14:00:34 <licquia> #startmeeting LSB Bug Triage 2014 Jun 27
14:00:34 <lsbbot> Meeting started Fri Jun 27 14:00:34 2014 UTC.  The chair is licquia. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:00:34 <lsbbot> Useful Commands: #action #agreed #help #info #idea #link #topic.
14:00:48 <licquia> good morning everyone
14:01:06 <mwichmann> hi
14:02:24 <licquia> #info lsb: 1 reopened, 4 pleasetest, 1 needinfo, 6 new
14:02:35 <licquia> #info fhs: 14 new
14:02:43 <mwichmann> speaking of fhs...
14:02:49 <licquia> ?
14:02:56 <mwichmann> think we're going to get it out w/5.0?
14:03:06 * licquia can't imagine why not
14:03:26 <mwichmann> do we have a proper location for it then? regular refspecs?
14:03:41 <mwichmann> !lsbbug 3951
14:03:44 <lsbbot> mwichmann: 04Bug http://lsbbugs.linuxfoundation.org/show_bug.cgi?id=3951 normal, P1, 5.0, mats, ASSIGNED , uplift FHS reference to 3.0
14:04:11 <mwichmann> I just pasted the sql to update, but it needs the permanent url for the entry
14:04:19 <licquia> i would imagine; think we have a mirror of it there already
14:04:32 * mwichmann checks
14:04:32 <licquia> i.e. the 2.3 fhs is already there
14:05:30 <mwichmann> it is, I guess we've had this in place for a while
14:05:33 <mwichmann> so go ahead with this?
14:05:39 <licquia> i'd say so
14:06:03 <licquia> i think you can assume that the ref will look like the 2.3 ref does now, with s/2.3/3.0/
14:06:34 <licquia> (that is, assuming you need a version-specific ref for that)
14:06:46 <mwichmann> no
14:06:57 <licquia> if you can just point to refspecs/fhs.shtml, all the better
14:07:05 <mwichmann> we still have the irritating exposure of shtml
14:07:13 <mwichmann> I'd /rather/ it be refspecs/fhs
14:07:23 <mwichmann> but that's just more busy-work we can put off
14:07:47 <licquia> that shouldn't be hard, actually
14:07:56 <licquia> if you want that to be the reference, should be easy
14:08:03 <licquia> add a note to the bug
14:08:17 <mwichmann> I think it would be better
14:08:22 <licquia> (or file another, whatever)
14:09:12 <mwichmann> anyway, okay to jump back to the list; guess it's just the two of us (recall Russ saying something about travel - that NC trip)
14:09:38 <licquia> #action licquia add refspecs.lf.org/fhs and refspecs.lf.org as better refs to FHS and LSB specs
14:10:00 <licquia> yup; did poke in main channel in case jbj wanted to join us
14:10:20 <licquia> shall we do the usual, or do 5.0 stuff first?
14:10:41 <mwichmann> I flipped one bug to NEW (3465)
14:11:01 <mwichmann> it said we needed a triage decision, but never happened
14:11:13 <licquia> k
14:11:18 <licquia> let's do the usual first
14:11:21 <mwichmann> sure
14:11:34 <licquia> !lsbbug 3033
14:11:36 <lsbbot> licquia: 04Bug http://lsbbugs.linuxfoundation.org/show_bug.cgi?id=3033 normal, P2, 4.1-updates, licquia, REOPENED , T.c_iflag and T.tcgetattr fail when using pseudo ttys (on some systems)
14:12:14 <licquia> this is the deal where there's a patch, but it didn't seem to get applied upstream
14:12:27 * licquia didn't send the query upstream
14:13:52 <licquia> #action licquia send 3033 query to glibc upstream
14:14:04 <licquia> !lsbbug 3195
14:14:06 <lsbbot> licquia: 04Bug http://lsbbugs.linuxfoundation.org/show_bug.cgi?id=3195 normal, P2, ---, licquia, PLEASETEST , LSB packages with unowned directories uninstallable on some rpm implementations.
14:14:25 <mwichmann> I guess it worked for lsb_tester
14:14:25 <licquia> that actually has progress; the fixed qmtest is in for 4.1
14:14:41 <mwichmann> enough to close?  or does this involve sweeping up more?
14:14:47 <licquia> cool; saw some activity in the other channel on that, wasn't sure what the result was
14:14:58 <mwichmann> guess one could ask explicity
14:16:19 <licquia> k
14:16:42 <licquia> #action licquia ask lsb_tester about 3195 fix
14:17:02 <licquia> !lsbbug 3465
14:17:04 <lsbbot> licquia: 04Bug http://lsbbugs.linuxfoundation.org/show_bug.cgi?id=3465 normal, P2, 5.0, licquia, NEW , missing verr/vwarn functions
14:17:24 <licquia> this is the bug you flipped to NEW manually?
14:17:29 <mwichmann> yes
14:18:11 * licquia reads what the decision is to make
14:18:22 <mwichmann> it's a split thing
14:18:55 <mwichmann> "set completeness" (this bug) divided from the earlier bug which also pointed out an error
14:19:05 <licquia> right
14:19:12 <licquia> so the decision is "in or out"?
14:20:05 <mwichmann> "complete the set" vs. "drop (deprecate) the existing partial set"
14:20:08 <mwichmann> I think
14:20:10 <licquia> ah, i see; the 'v' versions use an explicit va_args
14:20:11 <mwichmann> too long ago now
14:20:52 <licquia> while the non-v versions use variable args the "syntax" way
14:21:25 <licquia> and we only have the one
14:21:31 <licquia> (the one 'v' version)
14:21:41 <mwichmann> yes
14:21:42 <licquia> it is puzzling, but i'm inclined to postpone
14:21:50 <mwichmann> okay with that
14:22:00 <mwichmann> might have answered differently two years ago
14:22:09 <licquia> not sure it's the kind of thing that should hold up the release
14:22:26 * licquia sees it's p2, so no need to do anything
14:23:07 <mwichmann> I'd as soon be explicit and take if off 5.0 milestone
14:24:13 <licquia> k
14:24:47 <mwichmann> I'll mark it up later, move on
14:25:17 <licquia> !lsbbug 3953
14:25:19 <lsbbot> licquia: 04Bug http://lsbbugs.linuxfoundation.org/show_bug.cgi?id=3953 normal, P2, ---, licquia, PLEASETEST , add link checker job to puppet to run periodically
14:25:47 <mwichmann> is this stuff actually running now?
14:26:00 <licquia> not yet; it needs to be tested, then added to puppet
14:26:08 <licquia> (unless orc_fedo is running it somewhere)
14:26:36 <licquia> this is also wrapped up with bug 3954
14:26:39 <mwichmann> so this pleastest is not "everything done, just needs to be verified", but "some intermediate step needs testing before proceeding"
14:26:40 <licquia> !lsbbug 3954
14:26:42 <lsbbot> licquia: 04Bug http://lsbbugs.linuxfoundation.org/show_bug.cgi?id=3954 normal, P2, ---, licquia, PLEASETEST , link checker script should build cache ... more
14:27:19 <licquia> well, it needs to be deployed after testing; not sure that's an "intermediate step"
14:28:55 <licquia> #action licquia test patch in 3954 (also covers 3953)
14:29:28 <licquia> !lsbbug 3982
14:29:31 <lsbbot> licquia: 04Bug http://lsbbugs.linuxfoundation.org/show_bug.cgi?id=3982 major, P2, ---, licquia, NEEDINFO , Unable to link Qt5WebKit using LSB 4.1 (lsbc++)
14:30:28 <licquia> still waiting on the list of missing freetype symbols
14:30:43 <licquia> should we do anything here?
14:32:03 <mwichmann> of coruse we should plan to quickly uplift freetype :)
14:32:15 <licquia> well, there's that, of course
14:32:29 <licquia> but should we just wait here, or ping, or...?
14:32:55 <mwichmann> I guess a ping and otherwise not act
14:33:00 <licquia> k
14:33:09 <mwichmann> this doesn't block anything
14:33:22 <licquia> #action ping re: list of freetype items on 3982
14:33:30 <licquia> !lsbbug 3991
14:33:32 <lsbbot> licquia: 04Bug http://lsbbugs.linuxfoundation.org/show_bug.cgi?id=3991 normal, P2, ---, herrold, NEW , Please grant: herrold RW access on the site
14:33:49 <licquia> this is the lanana site
14:34:02 <licquia> major obstruction: setting up auto-pull on lanana
14:34:22 <mwichmann> obstruction?
14:34:23 <licquia> orc_fedo could commit, but it wouldn't do anything until i went in and pulled
14:34:48 * licquia wonders if that's a bug or a feature :-)
14:34:49 <mwichmann> "just work", or actual blockage?
14:34:58 <licquia> just work, pretty much
14:35:10 <licquia> post-commit on git, etc.
14:35:41 <licquia> #action licquia complete 3991 (grant orc_fedo access)
14:35:53 <licquia> #action licquia for 3991, also enable auto-pull on lanana site
14:36:39 <licquia> !lsbbug 3996
14:36:41 <lsbbot> licquia: 04Bug http://lsbbugs.linuxfoundation.org/show_bug.cgi?id=3996 normal, P2, ---, herrold, PLEASETEST , woefully stale page
14:36:51 <mwichmann> we have lsb/fhs snapshots going periodically, rather than on commit trigger
14:36:57 <mwichmann> easier to implement?
14:37:08 <licquia> might be
14:37:18 <licquia> cron job might be the quickest way
14:38:32 <licquia> so, old download page now contains just a link to the new download page
14:38:45 <licquia> want to verify that, and then we can close?
14:38:50 <mwichmann> which?
14:39:03 <licquia> 3996, there's a link to the stale page in the bug
14:39:15 <mwichmann> okay, that works
14:39:25 <licquia> #agreed close 3996
14:39:38 <licquia> !lsbbug 3997
14:39:40 <lsbbot> licquia: 04Bug http://lsbbugs.linuxfoundation.org/show_bug.cgi?id=3997 normal, P2, ---, licquia, NEW , LSB WG not enumerated at indicated URL
14:40:26 <licquia> that page probably just needs to get chopped, and point to the lf workgroups page
14:40:27 <mwichmann> hah, we've stopped existing even here
14:40:41 <mwichmann> yes, sounds like best solution
14:41:06 <licquia> #action licquia point wiki workgroup page at actual LF workgroup page for 3997
14:41:53 <licquia> !lsbbug 3998
14:41:55 <lsbbot> licquia: 04Bug http://lsbbugs.linuxfoundation.org/show_bug.cgi?id=3998 normal, P2, ---, licquia, NEW , implicit dependencies in dist checker cannot be over-ridden
14:42:04 <mwichmann> that page isn't especially up to date either, fwiw
14:42:22 <licquia> yeah, there's a difference now between workgroups and collab projects and so on
14:42:22 <mwichmann> on 3998, I'm not sure the statement is true,
14:42:53 <mwichmann> but it was giving us headaches with lsb_tester, as it kept trying to downgrade lsb-qm
14:43:06 <licquia> is there a way to override the default for non-tests?  (other than by hacking the manifest)
14:43:44 <mwichmann> I think if the bits are already locally installed, setting the actual test to local means the deps stay local too
14:43:54 <licquia> ok, worth noting
14:43:59 <mwichmann> I'm not sure, it just looks that way from the dist-checker display
14:44:17 <licquia> add this to the next vip rollup?
14:44:54 <mwichmann> I guess we should, can kick out later if, say, denis pops up and says "that's wrong, you can do foo, and it's documented in the wiki"
14:46:40 <licquia> #agreed attach 3998 to next vip release
14:46:57 <licquia> !lsbbug 4000
14:46:59 <lsbbot> licquia: 04Bug http://lsbbugs.linuxfoundation.org/show_bug.cgi?id=4000 normal, P2, ---, mats, NEW , LSB defines cups types that are going private
14:47:31 <mwichmann> the impact here is the devchk struggle - "native build"
14:47:55 <licquia> so these types have gone opaque, basically?
14:47:56 <mwichmann> only raise the question since these types seem to be unused by the rest of our cups setup
14:48:12 <mwichmann> private
14:48:31 <licquia> hmm, so we don't even have people using the types
14:48:52 <licquia> even in a "struct foo; foo *allocate_foo();" kind of way
14:48:59 <mwichmann> and renamed, to make it even easier, "public" ipp_request_u -> "private" (only conditional include) _ipp_request_u
14:49:17 <mwichmann> I don't see any yses
14:49:19 <mwichmann> uses
14:49:28 <mwichmann> but that doesn't mean someone doesn't find a way
14:49:32 <licquia> might be worth researching
14:49:50 <licquia> if it's truly not used, and we can find that in old docs, then we could just issue errata
14:50:07 <licquia> otherwise, should probably do something more interesting, like deprecate/remove
14:50:26 <mwichmann> we can't deprecate types, that's come up before
14:52:06 <mwichmann> thought there was an rfe on that, but I don't see it (there are other type questions, like whether to have a TypeStd)
14:52:22 * mwichmann likes LSB "Type" stuff much less after gtk3 experience
14:52:25 <mwichmann> :)
14:53:02 <licquia> well, by "deprecate" i mean "announce deprecation, drop at next major"
14:53:24 <mwichmann> yes, but where do you "announce"?
14:53:30 <licquia> yeah, not sure
14:53:40 <licquia> anyway, do you think this needs to happen for 5.0?
14:53:40 <mwichmann> it can go in a release note, but there's no sane way for people to find at random time
14:53:47 <licquia> true
14:53:52 <licquia> no compiler warning, for example
14:53:54 <mwichmann> maybe something we can eventually add to navigator
14:53:58 <licquia> perhaps
14:54:15 <mwichmann> if someone can solve devchk problem, we don't need to do anything, no
14:54:31 <mwichmann> it was just a "hey, wait, we know these are going away anyway"
14:55:14 <mwichmann> we can change the summary a bit and make it relative to a cups uplift (not that that's happening, but is a way to keep it documented)
14:55:45 <licquia> k
14:56:00 <licquia> i vote we make it "future", then
14:56:08 <mwichmann> sure, let's do that
14:56:28 <mwichmann> I wonder if we could find a volunteer to hack on devchk issues
14:56:37 <licquia> !lsbbug 4001
14:56:38 <mwichmann> dunno where such a volunteer would come from, but...
14:56:40 <lsbbot> licquia: 04Bug http://lsbbugs.linuxfoundation.org/show_bug.cgi?id=4001 normal, P2, ---, licquia, NEW , Failure in /olver/other/other/tests/other io_term_rw_case_d_scenario
14:57:14 <licquia> looks like there's a reported fix
14:57:48 * licquia asks
14:58:14 <mwichmann> one could hope
14:58:19 * licquia asks
14:58:45 <licquia> and marks needinfo
14:58:57 <licquia> #agreed mark 4001 needinfo; ask for snapshot olver-core results
14:59:06 <licquia> ok, that's that
14:59:30 <licquia> quickly, on 5.0 bugs, i went through and bumped the priority on a few bugs
14:59:39 <mwichmann> yeah, and just on time
14:59:41 <mwichmann> isn't this fun?
14:59:47 <licquia> heh
15:00:03 <mwichmann> denis took 4002 right away, but may have to make sure he understands what I'm complaining about
15:00:07 <licquia> basically, all non-p1 bugs dependent on p1 bugs that are part of the 5.0 rollup
15:00:23 <licquia> so, the views are starting to reconcile
15:00:55 * licquia counts 16 bugs marked p1 for 5.0 milestone, which includes the rollups
15:01:22 <licquia> the actual bug lists are different, but i think we're getting closer
15:01:28 <mwichmann> ok
15:02:01 <licquia> so, maybe next week we'll have several lists which all agree, and we can start moving other bugs to "future"
15:02:10 <mwichmann> can I just close 3706 (not P1)?
15:02:16 <licquia> (assume we all agree with the reconciled lists)
15:02:20 * licquia looks
15:03:24 <licquia> if it's not p1, and it's just a matter of a name, i'm ok with it
15:03:40 * licquia assumes someone would have bumped the priority if it was important to them
15:03:44 <mwichmann> also separate issue, this was related to 3512 which is P1, I peeled off lsb_release out of si, do you want me to push that up as a new branch?
15:03:53 <licquia> sure, let's do that
15:04:02 <mwichmann> hey, on 3706, it was only me caring, I think
15:04:06 <licquia> ping me if you need help with getting that working
15:04:17 <mwichmann> unfortunately, it works all too well
15:04:28 <licquia> mwichmann: in that case, since you are the one who cares, be my guest :-)
15:04:33 <mwichmann> I accidentally push branches up now and then due to typos, and it works just fine
15:04:52 <licquia> at least it's only an issue for people with commit access
15:05:02 <mwichmann> branch to be called lsb_release, or something slightly more general with lsb_release as a subdir?
15:05:13 <licquia> that's fine
15:05:19 * licquia waves to n3npq
15:05:26 <mwichmann> like if we wanted to maintain initd stuff sample also
15:05:41 <mwichmann> "that's fine" doesn't choose between the two :)
15:05:52 <licquia> you could call it "samples" or something like that
15:05:59 <mwichmann> lsb_samples?
15:06:08 <licquia> was just about to suggest that
15:06:19 <mwichmann> or lsb- we seem to have used that form more
15:06:25 <licquia> yup
15:06:28 <licquia> works for me
15:06:58 <licquia> alright, anything else before we close up?
15:07:02 <mwichmann> okay, I'll push that now, it is not "compelte" or anything, but contains the code and history without needing the rest of si
15:07:09 <licquia> excellent
15:07:23 <mwichmann> nothing from me
15:07:26 <mwichmann> n3npq ?
15:07:32 <mwichmann> "the bug bomber" :)
15:07:35 <licquia> heh
15:08:00 <n3npq> sorry about the noise, can't be helped
15:08:23 <mwichmann> no, I get it, just a little gently joshing
15:09:05 <licquia> we asked for it, we got it, for sure :-)
15:09:16 <licquia> alright, wrapping up
15:09:19 <licquia> #endmeeting