Log Mac Meeting October 8th 2008

Return to Previous meetings page : Previous_Mac_Meeting_logs

Return to October Meetings

[2008-10-08 14:59:30] <PhilippL> 1. Welcome new devs joining Mac OS X port 
[2008-10-08 14:59:35] <PhilippL> anybody new ?
[2008-10-08 15:00:00] <PhilippL> 2. Current state of the open child workspaces (CWS) 
[2008-10-08 15:00:16] <PhilippL> Currently I have no specific aqua CWS.
[2008-10-08 15:00:53] <hdu_hh> well ooo301gsl1 is currently osx only
[2008-10-08 15:01:28] <fne> i have no open CWS either
[2008-10-08 15:01:31] <PhilippL> Also if I had one the SVN migration would have put a hold on it.
[2008-10-08 15:02:51] <PhilippL> 3. Roundtable
[2008-10-08 15:03:29] <hdu_hh> OOo30 should be release really soon now...
[2008-10-08 15:04:44] <cloph> just a remark wrt svn: checkouttimes/cloning is very slow on Mac, so I recommend to keeping a tarball of any svn-master. When you need a new tree, unpack the tarball and then "svn switch" to the desired cws/master.
[2008-10-08 15:05:57] <PhilippL> true
[2008-10-08 15:06:04] <cloph> (unpacking tarball +  svn switch: clearly less than 10 MInutes here, checkout: 30 Min if I'm lucky, easily more - copy insted of tar: add another 10 minutes)
[2008-10-08 15:06:33] <PhilippL> 30 min, lucky you :-(
[2008-10-08 15:06:41] <PhilippL> say about 5 to 6 hours here.
[2008-10-08 15:06:55] * grahamperrin tunes in
[2008-10-08 15:07:23] <cloph> That was before everybody was starting to use svn... :-P -- a checkout from a local reposittory mirror took > 80 Minutes.
[2008-10-08 15:07:48] * grahamperrin alerts developers to http://qa.openoffice.org/issues/show_bug.cgi?id=94637 P2
[2008-10-08 15:07:48] <cloph> Wow 5 to 6 hours - thankfully it isn't /that/ slow here.
[2008-10-08 15:07:53] <IZBot> issue 94637: framework DEFECT UNCONFIRMED P2 an open edition of a file is not marked/recognisable as 'busy' to other users of the system; edits are lost or overwritten
[2008-10-08 15:08:19] <cloph> Did filelocking ever work? No idea...
[2008-10-08 15:08:47] <grahamperrin> cloph: potentially serious, IMHO
[2008-10-08 15:09:03] * grahamperrin looks forward to release, in any case :)
[2008-10-08 15:09:09] <cloph> grahamperrin: But /very/ late for the release..
[2008-10-08 15:09:48] <hdu_hh> mav did some changes for file locking on osx i92735
[2008-10-08 15:09:54] <grahamperrin> cloph: certainly.
[2008-10-08 15:10:48] <cloph> (and I have no idea whether it worked in the past, and Multi-User Mac Installations (multiple users working simultaneously) are rather rare, aren't they?)
[2008-10-08 15:11:54] <cloph> issue 92735
[2008-10-08 15:12:01] <IZBot> porting DEFECT VERIFIED FIXED P3 Saving to samba mounts does not work: Could not create backup copy. http://qa.openoffice.org/issues/show_bug.cgi?id=92735
[2008-10-08 15:12:17] <grahamperrin> cloph: I haven't had an opportunity to test in AFP and/or SMB/CIFS environments but my P2 concerns were:
[2008-10-08 15:13:09] <grahamperrin> 1) if the issue is true in single computer multiple user environment then we should check whether similar is true for (more prevalent) file server environments
[2008-10-08 15:13:49] <grahamperrin> and 2) defocusing from the un-likelihood of two users of a single computer using OOo to access the same file at the same time ...
[2008-10-08 15:14:10] <grahamperrin> ... the fact that a file is not marked as 'busy' (or whatever) can be problematic to other applications
[2008-10-08 15:14:15] <grahamperrin> backup applications come to mind
[2008-10-08 15:14:18] <grahamperrin> et cetera
[2008-10-08 15:15:00] <grahamperrin> yes and now (common use case) mobile home directory sync with Mac OS X Server, which sort of falls under umbrella of backup
[2008-10-08 15:15:33] <grahamperrin> hdu_hh: thanks, I'll cc mav on 94637
[2008-10-08 15:15:33] <cloph> I don't see how a backup or sync operation should fail if there is /no/ locking.
[2008-10-08 15:15:44] <hdu_hh> grahamperrin: I already did
[2008-10-08 15:16:01] <cloph> (but that doesn't mean I don't take such an issue serious)
[2008-10-08 15:16:04] <grahamperrin> hdu_hh: thanks and thanks
[2008-10-08 15:16:44] <grahamperrin> cloph: thanks for your consideration
[2008-10-08 15:16:50] <hdu_hh> grahamperrin: you"re welcome
[2008-10-08 15:17:26] <grahamperrin> cloph: backup/sync may be of file whilst in a possibly inconsistent state (yes?)
[2008-10-08 15:17:28] * grahamperrin thinks
[2008-10-08 15:18:28] <grahamperrin> users and system managers may gain false impression that a backed up/synced file has integrity, when that's not necessarily true
[2008-10-08 15:18:35] <cloph> The state is not inconsistent, it is just the file as it was before editing. And backup is read-only, it doesn't modify the file it backs up.
[2008-10-08 15:18:45] <grahamperrin> cloph: OK
[2008-10-08 15:20:27] <PhilippL> Anything left to discuss ?
[2008-10-08 15:20:51] * grahamperrin apologies for sprawling mid-meeting discussion
[2008-10-08 15:21:12] <PhilippL> ? Since you were on topic why apologise ?
[2008-10-08 15:21:26] <grahamperrin> :) not sure quite how on-topic it was/is
[2008-10-08 15:22:12] <PhilippL> Ok, if no other open points remain, I guess that's it for today ?
[2008-10-08 15:22:15] <MechtiIde> One question for the Release of 3.0 "no known stoppers" on Intel and PPC?
[2008-10-08 15:22:26] <grahamperrin> cloph et al: extend the thought: if a backup or some other operation properly marks a file that a user is editing as busy during (say) a backup, then OOo may be unable to save that file when it needs to
[2008-10-08 15:22:30] <PhilippL> And the question is :-)
[2008-10-08 15:22:49] <MechtiIde>  "no known stoppers" on Intel and PPC?
[2008-10-08 15:23:04] <cloph> I'll try to get anouther round of feedback for the languagepacks - and if there is none, I'll search for a QA to get the current state into a master, maybe then the bigger masses will complain about bad spelling and wording in dialogs :-)
[2008-10-08 15:23:08] <PhilippL> I know of none, but I probably wouldn't about PPC.
[2008-10-08 15:23:22] * grahamperrin knows of no show-stoppers on Intel
[2008-10-08 15:23:39] <cloph> MechtiIde: http://qa.openoffice.org/issues/show_bug.cgi?id=94637 - but not declared as stopper (yet?)
[2008-10-08 15:23:44] <IZBot> issue 94637: framework DEFECT UNCONFIRMED P2 an open edition of a file is not marked/recognisable as 'busy' to other users of the system; edits are lost or overwritten
[2008-10-08 15:24:31] <grahamperrin> I shouldn't class my 94637 but maybe be prepared for side-efefcts to be difficult (for users more than us) to diagnose
[2008-10-08 15:24:44] <grahamperrin> * shouldn't class my 94637 as a stopper
[2008-10-08 15:24:46] <cloph> (I personally don't think it is a stopper - unless locking wouldn't work for remote volumes with properly setup-locks)
[2008-10-08 15:24:51] <MechtiIde> no one confirmed it
[2008-10-08 15:25:17] * grahamperrin does suspect that AFP will handle busy-ness with grace 
[2008-10-08 15:27:39] <MechtiIde> grahamperrin, I think an unconfirmed Issue can't be a stopper
[2008-10-08 15:28:02] <grahamperrin> MechtiIde: I agree, not a stopper; everyone: please go ahead with the show
[2008-10-08 15:28:49] <PhilippL> Anything else to discuss ?