ReleaseStatus Minutes 2010-05-03 IRC log

From Apache OpenOffice Wiki
Jump to: navigation, search


(14:54:26) Das Thema für #oooreleases ist: Release-Stauts-Meeting, every monday at 15:00 Hamburg Time (13:00 UTC in summer)
(14:56:44) _Nesshof_ [~martin@nat/sun/x-zbjjzbjrkjnwogbu] hat den Raum betreten.
(14:56:57) ja_: Moin
(14:57:48) _KAMI_ [~kami@host-50-018.comunique.hu] hat den Raum betreten.
(14:57:59) dtardon [~dtardon@nat/redhat/x-pglgbbnownqgjfde] hat den Raum betreten.
(14:58:26) rtimm [~Ruediger@nat/sun/x-gsxfhilzwtundmim] hat den Raum betreten.
(14:59:40) ml1 [~ml93712@nat/sun/x-ejizolgknxfiwwvi] hat den Raum betreten.
(14:59:47) ml1: hi
(14:59:49) kai_a [~ka@nat/sun/x-jubzgprswemzqqpc] hat den Raum betreten.
(15:00:33) mdamboldt: Hi
(15:00:54) UweL [~chatzilla@nat/sun/x-wiaeconzfkabvylo] hat den Raum betreten.
(15:01:40) mdamboldt: Welcome to todays release status meeting. Let's see whats on todays agenda.
(15:01:46) mdamboldt: First of course the 3.2.1 release
(15:02:07) mdamboldt: A couple of CWS have been approved by QA the past days.
(15:02:53) mdamboldt: One last one is remaining right now, CWS ooo321l10n4. Which is targeted to get it's QA approval during today.
(15:03:27) mdamboldt: So it's looks very good we will be able to start the RC1 round (OOO680m17) tomorrow.
(15:04:10) MechtiIde: why we always have the l10n fix always in the RC and not before?
(15:04:10) mdamboldt: On todays agenda Stefan Weigel again proposed another stopper. Issue 111205
(15:04:15) IZBot: Database access DEFECT NEW P2 Inserted data is not shown. Error "No data is available" http://qa.openoffice.org/issues/show_bug.cgi?id=111205
(15:04:37) MechtiIde: mdamboldt, for issue 111205 is a patch available
(15:04:39) IZBot: Database access DEFECT NEW P2 Inserted data is not shown. Error "No data is available" http://qa.openoffice.org/issues/show_bug.cgi?id=111205
(15:04:45) Stefan_W: ...since a patch is available.
(15:04:48) MechtiIde: and testet by Stefan_W and me
(15:04:58) MechtiIde: we can approve it
(15:06:16) ***mdamboldt Wondering where the patch is, can't find it attached to the issue nor can I find a CWS where the Issue is registered.
(15:06:53) Stefan_W: somebody called oj created a build with a patch
(15:06:57) MechtiIde: ask oj because he makes us a build with the patch available
(15:07:36) MechtiIde: mdamboldt, he tagged it as a patch for dba321b
(15:08:13) MechtiIde: I don't know how he want to organize this mdamboldt
(15:09:02) Stefan_W: oj presented this patched build last week Thursday
(15:09:41) ja_: Mechtilde: CWS dba321b has been integrated into OOO320m16
(15:10:12) MechtiIde: yes, thats because I said " a patch for dba321"
(15:10:38) MechtiIde: It is irritating I know
(15:11:00) MechtiIde: is it a help I post the link to the build?
(15:11:29) rafaella [~Rafaella@nat/sun/x-athnrinpqlncrfyu] hat den Raum betreten.
(15:11:32) ***Stefan_W got this build ftp://qa-upload.services.openoffice.org/dba321b/ and in this build 111205 is fixed
(15:12:41) MechtiIde: maybe msc_sun can give informations
(15:13:14) enoki [~enoki@i114-189-83-251.s10.a028.ap.plala.or.jp] hat den Raum betreten.
(15:13:38) _Nesshof_: do we have a patch as source code ?
(15:13:39) msc_sun: MechtiIde: not more information then you and stefan provide.
(15:13:54) _Nesshof_: afaik oj is not in the office this week
(15:13:59) MechtiIde: is oj available, msc_sun ?
(15:14:01) Stefan_W: grmpf :-(
(15:15:14) mdamboldt: Stefan_W: The binaries oj provided to you via FTP are a modified CWS dba321b but not the one we already integrated.
(15:15:39) MechtiIde: it's a modified version, yes
(15:15:57) mdamboldt: Stefan_W: So we don't have a 3.2.1 CWS at hand containing that fix.
(15:16:15) MechtiIde: that's bad :(
(15:16:32) mdamboldt: So no CWS to test, approve, nominate and integrate into the RC1 build.
(15:17:10) MechtiIde: why we always have the l10n fix always in the RC and not before?
(15:18:26) kai_a: MechtiIde: oj is away for the whole week
(15:19:00) mdamboldt: So let me try to summarize 111205 status here again.
(15:23:46) ja_: ping
(15:23:47) mdamboldt: We already discussed this issue in the issue itself and we already had it nominated on the mailinglist some days ago. As Ocke (oj) and Marc (msc_sun) already told, the fix is complex and there is a high risk of breaking things. oj himself clearly stated to me that it sounds far to risky to put such a change into 3.2.1 release this late. Thats the main reason why I rejected it on the mailing list already. At the same time I asked oj to
(15:23:47) mdamboldt: care for a fix on the 3.3 code line in the near future. Indeed it looks like he picked up that ball instantly and continued to work on a fix (for 3.3) release. Unfortunately he provided his first findings / patch based on a 3.2.1 CWS. Of course one can assume that this means a 3.2.1 fix is easily possible. But I think it's not. And I still wouldn't like to integrated any changes related into 3.2.1 now based on current informations.
(15:25:06) MechtiIde: -1
(15:25:19) MechtiIde: esp without informations about the risks to test it
(15:25:47) rtimm: mdamboldt: sounds reasonable, especially since we do not have OJ at hand now
(15:25:58) UweL: +1
(15:26:16) MechtiIde: then I have problems to approve 3.2.1
(15:26:40) Stefan_W: I cannot judge this. The only thing I can tell, is that this is a very severe bug.
(15:31:57) Stefan_W: (many, many users must stick with 3.1.1 due to this bug) However, I am not a release expert and not a developer. So I must rely on experts (you) judgement.
(15:32:36) FrankS [~FrankS@nat/sun/x-jehfxmwjwdovzkem] hat den Raum betreten.
(15:32:41) FrankS: Hi
(15:33:06) mdamboldt: Hi Frank!
(15:33:23) kai_a: FrankS: could you provide us some more background information on OJ's patch, please?
(15:33:33) FrankS: sure ....
(15:33:46) FrankS: I have no background 'bout the patch, but ...
(15:33:49) FrankS: I know its context
(15:33:58) FrankS: that's pretty difficult code, which is prone to regressions
(15:34:03) FrankS: (as we saw in the not too distant past)
(15:34:22) FrankS: and, Ocke is the only person deeply involved in that code
(15:34:38) FrankS: While it might be that Ocke's recent patch attempt fixed the problem,
(15:34:53) FrankS: it is pretty impossible for me to judge, without tremendous effort, the potential for regressions
(15:35:18) FrankS: so, my opinion simply is: Don't do it
(15:35:28) FrankS: and I say this with big regrets, since this is a nasty ugly terrible bug
(15:35:56) ***MechtiIde then have problems to approve 3.2.1
(15:35:57) FrankS: but we simply cannot overlook, in the current situation, the possible consequences
(15:35:58) FrankS: .
(15:36:20) MechtiIde: It's a bad situation
(15:36:35) FrankS: yes
(15:36:45) FrankS: but we won't make it better by introducing more regressions
(15:36:59) FrankS: and in fact we cannot be sure not to do this
(15:37:05) MechtiIde: FrankS, can you give a hint whre we can expect problems so we can test it in the build?
(15:37:40) FrankS: that's a central component of DBA, in practice, everything related to forms / row sets would have to be tested
(15:37:52) FrankS: and *everything* is, as experience from the past shows, more than we can list ...
(15:38:43) kai_a: MechtiIde: if Frank gives a warning not to integrate it, I really wouldn't do it...
(15:39:16) MechtiIde: kai_a, I know that but I also want to understand
(15:39:30) MechtiIde: only then I'm able to communicate it
(15:39:32) kai_a: MechtiIde: even OJ seemed to be quite unsure about it
(15:40:21) jj293: MechtiIde, AFAIK 3.2 also contained this bug. So is it really wise to not approve 3.2.1, given the high risk involved in fixing it and the little time to verify?
(15:40:34) _KAMI_ hat den Raum verlassen (quit: Quit: Leaving.).
(15:40:48) MechtiIde: jj293, 3.2 contains more then this bug
(15:41:08) MechtiIde: a little bit seems to be fixed but with a bad result
(15:41:24) MechtiIde: and the patch will complete the fix
(15:41:25) jj293: MechtiIde, of course
(15:41:37) MechtiIde: this is the way the users see it
(15:41:55) MechtiIde: that#s why I want more information
(15:42:03) Stefan_W: jj293 this bug was hidden behind another bug, that has been fixed in m16. That´s why we are so short of time.
(15:42:13) FrankS: mechtilde: I am not sure what more information I could give you - except
(15:42:14) MechtiIde: to be able to tell the user why the fix doesn't work as expected
(15:42:32) FrankS: that this is a way too risky fix for my taste
(15:42:35) MechtiIde: FrankS, I understand your informations, you give here
(15:42:57) MechtiIde: so there seems to be no chance to do some more test
(15:43:07) MechtiIde: also from my side
(15:44:24) paveljanik [~Pavel@unaffiliated/paveljanik] hat den Raum betreten.
(15:46:42) mdamboldt: I think one of the important things here is to continue keeping an eye on this. Means to get the fix into a CWS and to QA that CWS as best as we can. Finally when that is complete, we can integrate it into master code line and more people will get hands on it. All this should happen soon on the 3.3 code line to make sure it gets in there without having regressions related to the fix.
(15:47:35) MechtiIde: that is a must
(15:47:54) MechtiIde: if we have to improve it with that patch
(15:48:17) MechtiIde: it should be then DEV300_m7x
(15:48:20) MechtiIde: ;-)
(15:49:07) mdamboldt: Mechtilde: Regarding the last L10N CWS for 3.2.1, well the last fixes were simply late again. The biggest amount of 3.2.1 L10N stuff is already in the master. The last CWS ooo321l10n4 contains only some few late changes. The majority is already in.
(15:50:37) MechtiIde: then let us try it mdamboldt ;-)
(15:52:44) mdamboldt: There is another item regarding 3.2.1 release. Especially in point of view to the upcoming RC and it's Windows installation sets which are usually signed.
(15:53:32) _Nesshof_: we will need this release some more time for the process of igning because the process has changed for this
(15:54:31) _Nesshof_: so I expect that we will provide the signed download set for windows with a delay compared to earlier releases
(15:54:51) dtardon hat den Raum verlassen (quit: Quit: this parrot is dead).
(15:55:10) _Nesshof_: I think I will start with OOo_wjre with a block of 10 lacalization or so
(15:55:28) _Nesshof_: this seems to be a package for a day
(15:55:58) _Nesshof_: so I would start with en_US, de, fr, ja, it, es, etc.
(15:56:30) _Nesshof_: I'm open for suggestion for a certain sequence of langs
(15:56:45) MechtiIde: pt_br
(15:57:07) MechtiIde: and this is for the Win build?
(15:57:14) _Nesshof_: yes
(15:57:36) ml1: we will have 40 langs this time
(15:57:39) MechtiIde: Can we expect the Linux/Solaris/Macbuilds in the normal time?
(15:57:49) paveljanik hat den Raum verlassen (quit: Quit: This computer has gone to sleep).
(15:57:53) _Nesshof_: the remaining OS will come within the usual timeframe
(15:58:01) ml1: + 68 langpacks
(15:58:03) MechtiIde: and the source code?
(15:58:13) _Nesshof_: also as usual
(15:58:28) MechtiIde: ok this is for the record
(16:00:32) mdamboldt: Ok, so everybody should keep these changes in mind. The RC1 bits for Windows will become available in pieces step by step.
(16:01:00) mdamboldt: Are there any other items for todays meeting?
(16:01:35) mdamboldt: For the records, Kurt Zenker (kt) will be running the OOO320m17 build.
(16:01:45) mdamboldt: kt should read kz
(16:02:16) ja_: ...and DEV300m78 will be built by by Ivo Hinkelmann (ihi)
(16:02:35) mdamboldt: Ok, sounds thats all for todays meeting.
(16:02:36) mdamboldt: bye

Personal tools