13:01:54 <stickster> #startmeeting Workstation WG
13:01:54 <zodbot> Meeting started Mon May 22 13:01:54 2017 UTC.  The chair is stickster. Information about MeetBot at http://wiki.debian.org/MeetBot.
13:01:54 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
13:01:54 <zodbot> The meeting name has been set to 'workstation_wg'
13:01:57 <stickster> #meetingname workstation
13:01:57 <zodbot> The meeting name has been set to 'workstation'
13:02:01 <stickster> #topic Roll call
13:02:03 <stickster> .hello pfrields
13:02:03 <mcatanzaro_> .hello catanzaro
13:02:04 <zodbot> stickster: pfrields 'Paul W. Frields' <stickster@gmail.com>
13:02:07 <zodbot> mcatanzaro_: catanzaro 'Michael Catanzaro' <mcatanzaro@gnome.org>
13:02:17 <ryanlerch> .hello ryanlerch
13:02:18 <zodbot> ryanlerch: ryanlerch 'Ryan Lerch' <rlerch@redhat.com>
13:02:23 <mclasen> .hello mclasen
13:02:24 <zodbot> mclasen: mclasen 'Matthias Clasen' <mclasen@redhat.com>
13:02:47 <juhp_> .hello petersen
13:02:48 <zodbot> juhp_: petersen 'Jens Petersen' <petersen@redhat.com>
13:02:49 <stickster> #info stickster is coming in cold after three weeks away, so pardon the mess
13:03:22 <mcatanzaro_> stickster: If you could sneak https://pagure.io/fedora-workstation/issue/15 into the agenda for today, that would be super.
13:03:26 <mcatanzaro_> And welcome back!
13:04:40 <stickster> mcatanzaro_: Sure
13:04:51 <stickster> #chair mcatanzaro_ ryanlerch mclasen juhp_
13:04:51 <zodbot> Current chairs: juhp_ mcatanzaro_ mclasen ryanlerch stickster
13:05:08 <stickster> I pinged cschaller, anyone seen otaylor?
13:05:21 * mclasen summons cschaller
13:05:24 <stickster> rdieter looks to be out
13:05:28 * cschalle hi
13:05:32 <stickster> #chair cschalle
13:05:32 <zodbot> Current chairs: cschalle juhp_ mcatanzaro_ mclasen ryanlerch stickster
13:05:43 <stickster> brilliant, let's go
13:05:55 * stickster sees how far his keyboard magic has deteriorated
13:06:53 <stickster> #topic gnome-nautilus-terminal by default
13:06:58 <stickster> #link https://pagure.io/fedora-workstation/issue/10
13:07:10 <stickster> oh, is kalev around?
13:07:18 * mclasen hasn't seen him yet
13:07:46 <stickster> I don't see him online either. But is there someone with a recent install that can verify whether this package was included by default? cschalle?
13:07:58 <cschalle> will check, gimme a sec
13:08:08 * mclasen is on rawhide
13:08:19 <cschalle> no, it is not there
13:09:20 <stickster> Hm, OK. mclasen, can you follow up with Kalev to see how to get this done? I'm not sure anyone's ever thrilled with throwing a new package in after Beta freeze, but it looks like we have to have that discussion
13:09:38 <mclasen> yes
13:09:46 <mcatanzaro_> It should be uncontroversial to add such a small package.
13:10:02 <cschalle> at the same time maybe he can add the workstation-repositories package to the workstation comps
13:10:04 <stickster> this should not be a big disturbance in the Force, but plenty of people have said that in the past and been wrong (even if I'm probably not in this case)
13:10:13 <stickster> cschalle: That's coming up on the agenda, don't worry ;-)
13:10:21 <cschalle> ok :)
13:10:38 <stickster> #action mclasen follow up with kalev to get this package into the complement as was expected last month
13:11:01 <stickster> #topic fedora-workstation-repositories not in default install
13:11:03 <stickster> #link https://pagure.io/fedora-workstation/issue/17
13:11:08 <stickster> cschalle: here you go
13:12:29 <cschalle> I thought we had this in the defaults for F25, so I am suprised it is missing in F26
13:12:30 <stickster> sgallagh pointed out on list he hadn't seen any Change filed for this package. Is filing a Change a requirement for any alteration to default repositories? Just for those that a WG decides on? Some other criterion?
13:13:07 <stickster> cschalle: I don't think it was, but I could be wrong
13:13:16 <sgallagh> stickster: Officially, Fedora bans any package from shipping new repos except for the fedora-repos package.
13:13:18 <cschalle> since the WG chooses which repos to enable, I would assume we decide to include this package too
13:13:23 * mclasen not particularly interested in extra paperwork
13:14:01 <sgallagh> When the topic of adding extra repos for Workstation came up, I remember mentioning to you guys that they'd need to go into fedora-repos or else get permission from FESCo to ship another repo.
13:14:11 <sgallagh> A Change would have been the ideal process for that.
13:14:33 <stickster> cschalle: Refresh my memory: How did we end up with fedora-workstation-repositories in the first place?  Were we directed away from fedora-repos by anyone?
13:14:35 <mcatanzaro_> We already got approval for this in the past, right, and just forgot to enable it?
13:14:42 <sgallagh> Mostly I'm concerned that rel-eng is going to come down on this hard.
13:14:59 <cschalle> stickster, the 3rd party policy is that it applies per edition, so we wanted a separate place for them
13:15:20 <stickster> cschalle: right, but why this and not a fedora-repos-workstation subpackage?
13:15:32 <cschalle> stickster, what is a subpackage?
13:15:43 <sgallagh> cschalle: Built from the same SRPM into a separate binary RPM
13:15:44 <stickster> cschalle: just like -libs or -doc or... etc.
13:16:33 <ryanlerch> https://apps.fedoraproject.org/packages/fedora-repos
13:16:36 * mclasen can't deal with the bureaucracy today
13:16:43 <cschalle> well a) I was not ever aware that there was a preference for subpackages, and b) I guess setting up a new package than to get co-maintainership of another package
13:16:44 <ryanlerch> ^ there is a -rawhide subpackage already
13:18:08 <juhp_> ryanlerch, sure
13:18:08 <sgallagh> To avoid the bureaucracy question, can we just ping dgilmore and mboddu and ask them to review the current state of things and then look at a longer-term solution for F27?
13:18:29 <sgallagh> I don't want anyone getting surprised by this
13:18:59 <stickster> sgallagh: I agree we should do that, but before we do, let's also see what the fallout is for f26 without this package
13:19:05 <stickster> well, by default at least
13:19:16 <cschalle> well I am fine with adding a change request, my argument would be that this was implicitly approved through the 3rd party proposal, and I don't really have any strong preferences on how it is packaged as long as adding/removing repos don't become a huge process
13:19:47 <stickster> cschalle: in the case of a subpackage, would be simply proposing a fix via a pull request in Pagure I think
13:19:52 <sgallagh> (FWIW, the main reason for forbidding the dropping in of new repos is because it means that it opens the door for unknown of overriding software to get installed on people's machines)
13:20:15 <sgallagh> In this case, the Workstation WG repos are reasonably trustworthy and vetted (right?)
13:20:27 <sgallagh> stickster: Yes, exactly that
13:20:30 * stickster also points out that the Change request stuff is all past for F26, so that's more about F27 too
13:20:41 <juhp_> indeed
13:20:47 <mcatanzaro_> It's the Google Chrome repo plus one copr for... something weird...  PyCharm?
13:20:58 <cschalle> sgallagh, yes, nothing goes in there without the explicit approval of the working group
13:21:06 <sgallagh> Yeah, I don't want to force bureaucracy... we just have legitimate reasons for not allowing packages to add repos.
13:21:15 <sgallagh> Best if that gets the occasional double-check.
13:21:43 <cschalle> sgallagh, the 3rd party policy talks about a lot of those things like 'no overwriting system libs etc'
13:22:43 <sgallagh> ack
13:23:19 <mcatanzaro_> It is PyCharm, and it has not been updated since 2015
13:23:36 <juhp_> what is PyCharn for btw?
13:23:39 <stickster> mcatanzaro_: cschalle: So am I correct that nothing terrible happens to users who don't get this package installed right away, except it delays our plans for introducing it. (I have concerns about copr but will take those up offline.)
13:23:43 <stickster> juhp_: IDE for Python development
13:23:45 <juhp_> erm PyCharm
13:23:45 <sgallagh> juhp_: It's a really nice Python IDE
13:23:49 <juhp_> ah okay
13:24:09 <juhp_> ah yeah rings a bell
13:24:20 <mcatanzaro_> stickster: Yup
13:24:37 <mcatanzaro_> And I am wrong, it's being regularly updated... but not on rawhide. I guess that's just how coprs work.
13:24:38 <stickster> It was one of our early choices for "bring cool stuff to developers since they are an important audience for Workstation"
13:25:07 <stickster> mcatanzaro_: yeah, if I were running a copr, I'd probably shy away from doing Rawhide builds unless I had a lot of time on my hands
13:27:38 <stickster> cschalle: So is the only thing to do here: Check with mboddu about adding -workstation subpackage to fedora-repos, or some other fix as recommended by rel-eng
13:29:03 <stickster> ?
13:29:28 <cschalle> stickster, sure
13:29:50 <stickster> cschalle: Do you want to take this up or should we assign elsewhere?
13:30:01 <cschalle> stickster, I can do it
13:30:13 * ryanlerch is not sure of the difference between a seperate package or a subpackage in this instance. they will both funtion identically, right?
13:30:43 <juhp_> I guess it can be merged by a PR in pagure?
13:30:43 <stickster> ryanlerch: correct. the difference is a subpackage adheres to the guideline that repos don't get installed other than through fedora-repos
13:30:52 <stickster> juhp_: correct
13:30:55 <ryanlerch> ah, ok
13:31:04 <ryanlerch> thanks stickster
13:31:17 <stickster> #action cschalle check with mboddu about adding -workstation subpackage to fedora-repos as a solution to getting Workstation specific repos in
13:31:30 <sgallagh> stickster: The subtext there being "doesn't get installed without someone from rel-eng double-checking it"
13:31:35 <stickster> *nod
13:32:35 <stickster> sgallagh: right, I don't have a problem with another pair of eyes -- Workstation WG typically has our homework done for what we want to add.
13:33:04 <sgallagh> stickster: I also have no problem if rel-eng says "Go ahead, we trust you, make your own repo package"
13:33:08 <stickster> sgallagh: at the same time we should make sure that it's clear what that review is for -- not as an "imprimatur" on something the WG is charged to do
13:33:15 <sgallagh> I just want to make sure that conversation happens
13:33:17 <stickster> sgallagh: +1
13:34:11 <stickster> Like, does the repo work, what happens in upgrade scenarios, etc. -- those things (and others) make sense for review from a release bits POV
13:34:23 <sgallagh> yes
13:34:40 <stickster> OK, I think we beat this one to death :-)  Is there anything else to cover on this topic before we move on?
13:34:53 <sgallagh> And if (for example) we ended up changing the default repo location or something, we'd want there to be cross-communication so that everything updates together
13:34:57 * sgallagh is done
13:35:08 <stickster> sgallagh: Agreed, another good point
13:35:35 * sgallagh remembers there was a pitch to switch over to /usr/lib/ a while back...
13:35:49 <sgallagh> /usr/share, rather
13:36:02 <stickster> oh right, for systemwide repos
13:36:14 <stickster> anyhoo... OK, 20 min until I turn into a pumpkin (err, next meeting)... so let's move on
13:36:32 <juhp_> stickster, lol
13:39:27 <stickster> sorry guys, pagure was having a fit
13:39:53 <stickster> #topic Default apps decided by Workstation WG
13:39:55 <stickster> #link https://pagure.io/fedora-workstation/issue/17
13:40:15 <mcatanzaro_> So the problem here is that blivet-gui has appeared in our default install
13:40:42 <juhp_> wrong link I guess :)
13:41:08 <mcatanzaro_> #link https://pagure.io/fedora-workstation/issue/14
13:41:13 <juhp_> thanks
13:41:16 <mcatanzaro_> #link https://pagure.io/fedora-workstation/issue/14
13:41:40 <mcatanzaro_> It's an advanced disk partitioning tool that is not designed in line with the GNOME apps that make up our default install
13:41:42 <mcatanzaro_> #chair rdieter
13:41:42 <zodbot> Current chairs: cschalle juhp_ mcatanzaro_ mclasen rdieter ryanlerch stickster
13:41:58 <stickster> Hey rdieter! I didn't see you earlier, sorry
13:42:19 * stickster not using his little name-list fifo that auto-updates... should fix that at some point.
13:42:26 <mcatanzaro_> So we should want it gone. But I noticed that we don't appear to have any formal power to make it go away, other than the implicit understanding that Workstation WG gets to decide what is present in Workstation... there is not e.g. a blocker criterion we can use for this.
13:43:21 <ryanlerch> isnt blivet-gui used in anaconda now? does that have an impact on it being installed by default?
13:43:24 <mcatanzaro_> So my proposal is to add a sentence to the application and launcher guidelines (which are incorporated by reference into the blocker criteria) stating that applications installed by default must be approved by the WG.
13:43:28 * ryanlerch is not sure
13:43:36 <mcatanzaro_> And yes, blivet-gui is required by anaconda... that's the problem.
13:43:37 <juhp_> ryanlerch, I was wondering that too
13:43:42 <mclasen> ryanlerch: its a typical example of app/no-app confusion
13:43:51 <mclasen> if it is a part of anaconda, it should be just a part of anaconda
13:43:55 <mcatanzaro_> Maintainers need to hide the desktop file, split it to a subpackage, or otherwise figure something out
13:43:59 <mclasen> if it is a desktop application, it should not be a part of anaconda
13:44:12 <juhp_> right
13:44:55 <mcatanzaro_> They should *really* also uninstall anaconda at the end of the installation process, but that's not sufficient IMO as it wouldn't prevent blivet-gui from appearing in the live image, which is the first impression of our distro. We shouldn't permit it to be there either.
13:45:00 <stickster> Like there should be a blivet-app subpackage probably... and that would eliminate the issue
13:45:18 <juhp_> I often uninstall anaconda by hand after install...
13:45:31 <mcatanzaro_> So the problem is the package maintainers disagree, so we need a blocker criterion to use for this.
13:45:32 <juhp_> mcatanzaro_, +1
13:45:45 <stickster> #info anaconda requires blivet-gui so the .desktop file is being dragged in by default... a blocker criterion here would allow us to force a solution
13:46:01 <juhp_> is there a bug?
13:46:26 <stickster> I don't think blocker criteria should be abused in just any ol' case, but I think a criterion here that apps don't just show up randomly without WG oversight is quite rational
13:46:53 <mcatanzaro_> #link https://bugzilla.redhat.com/show_bug.cgi?id=1449752
13:47:16 <mcatanzaro_> So my concrete proposal is to add the following to our applications and launchers guidelines: "App launchers installed by default must be approved by the Workstation WG."
13:47:31 <stickster> mcatanzaro_: So we probably need to ask QA to add that criterion for Beta for F27, but we can also propose the subpackaging here to take care of the issue and request a Freeze Exception, right?
13:48:12 <mcatanzaro_> stickster: We could ask QA to add a new criterion, or we could just edit our existing guidelines. A blocker criterion already exists to ensure apps comply with those.
13:48:16 * stickster notes we should do this in a friendly way with anaconda/blivet folks, who are doing their best in an often thankless job
13:49:09 <mcatanzaro_> I've already asked nicely for them to add a subpackage... I think we should just propose it as a blocker and let them sort it out.
13:49:23 <stickster> mcatanzaro_: Ah OK -- if our guidelines are incorporated by reference that's fine, but we do want to call attention to the change so no one gets surprised
13:49:23 <mcatanzaro_> We wouldn't be discussing this now if that had worked. ;)
13:49:34 <stickster> lol, right on
13:49:52 <mcatanzaro_> I can mail desktop@ after making the change, would that suffice for calling attention to it?
13:50:12 <stickster> mcatanzaro_: I would also point out to test@ which is the QA list.
13:50:21 <mcatanzaro_> OK
13:51:04 <stickster> changing goalposts at the freeze time is a bit hinky so you may want to consult with QA on how they would like to see this handled in a reasonable and colleagial way
13:51:08 <stickster> collegial. Sorry
13:51:16 <cschalle> sidenote here though is that blivet-gui can actually handle our default lvm install, gnome-disks can't :(
13:51:30 <stickster> ruh roh
13:51:33 <mcatanzaro_> Yeah, we should probably stop using LVM. ;)
13:51:45 <mcatanzaro_> That is another longstanding issue we can discuss another time, though.
13:52:01 <stickster> mcatanzaro_: that's a different fight, and I actually disagree speaking as a user who's had to shuffle space
13:52:08 <stickster> but we'll deal with that later :-D
13:52:13 <mcatanzaro_> I'm not going to propose this as a beta blocker, but as a final blocker, so still plenty of time before freeze. (adamw you'll probably be interested in the above discussion.)
13:52:15 <stickster> mcatanzaro_: Can I #action you with this stuff then?
13:52:25 <mcatanzaro_> Yes
13:53:00 <stickster> #action mcatanzaro_ consult with QA and propose change in our guidelines for Final blocker, so we can get blivet-gui removed from the default Workstation install
13:53:10 <stickster> #undo
13:53:10 <zodbot> Removing item from minutes: ACTION by stickster at 13:53:00 : mcatanzaro_ consult with QA and propose change in our guidelines for Final blocker, so we can get blivet-gui removed from the default Workstation install
13:53:15 <stickster> #action mcatanzaro_ consult with QA and make change in our guidelines for Final blocker, so we can get blivet-gui removed from the default Workstation install
13:53:39 <mcatanzaro_> I'll mail test@ then.
13:54:16 * stickster thinks this doesn't sound like a difficult fix, but is often wrong judging by past results :-)
13:54:38 <stickster> OK, anything else on this?
13:55:08 <stickster> Allrighty then...
13:55:11 <stickster> #topic Next meeting
13:55:37 <stickster> #info Next meeting is 09:00am EDT (1300 UTC) on 2017-Jun-05, see you then!
13:55:54 <stickster> #action stickster file updates on all Pagure issues covered in this meeting
13:55:58 <ryanlerch> \0/
13:56:18 <ryanlerch> i will also update the Wallpaper ticket too with current progress
13:56:19 <stickster> Also, if you want something covered next time, mark the issue with "meeting" tag in Pagure please
13:56:24 <stickster> thanks ryanlerch !
13:56:31 <stickster> Thanks for coming everyone :-)
13:56:34 <stickster> #endmeeting