16:04:30 #startmeeting env and stacks wg (2014-01-14) 16:04:30 Meeting started Tue Jan 14 16:04:30 2014 UTC. The chair is mmaslano. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:04:30 Useful Commands: #action #agreed #halp #info #idea #link #topic. 16:04:39 #meetingname env and stacks 16:04:39 The meeting name has been set to 'env_and_stacks' 16:04:50 #chair abadger1999 pkovar tjanez samkottler bkabrda handsome_pirate hhorak juhp 16:04:50 Current chairs: abadger1999 bkabrda handsome_pirate hhorak juhp mmaslano pkovar samkottler tjanez 16:05:10 .hellomynameis skottler 16:05:12 samkottler: skottler 'Sam Kottler' 16:05:37 .hellowmynameis pingou 16:06:31 hi 16:06:32 I'm obviously here 16:07:01 * pkovar1 is here 16:07:47 hi 16:08:53 #topic init process 16:09:26 so we continue in discussion of https://fedoraproject.org/wiki/User:Mmaslano/Draft:Env_and_Stack_PRD 16:09:57 please, look at mailing list. Jens posted comment there 16:10:08 and tjanez posted there Vision statement 16:12:16 +1 for tjanez and his Vision statement 16:12:42 do we want to add something about improving tools and workflow? 16:12:45 tjanez: ? 16:14:45 It feels like tools and workflow are actually concrete result of stacks and environment to me, so I don't see a need to have them in vision. 16:15:29 +1 For tjanez and his Vision Statement. I agree with hhorak. 16:16:16 ok +1 for tjanez statement 16:16:55 more votes? 16:17:09 tjanez: do you want to vote for yourself? 16:17:24 are we voting on that tonight? 16:18:22 * tjanez is back 16:18:35 +1 16:18:43 +1 tjanez 16:19:50 +1 (again) 16:19:55 samkottler: I hope we can vote on improvements 16:20:00 +1 (obviously :)) 16:20:01 yeah, I'm +1 for it 16:20:06 Do we need to include anything of how we fit in with the other working groups, or on top of Fedora Core or the rings? 16:20:10 and final approval of PRD do on mailing list, maybe 16:21:13 #agreed add tjanez vission and mission statement (+7,-0,0) 16:22:00 #action tjanez will add section about vision and mission statement 16:22:21 drieden: I understand our WG should provide packaging mechanism, tooling, etc. 16:22:26 drieden: maybe we should add it 16:23:18 Does anyone think we need something about the environment in the Mission statement 16:23:33 or do you like it as it is? 16:24:43 * abadger1999 likes it 16:25:12 tjanez Do you have a suggestion of what you would like to see to cover the environment? 16:25:24 tjanez: I like that it is quite general, so I'm ok with the current wording. 16:26:21 drieden: No, I don't have a suggestion that's why I didn't put it in 16:26:33 Ok, let's move on then. 16:26:47 tjanez Ok, not a problem. 16:28:18 So, I have some open questions regarding the PRD 16:28:54 Could we merge the contents of "Document Structure" with "What this document describes" and delete the latter? 16:29:40 yeah, good idea 16:29:47 tjanez: +1 for merging 16:31:01 "! which are to be completed next year" do we want to be more specific? 16:31:46 (in Document Structure ) 16:32:41 pkovar, yeah, should be 2014 16:33:05 yeah 16:33:28 Yes, agreed. But do we want to put concrete dates in the PRD? 16:33:51 I don't think I've seen them in the Server and Cloud PRDs. 16:34:44 hhorak, I saw you changed "goals" to "tasks or goals". Care to explain the rationale? 16:34:49 I wouldn't put dates in the PRD either. That would belong in a schedule. 16:35:19 drieden: makes sense 16:36:32 Should we include what we expect from any of the other working groups? Should this be a separate section? 16:36:54 tjanez: tasks vs. goals were discussed on the last meeting, I understood that we agreed on goals, but wanted to make clear today anyway. 16:37:52 #info schedule will contain dates, not the actual PRD 16:37:58 drieden, I think we don't have any special expectations from other WGs. However, other WG probably have expectations/requirements from us. 16:38:36 hhorak, yes, I remember. That's why I don't understand why you changed it to "Tasks or goals" 16:39:08 Or does it mean we should decide and pick one today? 16:39:28 tjanez okay, I wasn't sure. Thanks! 16:41:50 tjanez: after re-reading some of the items feel more like goals, some more concrete like tasks, so no expression seemed correct for every-thing.. 16:43:29 * mmaslano needs to leave in ten minutes. Could someone end the meeting, when it's finished? 16:44:01 hhorak: Aha, thanks for explaining. The "Tasks or goals" still sounds odd, maybe "Tasks and goals" or "Tasks, Goals" 16:44:29 tjanez: great, that sounds better. 16:44:41 mmaslano: I'll hit \#endmeeting for you 16:44:47 hhorak: The harder way of course is to improve/reword the parts of the PRD that are written to much like concrete tasks 16:45:07 abadger1999: great, thanks 16:45:13 mmaslano, 16:45:40 tjanez: I'd go with easier way :) 16:46:03 Another thing with the PRD that I still don't like are the "Build systems" and "SCL" sections 16:46:18 hhorak, ok :) 16:46:39 tjanez: Build system -> there could be some generic statement - improve bugs, which bothers developers 16:47:02 hhorak mentioned better logging in mock or koji 16:47:18 but I don't we need to actually list all problems 16:47:24 at least not now 16:48:21 Ok, fair enogh. 16:48:30 What about the SCLs? 16:48:38 tjanez: about scl, I do not know, what will happen in Fedora with scl. I'm currently working on upstream of collections, because we need to provide some place for people who already build their own stuff 16:50:11 Should we rather put something like "will cooperate with SCL authors/SIG/... and use them to provide alternative/parallely installable software stacks in Fedora" 16:50:23 tjanez: we might 16:50:28 abadger1999: what do you think? 16:50:41 mmaslano: We should talk about that -- the more I've worked on scls the more I think scls in fedora will only be marginally useful. 16:50:47 abadger1999: I read minutes from last few FPC meeting, but there wasn't any output 16:50:54 need to go 16:50:56 mmaslano: We really want a separate repository of scls... which might be upstream scls 16:50:58 bye guys 16:51:13 or might be a fedora repo that is similar in scope to upstream scls 16:52:32 tjanez: I think scls fall squarely in this WG because I think that we'll have scls in the main fedora repos but those won't be of as much use as scls in a separate repo. 16:53:11 tjanez: For instance, I think we're going to want scls from rhel to be built against fedora. Those won't go into the main fedora repos. 16:53:34 One thing that would be helpful in the PRD is to define what is meant by SCL, in case people reading it don't know what it is. Or provide a link to what the definition is. 16:53:49 tjanez: So I think we want to figure out what rules we want to govern a separate repository of scls (where rebuilds of rhel scls would be allowed). 16:54:07 drieden: see https://fedoraproject.org/wiki/User:Mmaslano/Draft:Env_and_Stack_PRD#Definitions_and_Acronyms 16:54:28 we probably want to move that section though 16:54:51 abadger1999, would you mind improving the SCL part of the PRD to become clearer? 16:55:09 pkovar Oops, sorry I didn't see that. But it might be helpful with more definition. 16:55:24 tjanez: If there's consensus that my idea above is the right way. 16:55:45 pkovar, I put Acronyms at the end to follow the Cloud's PRD 16:56:02 but it might be better to put it under "What this document describes" 16:56:17 Sorry, I meant "Document Purpose and Overview" 16:56:21 tjanez: +1 16:57:27 abadger1999, I don't know enough to make an opinion on that 16:58:04 The guidelines and the main fedora repo are: "these are the rules for scls produced by the fedora project for fedora and epel". This working group is dealing with: "enabling a separate repo to host SCLs that do not follow the guidelines/are not produced by fedora/(something else along those lines)" 16:58:19 tjanez: I don't know enough about what everyone else wants either. 16:58:26 Straw poll, maybe? 16:58:54 Do people who are present agree with this direction for SCLs vis a vis this group? 16:59:28 abadger1999: I see what you are hinting at and I like it 16:59:43 * abadger1999 is unhappily aware that mmaslano and bkabrda are probably the two most important people to get confirmation about this direction, though :-( 17:00:20 abadger1999, maybe put your idea in the PRD and ask for comments from mmaslano and bkabrda on the ML? 17:00:21 yes, we need scl maintainers to comment on this 17:00:50 How about I send that to the mailing list and we see what they say. 17:01:03 abadger1999, +1 17:01:09 abadger1999: sounds good 17:01:38 abadger1999: +1 that makes sense 17:01:53 abadger1999 +1 17:02:03 abadger1999, what about the other way in which the SCLs may interact with our WG, namely developing new features for SCLs (via scl-utils v2 or some other tool)? 17:02:51 Do you feel, this is in-scope, out-of-scope, or "we should cooperate with SCL maintainers on ..." 17:03:29 tjanez: hmm... I read last week's meeting notes as saying we weren't going to be dealing with scl-utils? But I do think that fedora (FPC? this WG?) should have a hand in developing it. 17:03:34 tjanez: we should definitely cooperate at least 17:04:13 scls are implemented as macros for rpm... traditionally, that's purely a distro's perogative to work on that. 17:04:13 abadger1999, hhorak, I agree with you 17:04:44 hhorak: +1 17:04:49 so as a distro we really need to be involved in the design and implementation. 17:05:27 I just don't know precisely which subgroup of fedora that would be. 17:05:39 so invent one 17:05:51 don't shoehorn it into an existing one if none of them fit 17:05:57 jwb: you mean, make a decision, not create yet-another-group, right? 17:05:59 oh 17:06:03 okay, you do mean that. 17:06:20 no, i mean make a group specifically for that purpose that can interact with the other groups 17:06:46 a subsubgroup 17:06:58 * jwb shrugs 17:07:06 jwb, you see Env-and-stacks WG too broad for that? 17:07:12 jwb: Yeah, that may work. Basically we need coordination between the various stakeholders which is something that is lacking right now. 17:07:55 a new group could do that... the danger to be avoided in that would be that the new group would end up being yet-another stakeholder instead of merely aiding in coordinating the existing ones. 17:07:58 tjanez, not necessarily. but if Env-and-stacks themselves don't think it fits then i'm not confident they'll produce the best thing. i'd rather have a group of people that feel they own something work on it 17:08:29 it's really just a matter of "enough interest to be willing to tackle the whole problem" 17:09:12 jwb, fair point 17:09:43 anyway, i'll stop armchair quarterbacking now :) 17:12:31 Okay, anything else? 17:12:51 One thing in the PRD SCL.org takes me to a lawyer website. Is this the correct URL? 17:13:26 I would like someone to take on improving the "Build systems" section. I can help but I need more info 17:14:20 SCL.org should be softwarecollections.org i believe 17:14:26 but the site is not up yet 17:15:21 drieden: good catch, that's only a shortcut of the project, we should probably use SCL Upstream or SoftwareCollections.org to not confuse others 17:15:40 maybe use SCL Upstream for now 17:16:06 hhorak pkovar I'm okay with either. If we reference a URL that isn't up yet, we should just state that in the PRD. 17:16:31 pkovar: +1 17:16:47 drieden: Good point. I like that. 17:17:02 pkovar: +1 17:17:10 Hmm... 17:17:28 I wonder what the legal issues about softwarecollections.org are. 17:18:00 ie: will they be like our COPR repos where RH is responsible and people have to agree not to upload legally questionable material. 17:18:24 If that's the case, we might be able to point to those repositories from Fedora in some way. 17:18:54 drieden: pkovar: changed to SCL Upstream for now 17:18:57 If that's not the case, then we probably need to have our own repositories for software collections. 17:19:26 hhorak: ty 17:21:37 abadger1999: afaict legal issues of softwarecollections.org were discussed and it shouldn't be problem to link to them from fedora. 17:21:58 hhorak: cool. Was that somewhere public that I can link to? 17:22:09 hhorak: because it'll need to land in front of fesco at some point. 17:22:17 so I'll need references to show them. 17:23:57 abadger1999: sorry, I don't have much info about that, nor I'm aware of any public policy. We'll probably learn as soon as it is made public, which should be quite soon. 17:24:14 hhorak, do you have any thoughts on the goals/tasks of "Build systems" section? 17:24:23 hhorak: Okay -- I'll mention it in my message to the list and maybe mmaslano can fill in some details. 17:26:28 tjanez: I have some concrete ideas, like getting more files from broken builds (core dumps), providing information about workers' usage and their resources 17:27:11 The statement with koji - a number of improvements for developers ... I think we should state the improvements, since this seems vague to me. 17:27:38 hhorak, yes, this is what we need :) 17:27:41 drieden +1 17:27:57 tjanez: but I'd rather use something general like "implement new features into build system hierarchy that would give more information to ordinary users and make their work easier" 17:30:12 hhorak, we can put something more general in the PRD, but it is also good that we know some concrete ideas 17:30:25 by we I mean our WG 17:32:38 tjanez: ok, so let's use the general statement and concrete ideas in the brackets. 17:33:54 hhorak +1 17:34:44 hhorak +1 17:35:45 I can be tasked to draft something to improve the "Build systems" section 17:36:26 #action abadger1999 to send out RFC about the SCL direction 17:36:31 tjanez: I just used what I wrote about but feel free to restate it if you want 17:36:39 #action tjanez to draft an improvement to the Build Systems section 17:37:01 #undo 17:37:01 Removing item from minutes: 17:37:13 * abadger1999 waits to see if more drafting is needed :-) 17:42:02 hhorak, thanks for adding it to the PRD already :). I might reword it slightly to make the PRD more cohesive, but its very good already. 17:42:20 * tjanez will need to go in a minute 17:43:03 tjanez: sure, reword it as you wish 17:45:00 Okay, anything else? 17:45:33 abadger1999: nothing from me 17:46:20 abadger1999, also nothing from me 17:46:34 Okay, if nothing else, I'll end the meeting in 60s 17:47:00 thanks for the meeting and working on improving the PRD 17:47:07 30s 17:47:14 20s 17:47:18 thank you, bye! 17:47:24 10 17:47:34 #endmeeting