17:59:12 #startmeeting FESCO (2014-01-29) 17:59:12 Meeting started Wed Jan 29 17:59:12 2014 UTC. The chair is sgallagh. Information about MeetBot at http://wiki.debian.org/MeetBot. 17:59:12 Useful Commands: #action #agreed #halp #info #idea #link #topic. 17:59:16 #meetingname fesco 17:59:16 The meeting name has been set to 'fesco' 17:59:21 #chair abadger1999 mattdm mitr mmaslano notting nirik pjones t8m sgallagh 17:59:21 Current chairs: abadger1999 mattdm mitr mmaslano nirik notting pjones sgallagh t8m 17:59:24 #topic init process 17:59:30 .hellomynameis sgallagh 17:59:31 sgallagh: sgallagh 'Stephen Gallagher' 17:59:36 .hellomynameis kevin 17:59:37 nirik: kevin 'Kevin Fenzi' 18:00:03 Hello 18:00:27 Greetings 18:01:06 * notting is here 18:02:30 oh hey it's that time again. 18:02:42 pjones: Time for wacky hijinks? 18:02:58 i'm around if needed. please ping. 18:03:11 hi 18:03:55 IIRC, mattdm is traveling today. Do we know if t8m will be around? 18:04:32 Let's proceed. 18:04:36 #topic #1197 Procedure for suggesting/approving different Products and/or WGs? 18:04:39 .fesco 1197 18:04:40 sgallagh: #1197 (Procedure for suggesting/approving different Products and/or WGs?) – FESCo - https://fedorahosted.org/fesco/ticket/1197 18:05:09 sgallagh: i believe mattdm said he would not be around 18:06:49 yeah we were waiting on a draft here... 18:06:50 Ok, without mattdm shall we defer this discussion another week? 18:07:02 I know that the KDE folks are getting a bit antsy about it. (With good reason) 18:07:06 related to this... 18:07:27 is it expected that spins will use this new process in the fedora.next world? or spins stay the same? or unknown? 18:08:03 I think we really need to differentiate spins from additional products 18:08:20 If we kept spins and products separate then spins would likely be second class citizens compared to products 18:08:30 Yes, that was kind of the point I was making 18:08:37 "class" is so... not important 18:08:39 IMHO, the spins process isn't very good currently. 18:08:49 If people care enough (and want to put in the work) to raise a spin to a product, then go ahead 18:08:56 it's a bit better in f20 than it was, but still not great. 18:09:03 If we removed spins and only had new products then I think the bar would be raised somewhat on what spins are. 18:09:26 * nirik nods. 18:09:29 "spnis" (i.e. different ISO with different package set, but not other modifications), should be easier to create than the full WG process - if anybody is interested in creating them, that is 18:09:45 mitr: Not according to all the flame wars about what spins get listed on the get.fedoraproject.org page, etc... 18:10:12 abadger1999: We'll have exactly the same flame wars about products; the differentiation is default/not, not spin/product 18:10:47 mitr: only if some products are second class to other products... So far, I haven't seen that we have that differentiation. 18:10:54 there are two sets of spins - different desktops and packages on top of desktops to server some purpose... maybe the first class should be product and for second let's go back to prepare some support to put it on top of products 18:11:18 abadger1999: There's already a talk of a KDE product, so we'll have the "Desktop vs. GNOME" discussion again it seems 18:11:36 mitr: only if we promote one above the other. 18:11:39 s/Desktop/Workstation/ 18:11:47 mitr: oh I see what you mean. 18:11:51 Naming of the Product. 18:12:31 It would be ideal if the two products could select different target audiences, but that might be asking for a unicorn 18:12:40 mitr: abadger1999: only if it goes past "talk of" 18:12:42 targeting a unicorn? 18:13:00 Unicorn Workstation: now with more glitter! 18:13:03 nirik: i would suspect the spins process can/should evolve, but it might need entirely redone if changes come to how repos are made 18:13:10 if spins want to pull packages out of coprs 18:13:10 etc 18:13:32 pjones: right... I'm rather tempted to oppose the creation of a KDE product (because it is technology-specific, not audience-specific), but I suspect that we can't really stop what a community wants to do any more than we can make a community do what they don't want to 18:14:16 mitr: that's the wrong question, though. it's never "what does $community want to do", it's "is $community's result fedora" 18:14:27 sure. The bad things about the current process are: it depends on a spins wrangler (single point of falure), until f20 there was no incentive to test once you were approved, and the bar to entry is high, but the bar to stay is very very very low. 18:14:33 pjones: true 18:14:38 Anyway... defer for a proposal? 18:14:46 +1 to defer 18:14:48 mitr: just doing it does not make it being part of fedora a fait accompli 18:14:54 yes, +1 to defer. 18:15:03 mitr: I think if we disallow that we're going to be defeating a primary advantage of multiple products... giving people a little extra room to express their desires for the distribution they want to create. 18:15:17 abadger1999: I think that's something we'll have to wait for the proposal to really judge 18:15:30 sure, defer, but we need to address this before long, so everyone do put on your thinking caps and come up with proposals. we don't need to wait for mattdm to make just one. 18:16:28 abadger1999: That goes back to the question by pjones - are we facilitating a community to work on whatever, or are we creating specific products that are useful _as products_? 18:16:47 mitr: I think we need to do both 18:17:07 sgallagh: I think so too, but I think it's also important that one goal is /in service of/ the other. 18:17:10 We need to have something we show off ("Products") while making sure we're the place people also want to try something new. 18:17:34 products are not new 18:17:47 in the sense that people aren't going to be like "OMG a product??" 18:17:48 products are not (necessarily) the mechanism for "a place to try something new" 18:17:48 mitr: Sure, then I'm definitely for the first choice. with the clarification that I'd imagine the majority of the products are useful to at least a few ;-) 18:17:50 Well, work done by the spins will hopefully be absorbed into products when they make sense too 18:18:32 products are a focus mechanism. they aren't an innovation catalyst outside of providing something stable for people to work from 18:18:46 jwb: Very well put. Thanks. 18:18:49 jwb: +1 18:19:29 jwb: yeah, that's pretty reasonable. 18:19:48 right 18:20:39 to narrow the field a bit... are we leaning towards spins and products being separate deliverables or having spins be something that feeds into/merges with the products? 18:21:19 eventually, do we want a spins process and a products process that operate independently of each other? 18:21:19 or a hybrid... a 'spins' product. ;) which makes all the spins. 18:21:20 The latter certainly presents a simpler message to the end consumers 18:21:44 I'm not sure they can't be both. Hypothetically, some spins might merge into a Product if they make sense there. 18:21:46 nirik: right round, baby, right round. 18:21:57 ha 18:22:24 spins will exist to feed into products, and spins will exist to be direct counter-points to products 18:22:25 sgallagh: Right ... but do we want to make that the expectation? You start off with a spin, gain critical mass, and then get promoted to product when you're ready? 18:22:34 So spins would be like secondary arches. 18:22:45 wait, "feed into" is entirely different than "promoted to" 18:22:49 abadger1999: Well, not necessarily. 18:22:53 jwb: to me that sounds incredibly confusing to our users 18:22:53 Using "spin" for "secondary product" is strange 18:23:01 sgallagh: sure - this is one example of how it could look 18:23:08 pjones, indeed! yet you can't prevent either without excluding the other 18:23:13 i.e. a Scientific spin could become part of the standard Workstation spin 18:23:26 perhaps you could make it less confusing by just making spins something the project does not do at all. 18:23:28 It wouldn't necessarily need to be a separate product. 18:23:48 sgallagh: What I'm asking is -- do we want to move in a direction that pulls spins and products closer together or do we want to move towards having them be entirely separate. 18:24:03 I'd rather see them be closer together. 18:24:37 I'd prefer separate - continue to make it fairly easy to create spins (== rel-eng + QA work mostly) but set the bar higher for products (establishing new infrastructure, design/integration work) 18:24:37 For one thing, I agree with jwb: ultimately (not immediately), I'd like to see them not be separate install media but instead a reasonable box to check in the installation of a product. 18:24:38 * abadger1999 leans that way as well. 18:24:47 Err -- I lean the same way as sgallagh 18:25:20 that's not exactly what i said, but sure that could be an eventuality. it's analogous to the server products "roles" in a way 18:25:30 abadger1999: "i don't know yet?" there are both more product-like spins that are designed for a specific audience (security, design suite, FEL, etc.) and ... technology-based spins ("i want my desktop to be X") 18:25:40 mitr: I'm not sure rel-eng and qa would agree that that's a feasible setup 18:26:01 pjones: Isn't that exactly what spins are today? 18:26:08 mitr: yes and they hate it? 18:26:16 i'd really prefer spins to be done on the fringe entirely, and not have fedora qa/rel-eng worry about them at all 18:26:32 they already seem overloaded with product stuff 18:26:55 jwb: likewise 18:27:20 Perhaps go a similar route as the *ubuntus (Kubuntu, Xubuntu, etc.) and just help them create a separate Project that is allowed to use our repos? 18:27:23 basically, i'm saying spins die as a concept and anything that isn't a product becomes a Remix 18:27:36 well.. 18:27:47 jwb: which still leaves the need for a "how do i create a new Product" proposal 18:28:01 notting, yes! but it gets rid of spins 18:28:09 jwb: I'd be okay with keeping the trademark/logo differentiation of spins vs remixes but in terms of creation, moving everything to the remix model. 18:28:23 sgallagh: "get your own damned web page" is certainly one way to answer "how do we decide what goes on the web page" 18:28:42 pjones: Arguably, that's the only reasonable answer... 18:28:53 But that is probably a Board decision\ 18:29:00 wait, what? 18:29:04 spins were created by fesco 18:29:06 it's certainly the easiest to scale up. 18:29:14 yeah, not sure how the board is involved there 18:29:27 jwb: I meant "deciding what goes on fedoraproject.org" 18:29:40 well, if you eliminate spins then you don't have to worry about taht either 18:29:40 It's a marketing/positioning decision. 18:30:07 i am of the opinion that we have enough of a marketing/positioning hurdle ahead of us with just products 18:30:08 I think that's only true if we choose to structure it that way 18:30:24 if we say "hey, do this thing on your own and use our repos", there's no marketing decision because it's not us doing it. 18:30:36 anyway, i'm just repeating myself now. i'll be quiet. particularly since i'm not even in fesco... 18:30:55 jwb: FESCo meetings aren't closed... 18:31:03 yes, but i'm repeating myself :) 18:31:18 I think nuking spins entirely might generate a lot of ill will, but I don't think we should decide right now. 18:31:23 pjones: Yeah, that's pretty much what I'm saying 18:31:52 nirik, perhaps. but the problems you've already highlighted are generating poor quality spins and ill will on the qa/rel-eng side 18:32:22 nirik, so i'm less inclined to be worried about the reaction of people taht are putting in minimal effort and expecting the project to carry the burden 18:32:24 yep. We did much better with f20 on qa tho... forcing signoff on each spin... 18:32:24 So I'm seeing several people in favor of eventually dropping spins and one person in favor of keeping them as entirely separate deliverables and a lot of... we don't want to make a decision yet. 18:32:57 Do we want to attempt to make any kind of resolution there? 18:33:09 Or maybe punt it to the next FESCo, since we're in the middle of election season. 18:33:31 I personally didn't come into the meeting having given this topic much thought... so I am -1 to making some decision right now. 18:33:32 sgallagh: maybe -- when do we want a resolution by? In time for F21 or not until later? 18:33:39 i think we're waiting on proposals/writeups in either case 18:33:53 have you even decided if F21 is products or not? 18:33:58 abadger1999: surely if we're doing something that affects Spins As We Know it, we need it by f21 18:33:59 sgallagh: +1 to punt for reasons nirik and notting said. 18:34:01 I think we repeated what we said last time, -1 to any decision now 18:34:19 jwb: Pretty sure f21 is products... I remember notting saying something along that line and everyone agreeing. 18:34:23 yeah, I think we do need to decide along with the product deliverables... 18:34:35 abadger1999, oh good. that memo was missed 18:34:52 abadger1999: I believe that the decision was "F21 is products unless it becomes obvious during the scoping that it's impossible" 18:34:59 ie, if we are making spins or something like them that should calculate into the needed resources for f21. 18:35:06 possibly missed because "nobody disagreeing with notting" doesn't sound like a strong decision 18:35:08 (Note: during scoping means "by mid-February") 18:35:28 sgallagh, impossible by when? 18:35:39 because it's always possible as long as you give yourself enough time 18:35:44 jwb: agreed. and sgallagh, for lack of a better memory on my part, that works for me 18:36:19 jwb: We want to keep as close to the current August date as possible. Adding a month might be reasonable, but if we determine it's impossible before 2015, then maybe we wait to F22. 18:36:33 Actually, s/we/I/. I shouldn't speak for FESCo there. 18:36:55 sgallagh, erh... i thought it was "it won't ship _before_ August". that doesn't mean "it will ship in August" 18:36:56 (only the first "we"...) 18:37:12 jwb: I don't think I contradicted that. 18:37:12 jwb: that's my understanding as well. 18:37:36 sgallagh, but that isn't a "current August date" 18:37:38 I meant we want to avoid being much later than that without reason, and if it's looking to be a LOT later, we change plans. 18:37:39 sgallagh: doing it in september is not "adding a month"; it's still consistent with our current statements. 18:37:45 jwb: Then I misspoke 18:37:51 * notting rings the "arbitrary, but certainly more than 15 minutes" bell 18:38:08 you aren't giving people a target. you are giving them an expectation that there won't be something before then, but not when there will be something after 18:38:15 * sgallagh sighs 18:38:23 Proposal: give this topic a week for people to come up with proposals 18:38:26 Let's start a discussion on DEVEL 18:38:42 pjones: +1 18:38:43 that can totally be one way to help come up with proposals! 18:38:54 pjones: -- I agree with you and jwb about the current schedule. 18:38:56 sgallagh: okay 18:38:57 pjones: Works for me. 18:39:10 pjones: +1 to proposal to defer and seek proposals 18:39:16 pjones: aye, +1 18:39:19 pjones: +1 18:39:50 * pjones +1 18:39:54 that's +5 so far. 18:40:08 Good enough for me. 18:40:24 #agreed Give this topic a week for people to come up with proposals (+5, 0, -0) 18:40:40 #action sgallagh to start discussion on mailing list about Spins vs. Products. 18:40:57 Moving on 18:41:01 #topic #956 Need to audit packageset for bundling of libiberty 18:41:04 .fesco 956 18:41:05 sgallagh: #956 (Need to audit packageset for bundling of libiberty) – FESCo - https://fedorahosted.org/fesco/ticket/956 18:41:51 abadger1999: Anything to do here, or defer a week for you to look at this again? 18:42:11 I've almost got all the bugs filed (doing last one now) 18:42:19 ah, cool 18:42:24 There's probably nothing further for fesco to do here? 18:42:28 Works for me 18:42:50 Excellent. then we can close. 18:42:59 abadger1999: we could say "iberty" a few more times, if it'd help. 18:43:13 *shudder* 18:43:17 #info abadger1999 has nearly finished filing bugs. Nothing more for FESCo to do. Close ticket. 18:43:24 #topic #1117 Generalize policy about privilege escalation and Administrator user accounts 18:43:27 .fesco 1117 18:43:28 sgallagh: #1117 (Generalize policy about privilege escalation and Administrator user accounts) – FESCo - https://fedorahosted.org/fesco/ticket/1117 18:43:56 Proposal: Since there is nobody with sufficient interest to generalize, and there's nothing really wrong about the current status, close the ticket == stop tracking this 18:44:25 Maybe we should ping halfie 18:44:26 Counter-proposal: Close this ticket and request that the Workstation WG address the issue 18:44:34 sgallagh: +1 18:44:52 sgallagh: we can't actually just /reassign/ it to them, right? 18:44:57 jwb: ^^ in case you're elsewhere. 18:45:31 well, this could affect cloud/server too? 18:45:32 sgallagh: -1 18:45:35 so fesco has been unable to do anything for 5 months, and they expect a WG that doesnt' even have an approved PRD to somehow fix this? 18:45:36 or what exactly were we asking? 18:45:50 uh... i guess if World Class Punting is what you're going for, sure 18:45:53 If we want to push it to a WG we should probably push it to Base desgin. 18:46:52 Maybe I'm misremembering. I thought this was mainly about PackageKit. 18:46:59 I'd be ok with asking base to look, or just closing and asking interested people to file a new ticket. 18:47:26 Ah, right sorry. This is the general escalation case. I need to read more carefully 18:47:42 This probably *is* a FESCo level decision then 18:47:44 The initial ticket was about software installation. But this ticket is because we were wondering if we should have some overarching policy rather than taking things case-by-case 18:48:21 I'm +1 to mitr's proposal. 18:48:36 nirik: Yeah, I'd either be for pining halfie on the ticket or closing and asking people to re-open. 18:48:39 Right now, "default to NO" seems to work pretty well 18:49:01 The default right now is that things requiring more privilege should come to fesco to be decided. 18:49:12 abadger1999: i'm fine with closing and leaving that 18:49:17 as am I 18:49:23 yeah, +1 to mitr's proposal as well. 18:49:31 mitr: +1 18:49:45 If somebody happens to have a generalized proposal, that'd be cool, but until somebody thinks that's worth their energy, one-offs will do. 18:50:14 assuming mitr's okay with it, that's +4 so far. 18:50:39 +1 mitr 18:51:11 * mitr is +1 for the record 18:51:39 #info The default right now is that things requiring more privilege should come to fesco to be decided. 18:51:58 Any other votes? We're at +5 18:52:06 +1 18:52:36 notting: ? 18:52:49 sgallagh: sorry, didn't add a clear +1. yes. 18:53:00 #agreed Since there is nobody with sufficient interest to generalize, and there's nothing really wrong about the current status, close the ticket == stop tracking this (+7, 0, -0) 18:53:06 #topic #1226 Workstation PRD for approval 18:53:09 .fesco 1226 18:53:10 sgallagh: #1226 (Workstation PRD for approval) – FESCo - https://fedorahosted.org/fesco/ticket/1226 18:53:19 We (FESCo) dropped the ball on asking questions in time 18:53:48 Proposal: Give Workstation WG $MORETIME to address concerns raised today. 18:54:17 sure. FWIW, I read the thing friday and thought it looked fine to me, I should have noted that in ticket. 18:54:28 sgallagh: that does appear to be what happened,. 18:54:42 jwb: have you an opinion? 18:54:46 Last week's concerns have been, IMHO, adequately resolved by the recent edits to the PRD. 18:54:49 the edited bits seem ok to me so far 18:55:11 I haven't seen the edits yet. 18:55:13 * sgallagh goes to look 18:55:17 I haven't had time to go through the very recent discussion in detail; so far it does seem to be veering towards FESCo being too controlling. 18:55:35 i have an opinion. i haven't seen anything additonal beyond my 4 questions that i feel needs addressing. 18:55:54 i would note that the edit said 'leverage the existing image/repo delivery mechanisms', but the list discussion implied shooting at least one of the delivery mechanisms in the head 18:56:12 notting, it didn't say leverage ALL of them 18:56:21 (I _do_ want to keep using the Change process for WG outputs, though - the "self-contained" version at least, to keep "the packager universe" aware) 18:57:12 Ok, so I've read the edits that were made and I'm reasonably comfortable with them. 18:57:20 If we wanted to call a vote...? 18:57:47 guess you'd say something like 18:57:49 * nirik hasn't had time to look at the edits, but if others are ok, fine. 18:57:53 Proposal: Approve the current version of the Workstation PRD 18:57:56 Proposal: vote to accept the PRd as it stands 18:57:59 or that, yeah 18:58:02 sgallagh: +1 18:58:13 sgallagh: +1 18:58:30 (I realized after my previous statement that the worst-case was just having it voted down, so I went ahead) 18:58:37 -1 18:58:39 +1 18:59:07 +1 18:59:19 +1 18:59:25 nirik: Can I assume you're 0? 19:00:09 yes, 0 19:00:17 My concern about the boundary between WG responsibilities and fesco responsibilities isn't adressed. 19:00:17 #agreed Approve the current version of the Workstation PRD (+5, 1, -1) 19:00:19 currently +5:-1:0 19:00:28 er 19:00:29 1 19:00:29 yeah 19:00:29 you've got it right 19:00:49 abadger1999: As noted in the document, it's permeable. Please continue discussing that. 19:00:55 19:01:02 The principal goals are established, which IMHO is the important part. 19:01:05 sgallagh: No worries -- Just clarifying why I voted -1 19:01:14 abadger1999: Perfectly reasonable. 19:01:17 abadger1999, we opened a ticket to discuss exactly that and you (as a group) closed it saying it was too generic 19:01:29 so really, just make up your damn minds 19:01:39 i am beyond frustrated at the moment 19:01:45 jwb: it is approved. 19:02:00 thank you for your time. i'm dropping off for now. 19:02:06 jwb: Right -- and the wokstation PRD is doing several of those things that we said should be brought up to feso on a case by case basis 19:02:15 that poorly phrased... 19:02:26 i disagree. you can raise objections with the WG on the WG list. 19:02:31 fesco said they would have to decide those things on a case-by-case basis 19:02:46 The workstation prd is treading that line in several cases. 19:03:09 but given that the WG haven't fucking done anythign at all yet, what i really think is unprintable. 19:03:53 that's actually a good thing - coming as close to crossing that line as possible is synonymous with "using all the discretion they've got" 19:04:14 pjones: But the question is whether the line is crossed or not. 19:04:27 Can we stop speaking generically? 19:04:40 I see that, but at the same time, let them actually do some work, you know? 19:04:43 abadger1999 raised some points in the FESCo ticket. Let's look at each of those individually (later) 19:05:09 sgallagh: that works for me. 19:05:18 For right now, I'd prefer to trust that the WGs aren't secretly plotting to undermine Fedora :) 19:05:45 Except Fedora Server. Those guys are just evil ;-) 19:06:08 We have the liaison reports to minimize secret plotting 19:06:31 Ok, let's move on. 19:06:38 We have no new business this week (for a change) 19:06:49 #topic Next week's chair 19:07:04 I am probably not going to be around next week due to DevConf-related events. 19:07:15 I suspect the same can be said for others of the group. 19:07:33 (Also, depending on travel weather, I may or may not be around for the following week) 19:07:47 are there enough missing that we should simply cancel the meeting? 19:07:48 * mmaslano too 19:08:10 note that that de facto means our punted agenda item goes two weeks rather than one 19:08:16 abadger1999: That's what I was trying to figure out. 19:08:20 * pjones will be available for the meeting. 19:08:22 * notting will be travelling the next two weeks. but if mmaslano/mitr/etc. can make this time every week, i suspect i can too (unless we all get dragged into something) 19:08:23 I still have no idea about my next week's schedule I'm afraid 19:08:37 Proposal: skip next week's meeting 19:08:38 * nirik should be available 19:08:52 (there's just too much uncertainty, and I'd rather you guys have a productive conference) 19:09:15 * nirik is ok with that, but we need a chair for the meeting after that. ;) 19:09:29 Proposal: plan to have the meeting and cancel if no quorum. 19:09:38 (Yes, I realize I'm restating the status quo) 19:10:01 sgallagh: I'd rather not - it's a very busy time for a lot of us, and knowing our week's schedule up front really helps 19:10:16 I'm happy either way. 19:10:49 We don't have any expected urgent business, so +1 to canceling 19:11:18 pjones: +1 19:11:41 sure, +1 (but note we still need a chair for feb 12th) 19:11:59 pjones: +1 19:12:03 +1 for cancel 19:12:03 nirik: Yes, clearly. 19:12:49 I see +5. 19:13:04 0 19:13:33 pjones: You're implicit +1? 19:13:35 yes 19:13:38 * pjones +1 19:13:38 #agreed skip next week's (Feb 5) meeting (+6, 1, -0) 19:13:56 Ok, so who wants to chair on Feb 12? 19:14:24 * nirik did last meeting. I guess I could again, but... 19:14:32 sgallagh: wants is such a nice word ;-) 19:14:35 sgallagh: I'll volunteer 19:14:48 #info abadger1999 to chair Feb 12 FESCo meeting 19:14:52 #topic Open Floor 19:15:58 Anyone have anything for Open Floor? 19:17:52 Closing the meeting in one minute, then 19:19:33 #endmeeting