17:01:54 <stickster> #startmeeting
17:01:54 <zodbot> Meeting started Thu Feb  4 17:01:54 2010 UTC.  The chair is stickster. Information about MeetBot at http://wiki.debian.org/MeetBot.
17:01:56 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
17:01:59 <stickster> #topic Roll call!
17:02:40 <stickster> sound off
17:02:44 <mmcgrath> off
17:02:46 <stickster> heh
17:03:07 * poelcat here
17:03:49 <stickster> I see ctyler, dgilmore, mdomsch, and walters_ also
17:03:52 <walters_> hello everyone
17:04:22 * ctyler here
17:05:05 <stickster> spot will not make it today -- as hopefully everyone knows by now, he and his wife just had their first child this past weekend
17:06:01 <ctyler> *good thing* he's not here, then!
17:06:09 <stickster> dgilmore mdomsch Are you guys here too?
17:06:10 * dgilmore hopes he is not here
17:06:12 <stickster> ha
17:06:14 <dgilmore> stickster: i am
17:06:17 <stickster> Good on ya mate
17:06:54 * mdomsch is here
17:07:02 <stickster> OK, it's 12:05. I want to proceed with one item for a maximum of 15 minutes, and then we'll go to Q&A
17:07:12 <stickster> #topic regional localized spins
17:07:29 <stickster> #link http://lists.fedoraproject.org/pipermail/advisory-board/2010-February/007900.html
17:08:06 <mmcgrath> I generally think "meh, why bother?" with this but realize that I don't have this problem so it clouds my judgement.
17:08:14 * dgilmore thinks that they are fine and necessary.  but i think that teams in the region they are intended for should be responible for producing and distributing them
17:08:17 <stickster> I think the questions are really:
17:08:52 <walters_> mmcgrath: yeah i agree on the latter
17:08:58 <stickster> 1.  Can we give a special blanket approval to spins that only change default language/regional settings?
17:09:02 <jwb> apologies for being tardy.  sick kids
17:09:20 <stickster> jwb: No problem
17:09:27 <stickster> 2.  Where should those spins be hosted?
17:10:10 <mdomsch> 3.  who makes them?  rel-eng or someone else?
17:10:19 <mmcgrath> stickster: one question I'd ask are these spins a convenience or is there really something inadequate with the anaconda language selection?
17:10:31 <mdomsch> and can we solve it differently?
17:10:45 <stickster> mdomsch: I would say that if the answer to (2) is not on our official space, then the answer to (3) is not rel-eng.
17:10:48 <mdomsch> (note that opensuse & ubuntu use a translated grub screen to start the install)
17:10:55 <jwb> stickster, agreed
17:11:11 <mmcgrath> mdomsch: how does that work?
17:11:19 <mdomsch> 2) I suspect we _can_ host them on FI space, just not the mirrors
17:11:20 <dgilmore> mmcgrath: big issue is that the spinds being livecds  boot into english.  if yo dont know any english then you might have trouble to change it to your native language.  and it has nothing to do with anaconda
17:11:31 <ctyler> On (1), two questions arise from the word "only" -- (a) what is the base for the comparison? The Desktop spin? and (b) what if there are some packages that are needed because of regional settings, e.g., for input methods?
17:11:56 <mdomsch> mmcgrath, not sure exactly, but the grub boot screen is translated IIRC
17:12:21 <mmcgrath> how have the other distros solved this problem?
17:12:27 <dgilmore> mdomsch: at least some of the use cases i can think of finding local hosting inside the target countries is financially better as international links can be very expensive
17:12:30 <stickster> dgilmore: mmcgrath: The other question is which fonts and other input method related material is on the Live spin by default.
17:12:49 <stickster> If there's not support for your language, then I'd imagine a new spin is very important to you when you go to local community groups handing out media.
17:12:54 <dgilmore> stickster: right some spinds will need to add fonts/input methods
17:13:09 <walters_> that's the part that concerns me more
17:13:17 <ctyler> there's actually a fair bit on the default spin, but I'm sure it's not exhaustive
17:14:21 * poelcat thinks we are getting off track... answer one question at a time? :)
17:14:39 <walters_> yeah agreed...
17:14:45 <jwb> personally, i think question 1 is the only board-level question
17:14:52 <mmcgrath> poelcat: yeah, perhaps we should put it this way: "Is anyone against us allowing language spins, but making the different local groups create and distribute them?"
17:15:14 <dgilmore> mmcgrath: im all for that
17:15:20 <mdomsch> mmcgrath, +1
17:15:22 <poelcat> +1
17:15:22 <jwb> i'm for that as well
17:15:26 <walters_> i'm not against it
17:15:56 <mmcgrath> then I say let them go forth and spin
17:16:02 <poelcat> what are the remaining *board* issues to discuss on this?
17:16:12 <stickster> I have no problem with regional groups making minimally altered spins for use in their locale, but I think those spins *must* still go through the Spins SIG
17:16:30 <jwb> stickster, agreed
17:16:34 <ctyler> just to clarify, they can alter package sets etc. in addition to local settings, with no Q&A?
17:16:35 <stickster> In other words, we don't have a concern about trademarks, as long as the Spins SIG feels the changes are minimal, to support local language.
17:16:46 <ctyler> stickster: +1
17:16:46 <poelcat> stickster: agreed as well... use existing established processes
17:16:56 <dgilmore> +1 from me
17:16:57 <stickster> ctyler: I would limit that only to required fonts or something directly in support of the local language
17:17:35 <dgilmore> ctyler: fonts and input methods
17:17:37 <stickster> And when I say "limit," note that I mean "limiting the blanket approval," not "limit what people can remix in general"
17:18:06 <ctyler> getting the Spins SIG to approve solves that -- we can give a blanket trademark approval and the Spins SIG can decide how much can change and whether there's a QA issue
17:18:27 <walters_> ctyler: sounds reasonable
17:18:49 <mdomsch> yep - each still needs to go through the Spins SIG
17:19:24 * poelcat notes 3 minutes left on this topic
17:20:01 <stickster> So the answer to 1. is "Yes, the Board will give a blanket TM approval for regionally-produced spins, provided that
17:20:46 <stickster> a. The changes are only those required to serve language and font settings, including font packages, and
17:21:06 <stickster> b. The Spins SIG approves the spins on a technical level to assure (a).
17:21:07 <stickster> "
17:21:12 <stickster> Is that correct?
17:21:15 <ctyler> +1
17:21:40 <mdomsch> +1
17:21:41 <dgilmore> stickster: yep
17:21:42 <poelcat> +1
17:21:42 <dgilmore> +1
17:21:42 <jwb> yes
17:21:45 <walters_> +1
17:21:50 <stickster> OK, thanks everyone.
17:22:06 <mmcgrath> +1 sorry
17:22:10 <mmcgrath> that went to a pm on accident :)
17:22:12 <stickster> Now we are at the end of 15 minutes and there seems to be a consensus that the Fedora Project is not required to host these spins
17:22:25 <jwb> well
17:22:28 <stickster> But no clear decision.
17:22:42 <jwb> i think there is consensus that it's not a Board level issue
17:22:46 <jwb> :)
17:22:47 <mdomsch> not required to - I agree; may be able to though.
17:22:56 <mdomsch> right, that's FI
17:23:15 * stickster can clear (2) off the agenda then
17:23:20 <stickster> Regional
17:23:23 <stickster> oops
17:23:53 <stickster> Give me a second, copy/paste slowness
17:24:10 <stickster> #agreed , the Board will give a blanket TM approval for regionally-produced spins, provided that (a) The changes are only those required to serve language and font settings, including font packages, and (b)  The Spins SIG approves the spins on a technical level to assure (a)
17:24:49 <stickster> #agreed The Board leaves the question of hosting to any regional creators to work out with the Infratructure and rel-eng teams
17:25:04 <stickster> Shall we move on?
17:25:09 <dgilmore> please
17:25:21 <stickster> #info caillon cannot make it today and sends his regrets
17:25:26 <stickster> #topic Community Q&A
17:25:39 <ctyler> wait, did we address (3)?
17:25:44 <ctyler> do we need to?
17:25:59 <walters_> ctyler: the locale community does
17:26:02 <jwb> ctyler, we did.  2 and 3 are up to FI and rel-eng
17:26:07 <mdomsch> 3) is a Spins SIG <-> rel-eng problem IMHO
17:26:11 <jwb> right
17:26:17 <stickster> right.
17:26:30 <mdomsch> at such a point when that can't be resolved, they'll kick it up
17:26:33 <stickster> right.
17:26:39 <ctyler> sounds reasonable, just wanted to ensure we were in agreement
17:27:21 <stickster> #agreed The Board leaves question about production to regional creators to work out with Spins SIG and rel-eng
17:27:52 * stickster awaits questions in other channel.
17:28:01 <stickster> We'll leave the line open for a few minutes
17:28:07 <stickster> jwb: Go ahead, you raised a hand over there.
17:28:38 <jwb> Has everyone seen the thread on the devel list about the target audience?
17:28:49 <stickster> Has anyone *not*?
17:28:58 <jwb> right :)
17:29:13 <jwb> so something that has come up there is the idea behind conflicting package changes
17:29:25 <jwb> we've disallowed this basically forever
17:29:44 <jwb> is that something we should revisit in light of allowing more flexible remixing?
17:29:59 <stickster> So the question is, should we allow conflicting package changes?
17:30:15 <jwb> yes, basically
17:30:30 <mmcgrath> jwb: FESCo decided on that question specifically or are they looking at things one conflict at a time?
17:30:32 <ctyler> so what would that look like, in practical terms?
17:30:34 <walters_> jwb: i'd have to scroll through the thread again, but concretely do you mean Conflicts: or multiple shared library versions?
17:30:35 * mmcgrath forgets the history on that.
17:30:55 <walters_> both right?
17:31:06 <dgilmore> jwb: i think its an issue to be taken up with FESCo not the board
17:31:08 <jwb> walters_, i think both
17:31:23 <poelcat> dgilmore: agreed
17:31:25 <mdomsch> "but I _don't want to link against libnss" ...
17:31:44 <jwb> dgilmore, in how it is implemented, possibly.  i'm wondering if the higher level goal behind it is something the Board needs to look at
17:32:11 <dgilmore> jwb: its a technical issue hence its a FESCo issue
17:32:28 <dgilmore> if FESCo asks for assisatnce or guidence then we should look at it
17:32:37 <jwb> fine.  retracted
17:33:25 <poelcat> jwb: did you feel there was anything else for the board to discus around the devel thread?
17:33:32 <mdomsch> I think it's a fair question to ask - how often have such "us vs them" conflicts arisen?  if many, the board may wish to provide some additional guidance
17:33:40 <stickster> There are some implications for people who end up on the receiving end of workload, like packagers, developers, and bug triagers. But it wouldn't work for us to start there before someone goes through the technical aspects.
17:33:45 <jwb> poelcat, yes, i just asked exactly what i felt needed to be asked
17:34:33 <jwb> i'm ok being wrong about it being a Board level issue
17:34:38 <walters_> well we'll definitely be taking up the feedback during the next SWG meeting, right?
17:34:47 <mmcgrath> mdomsch: it's funny because those things are rare, but increasingly I think people think that's what the boards job is... and little else.
17:35:00 <stickster> walters_: I would hope so
17:35:08 <dgilmore> mdomsch: the whole surf conflict recently was the first i rember
17:35:22 <jwb> dgilmore, i'm not talking about package names
17:35:42 <stickster> If we're agreeing this is a FESCo question, we should move on. There are others in the queue.
17:35:47 <jwb> yes, move on
17:36:03 <stickster> Strangely, the next question is also from a Board member -- go ahead mdomsch
17:36:07 <stickster> :-)
17:36:53 * mdomsch started a thread on advisory-board
17:36:58 <stickster> #info Board does not take up the question of conflicting packages at this time
17:37:05 * stickster making better use of zodbot, sorry.
17:37:20 <mdomsch> soliciting feedback.  Basically, I want to have yum provide a uuid in mirrorlist requests, to improve our ability to count the number of Fedora installations worldwide
17:37:24 <mdomsch> and over time
17:37:53 <mdomsch> our "unique IP address" metric is undercounting by, at my guess, at least 2x, probably closer to 6x
17:38:10 <dgilmore> mdomsch: only if its opt-in
17:38:23 <mdomsch> in my proposal, it's opt-out
17:38:29 <jwb> mdomsch, i'm concerned about opt-out
17:38:34 <mdomsch> smolt is opt-in
17:38:54 <mdomsch> opt-in will leave us with the status quo
17:38:55 <jwb> there's not great way to do opt-out without some kind of 'firstboot' like dialog for yum/PK
17:38:56 <mdomsch> at best
17:39:05 <dgilmore> mdomsch: opt-ot is not acceptable
17:39:09 <dgilmore> opt-out
17:39:16 <dgilmore> mdomsch: it has to be opt-in
17:39:25 <poelcat> dgilmore: why?
17:39:36 <ctyler> dgilmore: why? no trackable info here, really
17:39:55 <stickster> I think we should get expert legal advice on what these sorts of records mean, both here and other places like the EU
17:40:07 <jwb> ctyler, that's not entirely accurate.  it's trackable, but it's certainly not _easy_ to track an individual
17:40:30 <mdomsch> stickster, agreed
17:40:32 <walters_> stickster: yeah i don't feel i have a solid grasp of what standard practice is here
17:40:41 <stickster> How would you track a UUID back to a person?
17:40:50 <dgilmore> ctyler: it would be possible to track.  not trivially so but it would be possible
17:40:54 <stickster> You could trace it to an IP as I understand the problem mdomsch set out
17:40:57 <ctyler> mdomsch: what were you intending to record? just a DB of UUIDs that have been seen before?
17:41:24 <stickster> Then you *probably* could check other requests from that IP around the same time to see what was happening, i.e. wiki edits and from whom they came
17:41:49 <stickster> In order to do that, you would have to be somewhat of a trusted contributor yourself
17:41:52 <jwb> this is a slight derailment now...
17:41:58 <stickster> jwb: Yes, sorry.
17:41:59 <dgilmore> stickster: you can also track the different ips used by that UUId and work out travel patterns etc
17:42:23 <stickster> mdomsch: How about you and I round up with the right legal folks to ask specific questions?
17:42:24 <dgilmore> I personally feel it needs to be opt-in
17:42:25 <walters_> dgilmore: "you" being a mirror server?
17:42:34 <mdomsch> stickster, yes please
17:42:37 <stickster> dgilmore: I think opt-in makes it just as useless as what we have now, and thus pointless.
17:42:38 <mmcgrath> I personally don't care if the uuid gets added, but I find myself being paranoid for people before they even raise a stink about it because of what I went through with smolt.
17:43:05 <stickster> mmcgrath: Understood, I agree it's a concern. I don't want to see us sink a bunch of effort into something that's not going to get us better information.
17:43:18 <mdomsch> there are 2 standards here; 1) legal; 2) what our community expects
17:43:30 <mdomsch> we need to address 1) with the right people
17:43:40 <mdomsch> 2) we need to address as well, hopefully via fab
17:43:48 <walters_> i guess i feel what we need to do in the big picture is strongly encourage people to interact with fedora
17:43:49 <stickster> With 1) we can do 2) more readily.
17:43:56 <mdomsch> yes
17:44:02 <ctyler> I don't think you want to track anything (IPs included) except the fact that you have seen a particular UUID. Not evern *when* you saw that UUID.
17:44:03 <walters_> like with smolt, we don't clearly explain the benefits of sending the data
17:44:07 <mmcgrath> stickster: I think it would accomplish what mdomsch is trying to accomplish though.
17:44:21 <stickster> mmcgrath: *nod
17:44:38 * poelcat would still like understand why dgilmore says "has to be opt-in" + why decision was made to make smolt opt-in
17:44:45 <mdomsch> ctyler, implementation-wise, we currently build a map of "checkins over the last 7 days"
17:44:56 <stickster> poelcat: You might want to go back and check the archives of FAB (?) for smolt discussions
17:45:03 <mmcgrath> poelcat: I made smolt opt in to because I got tired of people complaining about it.  mdomsch might have a stronger will than I :)
17:45:05 <stickster> It was fairly extensive
17:45:08 <ctyler> mdomsch: tied to what? IP?
17:45:17 <mdomsch> ctyler, yes, as that's all I've got to go on
17:45:31 <ctyler> so if you do UUID, would you drop IP?
17:45:36 <stickster> OK, I think we've answered mdomsch's question, which is "What shall we do?"
17:45:47 <mdomsch> ctyler, no; need IP to generate the map (geoip)
17:45:53 <dgilmore> poelcat: because, in the event of a malicious user/event.  with the right data you could track a users travels and who he is via it
17:46:11 <stickster> #agreed mdomsch and stickster to talk to legal counsel about a revised statistics gathering method
17:46:31 <stickster> #agreed mdomsch and stickster to bring that information back to FAB for more discussion and, if applicable, planning
17:47:01 <poelcat> stickster: can you ask about whether it has to be "opt-in" too?
17:47:11 <stickster> poelcat: certainly
17:47:27 * ctyler is a lot less comfortable with IP+UUID then IP or UUID by itself
17:47:27 <stickster> #info mdomsch and stickster will set up a list of specific questions through FAB so we can gather the right ones.
17:47:33 <stickster> Let's move on
17:47:44 <stickster> We can focus on specifics on the lsit.
17:47:46 <stickster> *list
17:48:44 <stickster> The question queue is currently empty.
17:48:58 <mmcgrath> I thought maxamillion had one
17:49:22 <stickster> BobLfoot: Go ahead, you had a question
17:50:23 <stickster> BobLfoot: ?
17:50:34 <stickster> OK, we'll come back to him.
17:50:44 <stickster> maxamillion: Go ahead, you had a question
17:51:46 <stickster> maxamillion: ?
17:52:24 <walters_> our discussion of esoteric privacy issues made everyone else look away from the chat apparently =)
17:52:57 <ctyler> eyes glassed over everywhere, it seems
17:53:12 <stickster> Should I just paste it in, and we'll have at it?
17:53:19 <walters_> sounds good
17:53:51 <stickster> maxamillion asked, What is the specific, defined, and concrete end goal that the Board hopes to obtain with its current working group meetings?
17:53:51 <stickster> s/obtain/achieve
17:55:36 <stickster> (1) a clear set of proposals for ways to improve our focus on a useful product that people want to use, as part (but not all) of what the Fedora Project does
17:55:59 * poelcat thought we set some of that in our first meeting
17:56:11 <stickster> poelcat: yes, got a link handy?
17:56:21 * stickster is paraphrasing from memory which is probably not as helpful
17:56:48 <poelcat> basically address everything here: https://fedoraproject.org/wiki/Unfinished_Board_issues
17:56:56 <poelcat> the approach was
17:57:04 <poelcat> 1) SWG add items to page
17:57:13 <poelcat> 2) solicity community feedback (there was none)
17:57:27 <poelcat> 3) work through issues and make formal proposals to board as a whole
17:57:41 <poelcat> 4) when page is empty/done... SWG disbands
17:57:44 <poelcat> <EOM>
17:58:04 <poelcat> right now we are on step #3
17:58:24 <stickster> Thanks poelcat
17:58:55 <walters_> at a high level i see the SWG as a positive thing, trying to find and create rough consensus and refine existing structure, and I hope maxamillion doesn't continue to see it as a negative thing
17:59:17 <stickster> The ultimate goal is *not*, as it's been misstated, to disenfranchise people or devalue their contributions
17:59:22 <walters_> right
18:00:17 <poelcat> to be clear the SWG does not make decisions for Fedora or the board as a whole
18:00:18 <jwb> in my personal view, the SWG exists because the Board has more work than a weekly meeting can accommodate and not everyone can always make time for more than one meeting
18:00:19 * stickster reminds everyone that repeatedly in the past, conflicts have arisen (and continue to do so) where there's no clear rationale for deciding them, software being one of them
18:01:09 <poelcat> jwb: yes, that is why i originally proposed it
18:01:10 <stickster> We'd like to reduce that problem
18:01:21 <ctyler> the SWG is an accelerator, a way to spend more effort on some issues that have been around for a long time, progressing very slowly
18:02:17 <stickster> In part because the Board handles other issues as well, and sometimes being acountable to the community in the short term was causing us to lag on more long-term work.
18:04:44 <mdomsch> ultimately, I want to grow our contributor base, and grow our user base
18:05:09 <poelcat> is this a wrap for today?
18:05:31 <stickster> I think we've finished, there are no more questions in the queue.
18:05:51 <stickster> #info SWG background is found here: https://fedoraproject.org/wiki/Unfinished_Board_issues
18:06:15 <stickster> OK, I'm going to end the meeting in 15
18:06:32 <stickster> 5
18:06:34 <stickster> 3
18:06:35 <stickster> 2
18:06:36 <stickster> 1
18:06:38 <stickster> #endmeeting