14:00:00 <stickster> #startmeeting Workstation WG
14:00:00 <zodbot> Meeting started Wed Apr 27 14:00:00 2016 UTC.  The chair is stickster. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:00:00 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
14:00:00 <zodbot> The meeting name has been set to 'workstation_wg'
14:00:02 <stickster> #meetingname workstation
14:00:02 <zodbot> The meeting name has been set to 'workstation'
14:00:04 <stickster> #topic Roll call
14:00:05 <stickster> .hello pfrields
14:00:06 <zodbot> stickster: pfrields 'Paul W. Frields' <stickster@gmail.com>
14:01:24 <cschalle> .hello cschalle
14:01:25 <zodbot> cschalle: Sorry, but you don't exist
14:01:32 <cschalle> noooooo
14:01:33 <mcatanzaro> .hello catanzaro
14:01:34 <zodbot> mcatanzaro: catanzaro 'None' <mcatanzaro@gnome.org>
14:01:40 * otaylor is here now
14:01:43 <stickster> cschalle: "cschaller" ?
14:01:48 <cschalle> .hello cschaller
14:01:49 <zodbot> cschalle: Sorry, but you don't exist
14:01:49 <stickster> or "uraeus"
14:01:55 <cschalle> .hello uraeus
14:01:56 <zodbot> cschalle: uraeus 'Christian Schaller' <uraeus@linuxrising.org>
14:02:00 <cschalle> I am alive!
14:02:03 <stickster> cschalle: I know you better than you know yourself, apparently
14:02:10 <cschalle> finally the confirmation I been looking for my whole life
14:02:14 <stickster> haha
14:02:56 <mclasen> .hello mclasen
14:02:57 <zodbot> mclasen: mclasen 'Matthias Clasen' <mclasen@redhat.com>
14:03:05 <cschalle> .hello dolly
14:03:05 <zodbot> cschalle: Sorry, but you don't exist
14:03:30 <ryanlerch> .hello
14:03:30 <zodbot> ryanlerch: (hello <an alias, 1 argument>) -- Alias for "hellomynameis $1".
14:03:41 <ryanlerch> .hello ryanlerch
14:03:44 <zodbot> ryanlerch: ryanlerch 'ryan lerch' <rlerch@redhat.com>
14:04:02 <cschalle> the bot is not a big louis armstrong fan it seems
14:04:26 <stickster> apparently not
14:04:27 <otaylor> .hello otaylor
14:04:28 <zodbot> otaylor: otaylor 'Owen Taylor' <otaylor@redhat.com>
14:04:31 <stickster> OK, we have quorum, let's go
14:04:38 <stickster> o/ everyone by the way :-)
14:04:52 <stickster> I'm going to pre-empt one topic because jkurik reminded us
14:04:59 <stickster> #topic Approve release blocking deliverables
14:05:02 <stickster> #link https://fedoraproject.org/wiki/Fedora_Program_Management/ReleaseBlocking/Fedora24#Workstation
14:05:36 <stickster> The only things blocking on this list are the Live and netinst images. Given where we are with the new media writer, this list makes sense to me, although it could change for F25
14:05:58 <mcatanzaro> +1, revisit media writer for F25
14:06:34 <otaylor> stickster:  seems fine to me
14:06:48 <cschalle> yeah, the list seems fine, I still think we should offer the Media write for Windows even if its not 'the primary download'
14:07:06 <mclasen> .hello mclasen
14:07:07 <zodbot> mclasen: mclasen 'Matthias Clasen' <mclasen@redhat.com>
14:07:12 <mclasen> grr, xwayland crash
14:07:29 <cschalle> mclasen, you say hello and I say goodbye
14:07:49 <mclasen> so, do we just say +1 to the release blocking deliverables ?
14:07:50 <mclasen> I don
14:08:22 <cschalle> +1 deliverables
14:08:22 <mclasen> 't even know that we care about the netinst, but...
14:08:23 <stickster> correct, unless you see an issue with something either missing, or that shouldn't block but does.
14:09:12 * stickster thinks it adds little if any risk (since likely subsumed by Server requirement), and if it helps people with no media writer handy, great
14:09:27 <stickster> would rather not get stuck in a discussion about the value of netinst here :-)
14:09:35 <mclasen> for f25, we want to get the ostree image added to that list, I guess
14:09:43 <mcatanzaro> +1 status quo (release-blocking netinst)
14:09:46 <stickster> correct, f25 I think we'll see some substantial change here
14:09:54 <cschalle> stickster, also as I remember we needed netinst for Boxes
14:09:54 <mclasen> cschalle pointed out here that we may need the netinst for boxes, anyway
14:10:03 <stickster> Ah yes, that was the thing I was forgetting
14:11:28 <stickster> mclasen: +1?
14:11:33 <mclasen> yes, +1
14:11:46 <stickster> ryanlerch: ?
14:12:35 <ryanlerch> +1
14:12:36 * rdieter arrives late, sorry
14:13:11 <stickster> rdieter: Hiya, we are just making sure https://fedoraproject.org/wiki/Fedora_Program_Management/ReleaseBlocking/Fedora24#Workstation is correct -- so far +6, -0.
14:14:18 <rdieter> fine by me +1
14:14:27 <stickster> #agreed Release deliverable list looks fine, +7 / -0.
14:15:20 <stickster> #action stickster update the ticket at https://fedorahosted.org/workstation/ticket/9
14:15:30 <stickster> #topic Third party software policy draft
14:15:32 <rdieter> mention of "netinst for Boxes", is that different than the usual/normal netinst ?
14:15:42 <mclasen> no, just the usual netinst, afaik
14:15:50 <rdieter> k
14:15:52 <stickster> just the usual
14:15:52 <cschalle> rdieter, no it was just that we need netinst for the boxes usecase
14:15:59 <stickster> #link https://fedoraproject.org/wiki/Workstation/Third_party_software_proposal
14:16:06 <cschalle> +1
14:16:11 <cschalle> :)
14:17:38 <stickster> So the idea here was for folks to give a general +1 on the list so we could bring this draft up with the Council for review, discussions, and revision
14:17:53 <stickster> mcatanzaro already +1 on the list.
14:17:59 <mcatanzaro> +1
14:18:15 <rdieter> +1
14:18:23 <cschalle> +1
14:18:24 <stickster> #action stickster Add back in provision on third party repositories not breaking Fedora dependency chains
14:18:27 <ryanlerch> +1
14:18:33 <mclasen> +2
14:18:35 <mclasen> err, +1
14:18:35 <stickster> ha
14:18:43 <stickster> mclasen: You must vote in Chicago
14:18:56 <stickster> +1 from me too.
14:19:19 <mcatanzaro> I do want to ask about upstream vs. downstream xdg-app runtimes though. Seems a much harder sell to tell distros that xdg-app is cross-platform, please use a Fedora runtime for your app instead of a GNOME or generic freedesktop runtime.
14:20:38 <otaylor> +1
14:20:56 <stickster> cschalle: is there the notion of only one, or rather a small, controlled set of official runtimes?
14:21:03 <ryanlerch> if an app uses a generic gnome or freedesktop runtime, wont xdg-app just use that one?
14:21:07 <cschalle> mcatanzaro, well people can use whatever they want, but for most 3rd parties I think they prefer one that has a long term support commitment, so my expectation is that most 3rd parties will prefer a RHEL one, since they don't want to have to revisit their runtime on a rapid fire basis
14:21:28 <rdieter> mcatanzaro: in the end, does it really matter that much?  (assuming runtimes don't conflict name-wise)  who cares which runtimes are used?
14:21:30 <stickster> cschalle: that's waht I was getting at, a runtime that's aimed at not changing out from under a third party
14:21:33 <otaylor> mcatanzaro: the wording in the proposal is "the official xdg-app runtime" not "the official fedora xdg-app runtime"
14:21:40 <cschalle> stickster, I think we hope there will be a small core of de-facto ones, but there is no strick enforcement in xdg-app of a limited set of runtimes
14:21:48 <stickster> otaylor: Right, that's ambiguous to me
14:22:23 <cschalle> otaylor, but the two are likely to be the same, since we are the ones making it
14:23:26 <cschalle> for xdg-app to be successful it needs dependable runtimes and I don't see a viable path for that apart from a distro derived runtime
14:23:34 <otaylor> stickster: It's sort of meant to be ambiguous, since we don't have a final picture in place - what it's supposed to represent is that we don't want a proliferation of runtimes
14:23:43 <mcatanzaro> cschalle, that makes sense to me. rdieter, my concern is runtime proliferation, we don't want Skype to use a Fedora runtime, Chrome to use an Ubuntu runtime, Dropbox to use RHEL runtime, and each to have its own copy of WebKit... I was expecting we would tell people to use a distro-agnostic runtime to avoid this. otaylor, I think we need to define what the official runtime is. :)
14:24:02 <rdieter> mcatanzaro: sure, but I think that's outside the scope of this policy (initially)
14:24:28 <stickster> cschalle: otaylor: *: Would it be more accurate to say this: "An xdg-app hosted in official Fedora repositories must rely on either an official xdg-app runtime (e.g. provided by freedesktop.org or the appropriate xdg-app upstream), or an official runtime that is also available in Fedora repositorites." ?
14:24:44 <rdieter> which brings me to an editorial comment: for the "Rules" sections, I'd strongly prefer that each "rule" to be clearly marked as required/must or should/optional/advisory
14:25:04 <otaylor> cschalle: I not sure that the Fedora runtime that we use to build xdg-apps of fedora rpms can be the "official runtime" - because it needs to be compatible with each fedora release
14:25:08 <stickster> rdieter: Not a bad idea. Would you be willing to go through and markup appropriately? I'd be totally in favor of that.
14:25:15 <stickster> (the idea I mean, not just you editing) :-D
14:25:16 <rdieter> so on this specific case, we could offer an "advisory" suggestion on non-proliferation of xdg-app runtimes
14:25:18 <mcatanzaro> cschalle: I think we also need to think about the politics of this... it's going to be harder to compete with Snappy on Ubuntu if xdg-app is seen as some sort of trojan horse for making Ubuntu depend on Fedora.
14:25:45 <stickster> mcatanzaro: I think that's the point of having an upstream xdg-app runtime.
14:25:59 <cschalle> well I want the basis of a runtime to be an implementation details, I don't want there to be a 'Fedora runtime' as that will surely trigger the million runtimes situation, I rather have a medium lifecycle xdg-app runtime that happens to have been implemented using Fedora packages
14:26:19 <stickster> mcatanzaro: I mean, I think the point of the upstream runtime is to be not stuck in or branded with Fedora, but available through e.g. fd.o
14:26:46 <stickster> cschalle: And that would be hosted upstream?
14:26:50 <cschalle> mcatanzaro, trust me I been thinking about the politics a lot here, but there needs to be realism in what we are doing in terms of having resources to pull it off, and nothing would harm xdg-app more than badly maintained official runtime
14:26:52 <cschalle> stickster, yes
14:27:20 <rdieter> curious, is anyone currently tasked with making "official" runtimes?
14:27:32 <stickster> cschalle: OK, we should be clear about that in the draft. If the idea for implementation is that the official upstream would be hosted upstream and packaged in Fedora, like so many other things, let's just say that
14:27:43 <rdieter> if not, will there?  or rather, who's job is to do that then?
14:27:46 <cschalle> stickster, that is fine
14:27:48 <mcatanzaro> cschalle: I doubt it would make much difference whether we maintain the runtime upstream or downstream. Anyway if it's not branded as a Fedora runtime, and the Fedora packages are just an implementation detail, I think that makes it an easier sell....
14:27:48 <stickster> sorry, the official upstream *runtime*
14:28:05 <cschalle> rdieter, David King has been given the task of putting it together
14:28:12 <rdieter> cschalle: k, thx
14:28:24 <otaylor> stickster: I don't think that woudl be the plan no - because an "upstream" runtime would duplicate many packages in Fedora - it would be like the ultimate bundling issue..
14:28:29 <rdieter> so it should be easy to point to that work
14:28:39 <stickster> otaylor: *ANY* runtime could have that case though.
14:28:54 <mclasen> duplication is inherent in containerization at some level
14:28:55 <stickster> otaylor: The thing is that when you move from F25 -> F26, it doesn't mean your app breaks ;-)
14:29:07 <mclasen> its how you get the isolation that is one of the selling points of the exercise
14:29:09 <stickster> mclasen: +1, that's what I was getting at too.
14:29:43 <otaylor> What I'm saying is that I dont' think rebuilding a yocto-based runtime on fedora servers makes any sort of sense
14:30:11 <stickster> are we talking about rebuilding?
14:30:28 <otaylor> stickster: "the official upstream would be hosted upstream and packaged in Fedora" - not sure what you meant by that
14:30:42 <stickster> otaylor: I didn't mean rebuilding the sources of the upstream runtime.
14:31:01 <stickster> otaylor: I thought the idea was we treat that as a *payload* within a package
14:31:13 <otaylor> stickster: But I think it's either a) we just point people to a fedora-hosted fedora-based runtime b) we point people to a freedesktop.org-hosted fedora-based runtime c) we give an option for people to use a) or a freedeskop.org hosted non-fedora-based runtime
14:31:16 <stickster> Tinkering with it, to me, would induce a risk that the runtime is no longer going to act as expected.
14:31:21 <ryanlerch> what would be the difference between a fedora runtime and the gnome runtime?
14:31:58 <mcatanzaro> otaylor: +1 to (b), I think making this as distro-agnostic as possible will be key to success.
14:32:14 <otaylor> ryanlerch: A fedora runtime would be based on Fedora packages and might include some packages (Qt say) that are commonly used by apps but not part of GNOME
14:32:18 <stickster> I think b) reduces proliferation
14:32:36 <mclasen> ryanlerch: the fedora runtime is populated from fedora rpm, the upstream runtime is (currently) based on yocto builds
14:32:55 <stickster> But we should also realize nothing prevents people from *doing* c) and s/freedesktop.org/something else/ in the future
14:33:04 <mcatanzaro> The freedesktop runtime could conceivably contain both GNOME and Qt stuff, though. As those are the two major platforms we expect people to use to build apps.
14:33:16 <ryanlerch> so if an xdg-app requires a fedora runtime, the user (even if on another distro) needs to use the fedora runtime?
14:33:27 <cschalle> ryanlerch, yes
14:33:49 <otaylor> ryanlerch: yes. (And would be able to... so it's a "political" and disk space issue, not a feasibility issue)
14:34:27 <mclasen> ryanlerch: in part, that is how xdg-app is more portable than an rpm - you use the versions of the libraries it was built against when running it, instead of the whatever is on your host system
14:35:48 <mclasen> one way to look at a runtime is "standardized" bundle of app dependencies
14:36:06 <ryanlerch> mclasen: but might it end up being kinda that way anyway -- if a developer bundles their app on fedora, they might end up adding the fedora runtime as a dep
14:36:14 <cschalle> anyway, I think the upstream xdg-app implementation details goes a little beyond our current topic
14:36:23 <ryanlerch> sorry!
14:36:28 <stickster> #info stickster realizes the approval fo the draft to go to Council should have been noted a while back.
14:36:51 <stickster> #agreed Draft can go to council regardless of further xdg-app discussion, since that aspect is obviously still in motion
14:36:55 <stickster> #undo
14:36:55 <zodbot> Removing item from minutes: AGREED by stickster at 14:36:51 : Draft can go to council regardless of further xdg-app discussion, since that aspect is obviously still in motion
14:37:12 <mclasen> ryanlerch: which runtime/sdk you build your app against is an explicit decision - part of the metadata you feed into a build process
14:37:13 <stickster> #agreed Draft can go to council regardless of further xdg-app discussion, since that aspect is obviously still in motion (+7/-0)
14:37:21 <stickster> #topic xdg-app runtimes
14:37:32 * stickster doing a crap job of topic tracking :-)
14:38:49 <otaylor> stickster: Going back to the document, I'm not sure I understand why "Any xdg-app hosted by Fedora must be built using the official xdg-app runtime. " since the document is about "Tier 2" software as I understand it. The second sentence is a "strong recommendation" not a requirement - we could say "Third party are strongly recommended to use an official xdg-app runtime (list to be maintained by the Workstation working g
14:38:49 <otaylor> roup) to avoid ..."
14:38:49 <stickster> #idea otaylor - we either (A) point people at Fedora-hosted, Fedora-based runtime; (B) point people at freedesktop.org-hosted, Fedora-based runtime; (C) give an option for people to do either (A) or a freedesktop.org-hosted, non-Fedora-based runtime
14:39:35 <stickster> otaylor: Good catch. I'm going to make that change now in an act of total fiat ;-)
14:39:36 * linuxmodder late and  merely  staying utd
14:40:24 <cschalle> otaylor, I was trying to also cover the case of people wanting to host their xdg-app on Fedora infrastructure in the future, but I guess that can be left out for now as it is somewhat beside the scope of the doc
14:40:29 <stickster> otaylor: Wait, no
14:40:39 * stickster just re-read the doc -- we are apparently trying to cover two cases
14:41:05 <stickster> OK, that's a failure in the doc to be clear
14:41:18 <otaylor> stickster: Yes, it's trying to cover two cases, but I'm saying that the first one is outside the scope of the document (as I see it)
14:41:20 <stickster> otaylor: Sorry, I'm back to agreeing with you, thanks.
14:42:05 <cschalle> otaylor, stickster : actually it is part of the cope - Provide Fedora repositories that contain software packaged in ways other than RPM.
14:42:09 <cschalle> from the Goals section
14:42:58 * stickster makes an edit that should clarify
14:43:01 <otaylor> In any case, it could also be pointed to a list of runtimes maintained by the Working group in the same fashion - (I'm not sure it's the *same* list of runtimes, but that could be finesed later)
14:43:18 <cschalle> good idea
14:44:40 * stickster encourages people to make quick edits, save, and shout here to refresh :-)
14:44:44 <stickster> it's a wiki after all
14:45:46 <stickster> #info stickster moved mcatanzaro's suggestion out to the text as well, as noted on list
14:46:24 <stickster> So I'm getting lost on what we are trying to *decide* here about xdg-app guidelines, other than "what's the meaning of 'the official xdg-app runtime'".
14:47:04 <otaylor> I'l take an action to do some exploration about what it would look like to have a freedesktop.org hosted repository based on Fedora - what the policies, etc, would look around that.  (Not related to the third-party policy immediately, but seems like we need to get some clarity there)
14:47:40 <stickster> Oh, I forgot to:
14:47:54 <stickster> #chair otaylor cschalle rdieter mcatanzaro mclasen ryanlerch
14:47:54 <zodbot> Current chairs: cschalle mcatanzaro mclasen otaylor rdieter ryanlerch stickster
14:48:11 <mcatanzaro> Thanks otaylor
14:48:16 <cschalle> ok great
14:48:16 <stickster> otaylor: feel free to #acion yourself for the minutes
14:48:22 <stickster> um #action even.
14:48:41 <stickster> usage: #action <nick> <thing that nick's going to do>
14:49:28 <otaylor> #action otaylor to research what it would mean to have a freedesktop.org hosted runtime based on Fedora; what the policies would be around it, what buy-in we would need, etc, and will report back to the working group
14:50:35 <stickster> #action stickster condense notes in wiki and clarify otaylor is researching this part
14:50:50 <stickster> What else do we need to decide or act on here wrt. xdg-apps?
14:51:22 <mclasen> as an infomation, owen and I are meeting with the rel-eng team soon to talk about building xdg-apps
14:51:53 <mclasen> tomorrow, actually
14:52:07 <stickster> #info mclasen + otaylor meeting with rel-eng on Thu 2016-Apr-28 to discuss building xdg-apps
14:52:41 * mclasen has to run to another meeting
14:52:42 <stickster> #action stickster to move ahead with relaying draft to council after clarifying other wiki bits noted above
14:52:56 <stickster> cschalle: Anything else you can think of that we need to cover here?
14:53:06 <cschalle> no, I think we covered everything
14:53:38 <stickster> #topic Open floor
14:53:46 <jkurik> Thanks the whole WG for the review of release blocking deliverables.
14:53:49 <cschalle> I do plan to have a chat with mattdm to cover the recent media writer discussions as I think we need a general fedora policy to have sane requirements for Windows and Mac software to be shipped
14:54:00 <stickster> thank you for giving us the heads-up jkurik!
14:54:33 <stickster> cschalle: Agreed. It doesn't make much sense to set requirements that keep us from actually converting other OS users.
14:56:52 <stickster> OK, closing up then, thanks to everyone for coming! cschalle rdieter mcatanzaro mclasen otaylor ryanlerch
14:56:57 <stickster> #endmeeting