fesco
LOGS
18:01:59 <kalev> #startmeeting FESCO (2014-11-05)
18:01:59 <kalev> #meetingname fesco
18:01:59 <kalev> #chair dgilmore jwb kalev mattdm mitr mmaslano nirik sgallagh stickster t8m thozza
18:01:59 <zodbot> Meeting started Wed Nov  5 18:01:59 2014 UTC.  The chair is kalev. Information about MeetBot at http://wiki.debian.org/MeetBot.
18:01:59 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
18:01:59 <zodbot> The meeting name has been set to 'fesco'
18:01:59 <zodbot> Current chairs: dgilmore jwb kalev mattdm mitr mmaslano nirik sgallagh stickster t8m thozza
18:02:01 <kalev> #topic init process
18:02:09 <nirik> .hello kevin
18:02:10 <zodbot> nirik: kevin 'Kevin Fenzi' <kevin@scrye.com>
18:02:13 <sgallagh> .hello sgallagh
18:02:14 <zodbot> sgallagh: sgallagh 'Stephen Gallagher' <sgallagh@redhat.com>
18:02:16 <jwb> hi
18:02:16 <thozza> hello all
18:02:36 <kalev> let's see ... my first time chairing this :)
18:02:38 <t8m> hi
18:03:19 <kalev> looks like we have quorum at least
18:03:28 * mattdm is here
18:03:29 <sgallagh> kalev: Yup, you don't get out of it that easily :)
18:03:29 * stickster notes he will log but has to be afk today
18:03:43 <kalev> sgallagh: pff :)
18:04:15 <kalev> OK, let's get started
18:04:19 <kalev> we have two topics today
18:04:25 <kalev> #topic #1362 Clarify feature process as it relates to FPC
18:04:25 <kalev> .fesco 1362
18:04:26 <zodbot> kalev: #1362 (Clarify feature process as it relates to FPC) – FESCo - https://fedorahosted.org/fesco/ticket/1362
18:05:18 <kalev> tibbs: do want to give a short summary of what you'd like changed in the feature process and what else fesco can do to help make this better for FPC?
18:06:34 <sgallagh> I think the original proposal makes a reasonable request that any Change request that involves a packaging guideline modification must provide at least a first draft of that modification itself.
18:07:05 <sgallagh> Otherwise, we have Changes as a sort of backdoor into having FESCo demand that the FPC do something, which we really don't have control over.
18:07:38 <nirik> right, as well as people proposing the change having the best idea of what the change means/how it works.
18:07:49 <kalev> is it something that should be codified somewhere or just something to keep in mind when accepting changes?
18:08:28 <thozza> kalev: I think it should be part of the proposal
18:08:44 <nirik> well, it could be codified into the changes process?
18:09:04 <sgallagh> Yeah, I'd like to see it stated explicitly in the Changes Process
18:09:41 <sgallagh> (Which I realize means "having FESCo demand that <someone> do something", but at least <someone> in this case is lkely to be a FESCo member)
18:10:58 <t8m> sgallagh, +1
18:11:06 <nirik> +1
18:11:41 * sgallagh realizes he probably just volunteered
18:11:45 <kalev> sure, +1 -- does anyone want to write something up?
18:11:53 <kalev> aha, I can see a volunteer :)
18:12:07 <mattdm> On "demanding that FPC do something" .... I don't think that's the right characterization
18:12:38 <sgallagh> mattdm: Well, it kind of is. It's asserting that certain changes have to go in under specific time-constraints
18:12:39 <mattdm> There's a perfectly reasonable split where FESCo can say "we want this to be done; please figure out / approve guidelines for how it should be done"
18:12:53 <mattdm> sgallagh ... and sometimes they do!
18:13:22 <nirik> well, where there are cases where a change involves things noone on FPC is a expert in, we really need the change owner(s) to draft things...
18:13:29 <sgallagh> Sure, I have no problems with us approving it ahead of time
18:13:43 <sgallagh> But I think the onus MUST be on the Change proposers to provide the first draft
18:14:17 <mattdm> I'm in favor of that, sure.
18:14:20 <t8m> Do we have a concrete proposal to vote on? I think the original proposal was right at least in this regard.
18:15:17 <t8m> Sometimes the Change proposer does not realize that FPC guideline should be added though.
18:15:51 <t8m> The question is whether we should always demand him to propose such draft or not in this case.
18:16:42 <kalev> most of the time Changes owners are people who are driving changes
18:16:52 <kalev> I think it's reasonable to expect them to draft policy changes as well, when applicable
18:17:24 <thozza> Proposal: If the Change requires a Packaging Guidelines change, the owner of the change have to provide a draft of the change. This request has to be codified in the Change Process.
18:17:25 <nirik> sure, sometimes they won't need packaging changes, but it should be mostly obvious...
18:17:41 <nirik> thozza: +1
18:17:46 <sgallagh> thozza: +1
18:17:56 <kalev> thozza: +1
18:18:07 <thozza> I'm also +1 for the record :)
18:18:34 <t8m> OK, +1 with the reservation above, but I think that can be handled ad hoc
18:18:54 <thozza> t8m: sure
18:19:40 <kalev> looks like we have consensus and +5 votes
18:19:51 <kalev> mattdm: any objections?
18:20:02 <mattdm> kalev: sorry +1
18:20:22 <tomspur> So FPC should wait until the feature is approved or can it already approve the guideline in advance?
18:21:06 <nirik> I don't see a need to tie the two... if the change owner drafts something and FPC approves it, should be fine?
18:21:18 <sgallagh> tomspur: Yeah, the guideline can be approved beforehand
18:21:26 <thozza> I agree
18:21:28 <sgallagh> For non-controversial things, at least
18:21:53 <sgallagh> Obviously the FPC can opt to say "We want FESCo to weigh in first" on things
18:21:54 <thozza> It should not be used as argument for FESCo to approve such a change
18:22:02 <sgallagh> right
18:22:12 * nirik trusts FPC folks to do that as needed.
18:22:21 <kalev> #agreed If the Change requires a Packaging Guidelines change, the owner of the change have to provide a draft of the change. This request has to be codified in the Change Process. (+6,0,0)
18:22:32 <thozza> who will do it? :)
18:22:36 <kalev> sgallagh: were you saying you'd draft it?
18:22:47 <sgallagh> #action sgallagh to draft up an update to the Change Process
18:22:57 <kalev> great, thanks sgallagh
18:23:03 <thozza> sgallagh: thanks!
18:23:09 <kalev> ok, on to the next topic
18:23:13 <kalev> #topic #1365 A unique system-wide TMP directory for all programs and sane ways to retrieve the default
18:23:16 <kalev> .fesco 1365
18:23:18 <zodbot> kalev: #1365 (A unique system-wide TMP directory for all programs and sane ways to retrieve the default) – FESCo - https://fedorahosted.org/fesco/ticket/1365
18:23:33 <nirik> looks like there were comments added to this in just the last few minutes...
18:23:48 * nirik wonders if this is something we should punt on and see if we can reach some solution in ticket.
18:23:51 <kalev> not sure FESCo should do anything at this point here, quite a bit of discussion still going on
18:24:01 <t8m> sgallagh, I think you can amend the Change process outright
18:24:03 <kalev> people are just using the ticket for bringing different parties together to discuss the problem
18:24:24 <t8m> kalev, +1
18:24:26 <sgallagh> t8m: Sure; I suppose if the phrasing isn't perfect, we can iterate
18:24:38 <kalev> mschwendt: anything you'd want fesco to do at this point?
18:24:52 <mschwendt> kalev: I dunno what fesco thinks about it
18:24:53 <t8m> sgallagh, yeah
18:25:11 <mschwendt> is it considered bad that firefox overrides $TMPDIR if not set?
18:25:21 <mschwendt> is it considered bad that gnome-terminal discards env vars?
18:25:47 <mschwendt> do we want to rely on being able to customize installations via $TMPDIR env?
18:25:48 <t8m> mschwendt, +1 to both at least from my point of view
18:25:54 <kalev> I would personally say that Firefox shouldn't override TMPDIR in the environment it passes to launched apps, yes
18:26:37 <t8m> mschwendt, you added third question before I answered :) For the third question - I am not decided
18:26:47 <mschwendt> ;)
18:27:06 <nirik> I think it would be good for us to be consistent at the very least... the mismash now isn't good (IMHO)
18:27:42 <mclasen> mschwendt: gnome-terminal does not discard env vars
18:28:06 <mclasen> repeating it doesn't make it any truer
18:28:42 <mschwendt> mclasen: does it make a difference whether they get deleted or not set beforehand?
18:29:33 <mschwendt> mclasen: have you seen that gnome-terminal's mailto handler breaks XDG base dirs, too?
18:30:13 <mclasen> it does no such thing
18:30:27 <mschwendt> care to explain?
18:31:18 <t8m> mclasen, if you set some env variable, run gnome-terminal with this var set, the handler run through the dbus-activation won't see it
18:31:40 <mclasen> true
18:31:56 <mclasen> but blaming gnome-terminal for the mechanics of dbus activation is not useful
18:32:23 <t8m> yes, s/gnome-terminal/dbus-activation in gnome sessions/ should be done above
18:33:07 <t8m> there should be a way to pass environment from the process invoking the dbus-activation to the dbus-activated process
18:33:17 <mclasen> if you want tmpdir set for all processes in the session (including those launched via dbus activation), you need to make sure that tmpdir makes it into the dbus activation environment
18:33:29 <otaylor> t8m: activated processes are often shared by multiple activators
18:33:31 <mclasen> either by setting it before launching dbus-daemon, or by injecting it afterwards
18:35:02 <nirik> so, really to me this sounds like we need to hash out some solution if we want a system wide tmpdir in ticket...
18:35:03 <mschwendt> now it's getting somewhere. consistency would be essential here as not to start programs in changing environments that are different from the desktop.
18:35:08 * nirik doesn't think we are going to solve it here today.
18:35:18 * mschwendt neither
18:35:20 <sgallagh> Frankly, I've long wished for the ability to set an env var in one terminal and have it set in all my other terminals too
18:35:30 <sgallagh> Could we perhaps handle all env vars in the session bus?
18:36:00 <t8m> sgallagh, +1
18:36:52 <kalev> anyway, I don't think we can reengineer everything here today
18:36:59 * sgallagh agrees
18:36:59 <kalev> proposal: Give the discussion in the ticket more time and see next week if there's anything FESCo can do
18:37:21 <mschwendt> kalev: back to the initial question about the fesco ticket -> confirming/acknowledging the problem will be a first step. Firefox has been fixed, tells the ticket.
18:37:36 <kalev> ah cool
18:37:37 <mschwendt> that will help Claws Mail. By default.
18:37:51 <nirik> mschwendt: I think it is an issue and we should try and make things distro wide better. :) thanks for bringing it up
18:38:06 <t8m> kalev, +1
18:38:21 <mschwendt> I will continue talking to claws mail upstream and try to find out why they don't choose a different location for the file.
18:38:34 <kalev> sounds good, thanks mschwendt for working on this
18:40:08 <kalev> #info Give the TMP directory discussion more time and revisit it next week to see if there's anything FESCo can do
18:40:14 <kalev> anything else here?
18:40:28 <kalev> #topic Next week's chair
18:40:46 * kalev rubs hands. Someone else's turn next week :)
18:41:00 * sgallagh will be at a conference and unable to attend
18:41:08 <t8m> kalev, I would like to get the chair week after the next one, I am not sure I will be able to attend next week
18:41:16 <nirik> I can do next week
18:41:16 <kalev> sure
18:41:23 <kalev> awesome, thanks nirik
18:41:32 <kalev> #action nirik to chair next week
18:41:47 <kalev> #info t8m and sgallagh unable to attend next week's meeting
18:41:51 <kalev> #topic Open Floor
18:42:03 <kalev> anything anyone wants to bring up for the open floor?
18:42:34 <nirik> Thanks to everyones hard work on Beta.
18:42:42 <nirik> Seems to be getting pretty good reviews in the press.
18:42:42 <t8m> nirik, +1 to that!
18:42:54 <kalev> #info Thanks to everyone's hard work on Beta.
18:43:00 <sgallagh> nirik: Got any links besides Phoronix and InfoWorld?
18:43:44 <mattdm> http://www.pcworld.com/article/2842555/how-fedora-21-is-splitting-itself-inself-in-thirds-to-prepare-for-the-next-10-years.html
18:43:58 <mattdm> http://betanews.com/2014/11/04/linux-fans-fedora-21-beta-1-is-here-now-featuring-three-flavors/
18:44:14 <sgallagh> Nice
18:44:16 <nirik> yeah, several more...
18:44:22 <mattdm> http://www.lifehacker.co.uk/2014/11/05/fedora-21-beta-released-need-know-new-multi-flavour-linux-distro
18:45:07 <sgallagh> Cool, yeah none of those were on my watch-list
18:45:10 <sgallagh> Thanks!
18:45:40 <kalev> I have a huge grin on my face here after quickly skimming the reviews
18:45:44 <kalev> nice to see positive reviews :)
18:45:51 <sgallagh> Indeed
18:46:08 <kalev> oh, actually I had something I wanted to bring up for the open floor
18:46:36 <kalev> would anyone here mind if I try to get the packages that have broken deps in F21 dropped from the package collection?
18:46:49 <kalev> it seems rather bad user experience to ship something that cannot even be installed
18:47:09 <nirik> kalev: dropped or fixed before freeze. sure. ;) And also rememeber to nuke or fix rawhide versions at the same time
18:47:11 <kalev> I replied to a F21 Branched Report two weeks ago and generated some interest in broken deps and quite a few got fixed
18:47:17 <kalev> nirik: yep
18:47:46 <kalev> OK, I'll try to push this forward
18:47:56 <nirik> I think a few releases ago spot went to great effort to make the broken deps 0... but of course it didn't last. ;(
18:48:20 <t8m> kalev, nirik, in some cases if the packages are not in comps or deps of comps I don't think it would be a problem to have the deps fixed after release in updates
18:48:25 <kalev> yes -- I wonder if it could be a policy to just drop things before the final freeze that have broken deps
18:48:35 * spot rocks back and forth in his warm straitjacket.
18:48:40 <kalev> t8m: I agree -- but that doesn't mean we have to ship broken things in the base repo
18:48:41 <nirik> t8m: sure, but then they also could be re-added in updates. ;)
18:48:46 <kalev> t8m: can always add the package back once it's fixed
18:48:57 <nirik> spot: hey, at least it's warm! :)
18:49:03 <spot> :D
18:49:20 <t8m> so would the drop mean "drop for the release being released only"? I am +1 to that
18:49:53 <mschwendt> interesting topic ;-)   mind you, there still is no garbage collector meta package in the distribution
18:50:08 * nirik nods.
18:50:10 <kalev> t8m: right; anything can always be added back in the updates repo afterwards, even if it's dropped in the base "fedora" repo
18:50:41 <kalev> the distinction is that the "fedora" repo is frozen and cannot be fixed after we've shipped the F21 gold release
18:51:10 <t8m> kalev, OK
18:51:27 <sgallagh> Quick question about the Change Process I'm editing
18:52:01 <nirik> kalev: well, we will no longer have a "fedora" repo, it will be "everything" but same thing applies.
18:52:01 <sgallagh> I'd like to set Alpha Freeze as deadline for the first draft to the FPC. Is that reasonable?
18:53:27 <sgallagh> (or else the "Contingency Plan" goes into effect)
18:53:40 <nirik> sure. it shouldn't be too hard to have a first draft then
18:54:06 <t8m> exceptions can be always approved if reasonable
18:54:10 <sgallagh> right
18:54:27 <sgallagh> My thought is that Alpha Freeze is the time where we expect the Change to be at least testable anyway
18:54:39 <t8m> yes
18:54:43 <sgallagh> So presumably they are already implementing the new guidelines or well on their way
18:55:03 <kalev> I wouldn't be overly concerned when the deadline is, as long as there's a note there saying that the proposal owners should come up with a draft
18:56:19 <kalev> I'll close out the meeting in a minute if nothing else comes up
18:56:35 <sgallagh> https://fedoraproject.org/w/index.php?title=Changes%2FPolicy&diff=393926&oldid=390130
18:56:53 <sgallagh> jreznik: Would you mind making an update about this so that people are aware for F22 Changes?
18:57:47 <kalev> he just timed out
18:58:01 <kalev> sgallagh: looks good to me
18:58:22 <t8m> sgallagh, OK
18:59:37 <kalev> OK, thanks for coming everybody
18:59:46 <sgallagh> kalev: Thanks for chairing
18:59:48 <kalev> and thanks for bearing with me learning how to chair a fesco meeting
18:59:52 <kalev> #endmeeting