Commands: #action #agreed #halp #info #idea #link #topic. 16:01
zodbot The meeting name has been set to 'qa' 16:01
jlaska #topic Gathering 16:01
adamw yo 16:02
jlaska adamw: happy monday :) 16:02
jlaska who else can we convince to join us for a meeting? 16:03
* kparal convinced 16:03
adamw yeah, yeah, keep it short, whatever, i have important cellphone hacking to do :P 16:03
jlaska I can see that! :P 16:03
jlaska adamw: you must join the meeting using one of those phones 16:03
kparal adamw: you are a blogging superman :) 16:03
* maxamillion is here 16:04
adamw jlaska: gimme 2 minutes and i will 16:04
jlaska hi kparal, maxamillion 16:04
jlaska adamw: uh oh ... that wasn't meant as a dare :) 16:04
jlaska who else we have? wwoods Viking-Ice tk009 ? 16:05
jlaska alright, let's get started and hopefully they can join in 16:06
jlaska #topic previous meeting follow-up 16:06
jlaska #info jlaska to add bugzilla shared query links to Common_Bugs wiki page 16:06
jlaska #info beland added the links to the F-13 page already ... https://fedoraproject.org/wiki/Common_F13_bugs 16:07
jlaska The only other thing that occured to me was perhaps updating our IRC RSS feeds that zodbot tracks 16:08
* nirik can do that later today if you let me know what to change it to/etc. 16:08
jlaska I'll take an action item to grab the blocker bug RSS feeds and see if nirik can update zodbot to watch those links 16:08
jlaska nirik: thanks! 16:09
nirik no problem. 16:09
jlaska #action request zodbot tracking for updated F-13 QA RSS feeds (blocker bugs, common_bugs? etc...) 16:09
* adamw now meetingggggg on his phone :) 16:09
nirik I noticed the other day it spewed something about the f12 blockers... meant to look at it. 16:09
jlaska nirik: we can probably replace the F-13 ones with the new feeds? 16:09
jlaska adamw: you're nuts :) 16:10
jlaska okay, next up I have ... 16:10
adamw Pic available if proof desired! 16:10
jlaska #info wwoods and kparal to talk about design ideas 16:10
nirik f12 you mean? yeah. I think currently it's announcing any change to the f12 blocker 16:10
jlaska I think this came out of the AutoQA discussion last week 16:10
jlaska nirik: yeah, sorry ... replace the F-12 ones with the newer F-13 feeds 16:10
jlaska kparal: wwoods: do you have insight as to what was being tracked by this action item? 16:11
kparal jlaska: I believe it meant we should have consulted the future results database design? 16:11
kparal jlaska: which we didn't :/ 16:12
adamw http://www.happyassassin.net/extras/irc_meeting.jpg 16:12
kparal adamw: nice! 16:13
jlaska kparal: ah right ... hmm, what are your thoughts there. what is needed to do a good design review on that? 16:13
jlaska adamw: that's cute :) 16:13
maxamillion adamw: what irc client is that? 16:13
jlaska adamw: normally, I'd say this would impact your words/minute. But I don't think it will :) 16:13
adamw maxamillion: andchat 16:14
maxamillion adamw: interesting ... I'll have to check it out 16:14
maxamillion :) 16:14
adamw jlaska: i'm cheating, i'm typing on my pc now :P 16:14
wwoods jlaska: yeah we were going to talk about design ideas for the results database 16:14
kparal jlaska: I believe several should get involved, so we don't make any oversights right in the design. maybe some people that already designed some koji parts could help us 16:14
adamw maxamillion: it vibrates for nick alerts - stop buzzing my knee :D 16:14
maxamillion adamw: when did you get an android device? 16:14
maxamillion LOL 16:14
kparal *several people 16:14
adamw maxamillion: > #fedora-qa let's not derail 16:14
maxamillion rgr, sorry 16:14
jlaska kparal: re: koji ... do you mean like lmacken? 16:16
jlaska perhaps a micro-FAD would be helpful to get the right folks involved 16:16
kparal jlaska: I don't know about lmacken, for e.g. dmach hinted me to look at his kobo library several times. https://fedorahosted.org/kobo/ 16:17
jlaska kparal: aah right 16:17
wwoods ugh 16:17
wwoods xmlrpc with cookies? krb5? 16:17
kparal jlaska: it might be some inspiration for us, I think there is a hub part that store results 16:17
jlaska kparal: probably a good idea for us to outline what our needs are first? 16:17
Oxf13 kobo looked waaaay to overengineered for the stuff I needed 16:18
wwoods kobo is a great idea for building RHEL-internal apps 16:18
kparal I haven't studied it yet 16:18
wwoods but very poorly suited to fedora needs 16:18
kparal just looking for places where we can gather some ideas about the results DB 16:18
skvidal ugh 16:19
wwoods yeah it's still worth looking at other things that are around 16:19
kparal jlaska: yes, we should brainstorm about the needs first 16:19
skvidal b/c we need ANOTHER implemention of rpm header parsing 16:19
wwoods but there's no reason the results db needs to be able to parse RPM headers 16:19
kparal don't dump it everyone, we can have a look at some specific parts, not use everything at once :) 16:20
wwoods honestly there's no reason it needs to be anything more than, say, a TG app, as far as I can see 16:20
wwoods but that's still a very very broad canvas for design 16:20
jlaska in the interest of time ... is there a next step to keep track of here? 16:20
kparal wwoods already has some expertiese with irb frontend, so I suppose he will be the leader in this area :) 16:20
wwoods I'd suggest we should outline some requirements and goals before we start looking at tools/implementations 16:21
jlaska that seems sensible 16:21
kparal the next step should be the brainstorming meeting for those of us concerned I think 16:21
wwoods kparal: agreed 16:21
kparal we just must not forget about it :) 16:21
jlaska would you like me to organize a gobby meeting this week to review? 16:22
kparal sounds good for me 16:22
wwoods sure 16:22
jlaska #action jlaska to organize a gobby meeting to begin requirements/goal definition for a results db 16:22
jlaska alright ... last follow-up item ... 16:23
jlaska #info adamw update on security policy 16:23
jlaska The info I have is Adam moved the discussion from test@l.fp.org to devel@l.fp.org (http://lists.fedoraproject.org/pipermail/devel/2010-January/129978.html) 16:24
jlaska adamw: any other updates? 16:24
adamw sorry! 16:24
adamw nope, that's it - i did that before the weekend so haven't read the responses yet 16:25
jlaska adamw: no TXT'ing while MTG :) 16:25
adamw i intend to follow the same process i did for test list 16:25
jlaska okay, should we keep this on the radar for next week? 16:25
adamw do a new draft from the current responses, then send that, keep cycling until no-one has anything to complain about :) 16:25
adamw sure 16:25
jlaska #info adamw will continue monitoring feedback (http://lists.fedoraproject.org/pipermail/devel/2010-January/129978.html) and propose updated drafts as needed 16:26
jlaska alright, let's dive in ... 16:26
jlaska #topic Release validation wiki updates 16:26
jlaska adamw: I've got you on the spot for this as well, can you cover for rhe? 16:26
jlaska Just wanted to give a quick update on the wiki changes you and rhe have been working on 16:26
adamw yup 16:27
adamw so far we've added the installation validation testing wiki page: 16:27
adamw #link https://fedoraproject.org/wiki/QA:Installation_validation_testing 16:27
adamw a desktop validation testing wiki page, which is still full of duct tape: 16:27
adamw https://fedoraproject.org/wiki/QA:Desktop_validation_testing 16:27
adamw and modified the QA/Join page to mention this testing: 16:27
jlaska heh, the many uses of duct tape :) 16:28
adamw https://fedoraproject.org/wiki/QA/Join#Release_validation 16:28
jlaska adamw: I wanted to ask, would you like to have the QA schedule changed to reference 'release validation' and not specifically 'install validation' so there is room for other efforts? 16:29
jlaska okay, we can come back to that if needed, lemme keep on going 16:31
jlaska nice wiki updates rhe and adamw :) 16:31
adamw jlaska: i think that would be a good change 16:31
jlaska adamw: okay, I'll ping get in touch with poelcat later this week then 16:31
jlaska #action jlaska to discuss updating QA schedule with poelstra to reflect new 'release validation' focus 16:32
adamw there will be other types of validation we could do in future; there's already a couple of criteria we can test that don't really come under installation or desktop 16:32
jlaska I'd love to get off my butt and do the QA release checklist 16:32
jlaska to include in that mix 16:32
jlaska adamw: anything else on the wiki front? 16:33
adamw um, nothing springs to mind 16:33
adamw was that a leading question? :) 16:33
adamw i didn't get around to updating commonbugs yet :( 16:33
jlaska no, not a leading question 16:34
jlaska alright, moving on to another quick update ... 16:34
jlaska #topic Rawhide Acceptance Testing 16:34
jlaska we had another RATS drop last Thursday 16:34
jlaska jkeating composed a tree for rats_install (see rell-eng ticket#3292). 16:35
jlaska I had to correct a few typos with the rats_install patches designed to support providing stage2= and repo= ... but all-in-all it worked well 16:35
jlaska #link https://fedoraproject.org/wiki/Test_Results:Fedora_13_Rawhide_Acceptance_Test_2 16:35
Oxf13 did it pass? 16:35
jlaska Oxf13: well it passed, but there's a _but_ 16:36
jlaska bug#559597 prevents initramfs creation ... so it won't boot without that fixed 16:36
Oxf13 oh right 16:36
jlaska so one dracut-004-5 lands, we should be good 16:36
Oxf13 but you can use the images with a later rawhide tree 16:36
jlaska at least for the automated portion of rats_install 16:37
jlaska Oxf13: right, the packages from a newer rawhide 16:37
jlaska I don't think we need to update the install images for this, right? 16:37
jlaska Oxf13: so how would you like to proceed ... want a ticket to update the 'last known good' ? 16:38
Oxf13 jlaska: we probably shouldn't do that until the package tree has the right dracut level and we verify that 16:38
jlaska Oxf13: oh agreed 16:38
Oxf13 this does bring up a good question though 16:38
Oxf13 does "last known good" mean a set of images, a set of packages, or the combination of the two? 16:38
jlaska #info Oxf13 asked ... does "last known good" mean a set of images, a set of packages, or the combination of the two? 16:39
jlaska from a testing perspective ... it means images + packages 16:39
jlaska as soon as one changes, I think that adds a caveat to the 'last known good' 16:39
Oxf13 I suppose for users it's useful to have images which we expect to work, which can be tested against package sets which we're not sure of 16:39
Oxf13 all releng would be tracking and providing symlinks to is the set of images 16:40
jlaska Oxf13: true, so this is intended to increase install reliability ... but is still open to rawhide package churn 16:40
adamw i think it's as simple as putting a rawhide date on the 'last known good' listing 16:40
adamw and a note that says it was known to work with that day's packages 16:41
adamw may work with other days, but we can't say for sure 16:41
jlaska agreed, I think we just need some wording around this so it's clear 16:41
* poelcat thought LKG was a fully installable source, including images so people did not have to go all the way back to F12 for a reliable starting place 16:42
jlaska think it's time to get some documentation down for this 16:42
jlaska any volunteers :) 16:42
adamw i can do it if you let me know roughly what's needed 16:43
jlaska adamw: given the questions here, I think defining what 'last known good' means 16:43
jlaska it's a term we use a lot, but sounds like there is still some confusion around what's included 16:44
Oxf13 part of that was just working through the mechanics of actually doing it 16:45
jlaska true true 16:45
Oxf13 it's far easier to store a small set of images, and then reference an existing tree of packages somewhere 16:45
Oxf13 or let the images autopick the latest rawhide for more fun 16:45
adamw i think this is a case where we should do the mechanics and then i will write an explanation which sounds like we knew what we were doing all along =) 16:45
jlaska okay, in that case I think we're probably done then 16:46
jlaska Oxf13: are there other changes coming on the rel-eng side of building these? 16:46
Oxf13 um... 16:46
Oxf13 how do you mean? 16:46
jlaska how do we know when we're done with providing 'last known good'? 16:47
jlaska how can we [X] check it off 16:47
* adamw is the king of retrospective-justification-through-wikis 16:47
jlaska adamw: it's still 20/20 right? :) 16:47
Oxf13 jlaska: in my head there is a simple webpage that is maintained by QA 16:47
Oxf13 this webpage describes Last Known Good and the fact that rawhide no longer has images of it's own 16:48
Oxf13 onthis page would be links to the install image files, and to the tree of packages it was found to be good with 16:48
Oxf13 with instructions on how to combine the two into an install 16:48
adamw if you give me canonical locations for lng installer images and package tree i can totally make a wiki page for that. 16:48
adamw also, those instructions :D 16:49
jlaska Oxf13: ah that's helpful 16:49
Oxf13 and finally notes that without picking a specific repo of packages, the images will attempt to install the latest repo of packages on the mirror system 16:49
Oxf13 which has not been validated, and may not work 16:49
jlaska alright, adamw I'll be happy to work the details with you 16:49
adamw okay 16:50
adamw action it? 16:50
jlaska #info jlaska and adamw to work on drafting a 'last known good' wiki page 16:50
jlaska err ... 16:50
jlaska #action jlaska and adamw to work on drafting a 'last known good' wiki page 16:50
jlaska Oxf13: thanks ... will probably pick your brain again during the week 16:50
Oxf13 no problem 16:50
jlaska okay ... time for AutoQA update! 16:50
* jlaska hits the gong 16:51
jlaska #topic AutoQA project update 16:51
jlaska #topic AutoQA project update - deps/conflicts prevention 16:51
jlaska wwoods: you want to kick things off for the AutoQA update? 16:51
wwoods sure 16:51
wwoods Friday afternoon I got a working yum-based prototype of the depcheck code 16:52
wwoods http://git.fedorahosted.org/git/?p=autoqa.git;a=blob;f=tests/depcheck/depcheck 16:52
jlaska saweet! 16:52
wwoods a mere 147 lines, even 16:52
wwoods I need to start creating test cases to make sure it works as expected 16:53
wwoods typical runtime is ~20s 16:53
jlaska wow, you really brought that down from the 60s 16:53
wwoods ~15s for leaf-node packages (e.g. something like hamster-applet) 16:53
wwoods ~19-20s for packages with lots of prov/req data (e.g. gcc) 16:54
jlaska that's pretty darned good 16:54
wwoods but "20s" is a safe estimate 16:54
wwoods probably it can be sped up further 16:54
wwoods but premature optimization is the root of all evil, and all that, so.. later 16:54
jlaska heh 16:54
* jlaska stuffs a few #info tags into the mix 16:55
jlaska #info Friday afternoon I got a working yum-based prototype of the depcheck code 16:55
jlaska #link http://git.fedorahosted.org/git/?p=autoqa.git;a=blob;f=tests/depcheck/depcheck 16:55
jlaska not that this is blocking your efforts, but I'm waiting for some feedback from jhutar, but rpmfluff should be an official fedora package soon 16:55
wwoods excellent 16:56
wwoods yeah I think I'd like to use rpmfluff etc. to generate test packages/repos 16:56
wwoods for the test cases 16:56
jlaska that'll be great to see 16:56
wwoods yum's test cases involve creating objects which *simulate* test packages/repos/etc. 16:56
wwoods not sure if that makes a significant difference 16:57
jlaska just having a testsuite at all for this I think will pay off in the long run 16:57
wwoods definitely. 16:57
jlaska given the complexity 16:57
jlaska nice work! 16:57
jlaska wwoods: anything else on the depcheck front? 16:58
wwoods no, I think that covers it 16:58
jlaska sweet, alright ... kparal's up next 16:58
kparal ok 16:59
jlaska #topic AutoQA project update - rpmguard and resultsdb 16:59
kparal so I have updated a few tickets regarding rpmguard 16:59
kparal first is https://fedorahosted.org/autoqa/ticket/113 16:59
jlaska take it away kparal 16:59
kparal when comparing the exactly same packages the rpmguard will now just skip the comparison and print a notice that they are the same 16:59
kparal that could save a little computing power 17:00
jlaska aha, nice ... that test run confused the heck out of me :) 17:00
kparal it was connected to our problem of comparing the same package unintentionally 17:00
kparal and second ticket connected to that is https://fedorahosted.org/autoqa/ticket/114 17:01
kparal before we compared the current package with the latest package from stable, updates or updates-testing 17:01
kparal now we compare only to stable and updates, as per discussion on autoqa-devel 17:02
jlaska nice to see all this stuff getting hashed out 17:02
kparal and also we compare to the latest *previous* package (so it should not happen to compare two equal packages) 17:02
kparal this also allows us to run the test any time in the future, even against older packages (in that time already in stable or updates, not in updates-candidate) and receive the same results 17:03
jlaska oh so if I ran that test again, it should be grabbing ... yup, you just answered it 17:03
kparal this can be beneficial when we find some problem in autoqa/rpmguard and need to re-run the test suite again 17:03
jlaska #info I have updated a few tickets regarding rpmguard 17:04
jlaska #link ttps://fedorahosted.org/autoqa/ticket/113 17:04
jlaska #link ttps://fedorahosted.org/autoqa/ticket/114 17:04
jlaska #link https://fedorahosted.org/autoqa/ticket/113 17:04
jlaska #link https://fedorahosted.org/autoqa/ticket/114 17:04
adamw side note on rpmguard: we should talk about the privesc stuff at some point, since that looks like it may turn into a Real Thing 17:04
* jlaska gets flagged for incorrect use of meetbot 17:04
kparal adamw: privesc? 17:04
jlaska #info adamw discussed possible rpmguard and privilege escalation coordination 17:05
kparal ah 17:05
jlaska kparal: that's the security policy Adam'w been working 17:05
kparal just didn't know the abbreviation :) 17:05
Oxf13 jlaska: fyi, meetbot picks up links automatically, you don't have to #link them 17:06
jlaska Oxf13: ah, thx 17:06
kparal ok, that's all from me 17:06
jlaska kparal: also the results db discussion earlier, hopefully I can help get some of you and wwoods ideas on "paper" this week 17:07
jlaska kparal: thanks for the updates! 17:07
kparal e-paper = wiki :) 17:07
jlaska you got it 17:07
jlaska #topic AutoQA project update - install automation 17:07
jlaska #info Liam added a dvd_install.py script to the autoqa git repo last week 17:08
jlaska https://fedorahosted.org/pipermail/autoqa-devel/2010-January/000178.html 17:08
jlaska it raises some interesting questions for me on how we'll approach automating the install matrix 17:08
jlaska wwoods: if you have some time this week, I could use your input on that thread as well 17:09
wwoods jlaska: yeah I've been meaning to give that a closer look 17:09
jlaska ah thank you 17:09
jlaska you know the usual ... trying to figure out what path is the most sustainable and scalable for building new install tests :) 17:10
jlaska Liam has some more thoughts on the mailing list, I'll try to reply later today 17:11
jlaska so that's great news, hopefully more install tests to come soon 17:11
jlaska alright ... last up ... 17:11
jlaska #topic AutoQA projet update - packaging 17:11
jlaska #info Process akurtakov's feedback and remove all uncertain JAR files from the list 17:11
jlaska I've got the list down to as small as I can get it ... adamw suggested reaching out to GWT upstream for some additional feedback 17:12
jlaska so the plan this week is to Seek guidance from upstream GWT for ... 17:12
jlaska * Cases where multiple versions of a JAR are bundled 17:12
jlaska * Need for remaining bundled JARs? (see https://fedoraproject.org/wiki/User:Jlaska/gwt#Status_uncertain) 17:12
jlaska and ideally, I'd like to have one of the listed JPackage dependencies already started through the Fedora packaging process 17:13
jlaska https://fedoraproject.org/wiki/User:Jlaska/gwt#JPackage_Dependencies 17:13
jlaska aside from that ... I've been going through the gwt.spec and removing JAR files and linking them to their equivalents (and adding BuildRequires) 17:14
jlaska alright ... open discussion time 17:14
jlaska #topic Open discussion - <your topic here> 17:14
jlaska I'll close out the meeting in 2 minutes unless any topics come up 17:14
kparal ok, one topic from me 17:14
jlaska kparal: sure, what's up? 17:14
kparal package update acceptance test plan 17:14
kparal I just have a quick call for links :) 17:15
jlaska oh oh, right on 17:15
jlaska #topic Open discussion - Package Update Acceptance test plan 17:15
jlaska #info kparal put out a call for links 17:15
jlaska kparal: what sort of links are you looking for? 17:15
kparal I'm currently working on package update acceptance test plan, that means what should be our approach to decide if some package update is good or not 17:15
kparal I am currently looking into other distributions how they do it and trying to gather some inspiration 17:16
kparal if somebody is aware of some documented policies and have links for them, I will be very glad if you send them to me 17:16
kparal it is quite possible that I don't find everything that is available 17:17
jlaska kparal: I wonder if the MUST and SHOULD items could provide some insight - http://fedoraproject.org/wiki/Packaging:ReviewGuidelines 17:17
kparal yes, currently I have found mainly requirements for the initial review process 17:17
kparal but I haven't found many documents concerning package updates requirements 17:18
adamw i'm sure debian must have a policy 17:18
jlaska perhaps what wwoods is working on might be included in the mix? the depscheck stuff 17:18
kparal it is possible that other distributions also don't have strict policies about that 17:18
kparal but - if you know about such a policy and have a link, please send it to me. thanks :) 17:18
kparal that was all from me, just a call for links 17:19
adamw kparal: http://wiki.mandriva.com/en/Policies/SoftwareMedia 17:19
adamw kparal: i wrote that for MDV back when I was there 17:19
jlaska Oxf13: you have any handy links? 17:19
adamw kparal: see the definitions for the 'updates' repositories 17:19
kparal adamw: thanks, will look at that 17:19
adamw kparal: also see http://wiki.mandriva.com/en/Policies/Support , the full policy for maintainers about updates 17:19
adamw kparal: that was written by vincent danen when he was at mandriva, he also now works at red hat so you could always poke him for thoughts :) look for vdanen 17:20
kparal adamw: great, thx 17:20
Oxf13 jlaska: I don't 17:20
jlaska #topic Open discussion - <your topic here> 17:21
jlaska anything else to cover, if not let's close it out 17:21
adamw nothing frm me 17:22
jlaska alright gang, thanks for your time! 17:22
jlaska I'll follow-up to the list with minutes etc... 17:22
jlaska #endmeeting 17:22

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!