From Fedora Project Wiki

Fedora Project Board Meeting :: Tuesday 2008-07-08

Roll Call

Attendees: Everyone on #fedora-board-meeting

Discussion Summary

  • Board followup on CVS commits map and privacy considerations
    • OWNER: spot will open ticket with Infrastructure, geo data opt-in for FAS
  • Board answers community questions on:
    • Whether and how to quantify and use community work data
    • The #1 supercomputer in the world runs Fedora - http://www.top500.org/system/9485
    • Spreading other big installation success stories
    • Election improvements
      • Board consensus is that open nomination process is OK, provided nominee accepts
      • OWNER: John Rose will deliver a list of places where we need to document the above
    • FESCo role
    • Other activities Board members are involved with

IRC Transcript

  • Raw discussion is here
spevack stickster: do you want to at least start things off in fedora-board-public 08 Jul 14:00
spevack make an opening statement 08 Jul 14:00
--- ChanServ (ChanServ@services.) changed mode: +v spot 08 Jul 14:00
spevack then move in here? 08 Jul 14:00
stickster Uh, sure! 08 Jul 14:00
--- ChanServ (ChanServ@services.) changed mode: +v f13 08 Jul 14:01
* stickster notes the whole Board is here for the first time officially! 08 Jul 14:02
stickster s/first time/first time with this composition/ 08 Jul 14:02
f13 huzzah. 08 Jul 14:03
spevack inode0 points us to http://www.lanl.gov/roadrunner/ 08 Jul 14:03
spevack and asks if people know that it runs Fedora? 08 Jul 14:03
--- ChanServ (ChanServ@services.) changed mode: +v quaid 08 Jul 14:03
f13 I was not aware. 08 Jul 14:04
poelcat ready to start? 08 Jul 14:04
skvidal spevack: does tchung know anyone? 08 Jul 14:04
poelcat we have a couple of followups from last week's meeting 08 Jul 14:04
poelcat the first of which requires Jef(f) 08 Jul 14:04
spoleeba there are probably a lot of research clusters that run fedora 08 Jul 14:04
mdomsch the ones Dell sells tend to run RHEL or a variant thereof 08 Jul 14:05
stickster poelcat: Go ahead, let's get them handled so we can go to Q&A 08 Jul 14:05
--- poelcat (n=slick@fedora/poelcat) changed topic: privacy discussion around CVS commits 08 Jul 14:05
poelcat spoleeba: do you have an update? 08 Jul 14:05
spoleeba poelcat, update on what? 08 Jul 14:06
quaid 08 Jul 14:06
* quaid suddenly feels lost about the topic 08 Jul 14:06
spoleeba poelcat, the fact that we can't actually display the map for cvs commits 08 Jul 14:06
poelcat http://fedoraproject.org/wiki/Board/Meetings/2008-07-01#Guarding_Privacy 08 Jul 14:06
quaid step back question -- 08 Jul 14:07
spoleeba poelcat, no update... the map we have is stale.. i just have to ping the right admins to remove the map.. 08 Jul 14:07
quaid what is the exact problem, in our privacy viewpoint, with anonymous geo data? 08 Jul 14:07
poelcat 08 Jul 14:07
* poelcat notes followup for next meeting 08 Jul 14:07
spoleeba quaid, the cvs commits arent anon 08 Jul 14:07
spoleeba quaid, they are tied to account activity 08 Jul 14:07
spoleeba quaid, downloads...are anon 08 Jul 14:08
quaid so is the problem that we are not anonymizing it before revealing? 08 Jul 14:08
skvidal the problem is we're showing where people are coming from 08 Jul 14:08
quaid so, yes 08 Jul 14:08
f13 quaid: the problem is that cvs commits are contribution, rather than consumption 08 Jul 14:08
stickster quaid: And because FAS accounts may have geographic location attached, even though we didn't *intend* to pull data from FAS, we want to avoid appearance of not following our privacy policy 08 Jul 14:08
skvidal and in certain cases they could be the only contrbutor in their country 08 Jul 14:08
f13 quaid: I don't really want to know if there are people contributing to Fedora from certain countries. 08 Jul 14:09
h\h you can always use Ohloh for a quick overview ... :) http://www.ohloh.net/projects/fedora 08 Jul 14:09
skvidal and they might not want their gov't being able to know it 08 Jul 14:09
spevack this problem apply to the wiki edits maps also, then, yes? 08 Jul 14:09
spoleeba quaid, we would need to build in an opt in process in fas for people to choose to let us put them a map 08 Jul 14:09
mdomsch 08 Jul 14:09
* mdomsch nukes maps/cvs-commits/ 08 Jul 14:09
quaid oh please 08 Jul 14:09
mdomsch http://fedoraproject.org/maps/ no longer shows the maps 08 Jul 14:09
quaid if I 'm a gummint that cares 08 Jul 14:09
quaid I *know* already you are committing 08 Jul 14:09
quaid just from your IP traffic. 08 Jul 14:09
quaid what I want to know is ... 08 Jul 14:09
notting h\h: Ohloh Analysis Summary * Mostly written in shell script. heh. 08 Jul 14:09
quaid are we stopping this activity forever or just until we can properly anonymize and move to opt-in for aggregation? 08 Jul 14:10
spoleeba quaid, do we leak that ip information publically? 08 Jul 14:10
f13 quaid: we're stopping it until accounts opt-in to sharing their geographical location 08 Jul 14:10
quaid spoleeba: I'm just saying the Evil Gummit argument is ridiculous. 08 Jul 14:10
stickster spot spoke to this pretty well last week. spot, care to chime in? 08 Jul 14:10
quaid f13: thanks 08 Jul 14:10
spot 08 Jul 14:10
* spot needs to be careful what he "says" here. 08 Jul 14:10
f13 quaid: once we know that the contributor doesn't care, then we can make use of that data in various ways, such as commit maps 08 Jul 14:10
f13 at least that's my opinion 08 Jul 14:10
stickster *nod 08 Jul 14:10
quaid I was offsite and unavail last week so this is my first exposure to the discussion, fwiw 08 Jul 14:11
f13 nod 08 Jul 14:11
quaid f13: +1 08 Jul 14:11
spoleeba f13, do we leak ip logs in commit mails or similar? 08 Jul 14:11
spot but we can't have anything geographical that has the possibility of being tied back to contributors until they can opt in 08 Jul 14:11
spot and then we can only use geo data from the contributors who have opted-in 08 Jul 14:11
f13 spoleeba: I don't believe so, but I wouldn't hold me accountable for that. 08 Jul 14:11
quaid makes sense, I just wanted to be sure we were going in the right direction v. full stop 08 Jul 14:11
spoleeba f13, we should check that..since this basically comes down to keeping the ip addresses in the logs..private 08 Jul 14:12
f13 sure. I haven't been on cvs-commits-list for a long time. 08 Jul 14:12
spoleeba f13, if we are leaking the ips... then the hand wringing over geoip is pointless 08 Jul 14:12
stickster Is there anything preventing us from posting broader regional data like NA, SA, EMEA, etc.? 08 Jul 14:12
stickster As opposed to a geo map? 08 Jul 14:12
f13 spoleeba: I don't believe we're leaking th ough, looking at some of these commits 08 Jul 14:12
mdomsch I wouldn't think so, that's aggregation 08 Jul 14:13
f13 stickster: I would avoid doing such things 08 Jul 14:13
spevack stickster: a related question from JonRob to this topic, when you are ready to move from the specific to the general case 08 Jul 14:13
f13 stickster: we want to do it right (opt-in) rather than "what we can get away with" 08 Jul 14:13
spoleeba stickster, im not sure there's a lot of value in a patchwork with areas that large 08 Jul 14:13
quaid spoleeba: I don't see any IPs in any CVS commit messages that I get 08 Jul 14:13
spot stickster: i'm not sure how valuable that data is, other than to say things like "we got commits." 08 Jul 14:13
quaid I'm sure it's valuable 08 Jul 14:14
stickster spoleeba: There's some possible value to the way we spend community architecture funds... but I don't mean to suggest we should try and "get away with" anything. 08 Jul 14:14
spoleeba quaid, regardless of the evil government argument.. the fact that matters to me..was that we do right by our privacy agreement 08 Jul 14:14
quaid marketing focus -- where do we have activity, where do we not 08 Jul 14:14
ctyler backup to a claim of being global 08 Jul 14:14
spoleeba quaid, over all of asia? versus all of europe? 08 Jul 14:14
f13 is there any reason to play "how much can we get away with" games now, instead of waiting until we get opt-in/out capability later? 08 Jul 14:14
quaid IMO, Fedora Project is likely within it's rights to manually anonymize whatever data we have in the systems. 08 Jul 14:14
stickster OK, we're 14 minutes in. spoleeba, can you handle getting this map dismantled by next week? 08 Jul 14:14
skvidal f13: +1 08 Jul 14:15
spoleeba quaid, im not sure thats fine grained enough to be useful... it look like a risk map 08 Jul 14:15
quaid spoleeba: as granular as we can get it 08 Jul 14:15
spot i would still feel a lot better if we only used data from contributors who have opted in. 08 Jul 14:15
quaid spoleeba: you are a scientist, not a marketeer :D 08 Jul 14:15
spoleeba stickster, mdomsch already nuked it 08 Jul 14:15
spevack those maps and the data behind them are the single best piece of information that we have when it comes time to talk $ with RH VIPs. 08 Jul 14:15
quaid spevack +1 08 Jul 14:15
quaid that's what I mean 08 Jul 14:15
quaid people who use slidedecks as "in depth analysis" 08 Jul 14:15
mdomsch spevack, but that need not be public 08 Jul 14:15
quaid nope 08 Jul 14:15
f13 usefulness doesn't override responsibility 08 Jul 14:15
spot then, lets get an opt-in system for people to let us use their country code. 08 Jul 14:15
spoleeba spevack, im not disagreeing what that...but i think they need to be some what more fine grained than..continent coloring 08 Jul 14:15
poelcat spoleeba mdomsch is there any followup here? 08 Jul 14:15
spot that data exists in FAS2 08 Jul 14:15
mdomsch poelcat: follow-up is: 08 Jul 14:15
spevack spoleeba: +1 08 Jul 14:16
quaid f13: agreed, just arguing that it *is* useful, where some here are saying they don't see a use. 08 Jul 14:16
spoleeba poelcat, until we get the log parsing scripts reworked..there's nothing here 08 Jul 14:16
mdomsch 1) contributor maps have been removed 08 Jul 14:16
spevack mdomsch: true, but I'd prefer that as much be public as possible, always, based of course on the legal requirements that we must comply with 08 Jul 14:16
mdomsch 2) FAS work needed to let users "opt in" to having their geographic location used in public 08 Jul 14:16
spoleeba poelcat, basically...before we put up any new map concepts..we need to approve them at the board level to make sure we dont screw up privacy data 08 Jul 14:16
mdomsch EOL 08 Jul 14:17
stickster Who's going to own #2 on mdomsch's summary? 08 Jul 14:17
spot i will open a ticket with infrastructure 08 Jul 14:17
spot they already love me. ;) 08 Jul 14:17
stickster You're a lovable guy. 08 Jul 14:17
skvidal 08 Jul 14:17
* skvidal just pinged folks in #fedora-admin 08 Jul 14:17
poelcat okay... other followups were longer term 08 Jul 14:18
quaid ha, only four clicks to get to the privacy policy from fp.o 08 Jul 14:18
poelcat and may want to touch on them here during the Q&A 08 Jul 14:18
poelcat which were: board elections and future trademark usage 08 Jul 14:18
spevack are we ready for Q&A? 08 Jul 14:18
poelcat followups with board after discussion on f-a-b will be on 2008-08-05 08 Jul 14:18
mdomsch quaid: I can get there in 3 :-) 08 Jul 14:18
spot 08 Jul 14:18
* spot was born ready. ;) 08 Jul 14:18
stickster poelcat: Yeah, those seem like they can easily come up as Q&A fodder 08 Jul 14:19
quaid mdomsch: well, only if you scroll where I clicked on the the ToC on the Legal page 08 Jul 14:19
poelcat spevack: that is all from the desk of the secretary 08 Jul 14:19
spevack question from JonRob which is a natural segue from the previous topic. I shall summarize: 08 Jul 14:19
f13 I'm ready. 08 Jul 14:19
spevack what is the board's thoughts about how we should use contributor data? 08 Jul 14:20
spevack in particular: 08 Jul 14:20
spevack with regard to figuring out/monitoring how much "work" any given contributor is doing? 08 Jul 14:20
spevack EOL 08 Jul 14:20
f13 I think we should be using it in the most 'anonymous' way possible, and the least intrusive way possible. 08 Jul 14:20
spot well, we can't use data covered by the privacy policy 08 Jul 14:20
quaid We're pretty limited by this in terms of anything not a public forum: http://www.redhat.com/legal/privacy_statement.html#share 08 Jul 14:20
quaid yeah, what spot said 08 Jul 14:21
spot but we can say that spot had 40000 commits last night. 08 Jul 14:21
f13 we'dl ike to show that "work" or "contribution" is done by areas of the world, and by areas of SIGS, but without revealing specifically who is doing which specific work. 08 Jul 14:21
quaid because that is a public forum 08 Jul 14:21
mdomsch slow night for spot 08 Jul 14:21
stickster It gets touchier to produce these statistics when SIGs dwindle in size 08 Jul 14:21
h\h f13, while the data who is doing what is already public in the cvs commit log anyway.. 08 Jul 14:22
spoleeba spevack, different subproject areas need to figure out what makes sense as a unit of work 08 Jul 14:22
spoleeba spevack, obviously commits for "packaging subproject" but wtf do you we stand up as a unit of work for ambassadors? 08 Jul 14:23
stickster I think spoleeba would agree there's some value in making sure contributors aren't burning out -- that doesn't mean forcing people to change jobs, but simply making sure we don't let people get overburdened simply because no one notices how much they're doing. 08 Jul 14:23
spot from my perspective, any data that we gather from a user making public contributions (owning packages, making cvs commits) is fine. 08 Jul 14:23
mdomsch h\h, agreed, but as you recognize too, there's more to contributing than CVS commits 08 Jul 14:23
h\h right 08 Jul 14:23
skvidal stickster: it feels like the opposite t oem 08 Jul 14:23
ctyler some activity is really hard to centrally monitor 08 Jul 14:23
skvidal it feels like tracking user contributions will make it feel like a job 08 Jul 14:24
spoleeba mdomsch, i think we should count average length of sentences in mailinglist posts 08 Jul 14:24
stickster skvidal: There's value in making sure people *do* burn out? 08 Jul 14:24
quaid what we want to be careful of is having it become a negative pointer, only a positive pointer -- it shows when people are working hard, but doesn't make a spectacle out of people who are e.g. not responding to their bug reports 08 Jul 14:24
skvidal any time I've been faced with daily/hourly status reports in jobs it always feels like I'm being watched 08 Jul 14:24
stickster skvidal: no, no -- I'm not talking about making people track their work. 08 Jul 14:24
spot i would rather not use stats to tell people when they are failing. 08 Jul 14:24
f13 skvidal: +1 08 Jul 14:24
f13 stickster: no, but you're tracking it /for/ them 08 Jul 14:24
h\h http://www.ohloh.net/projects/fedora/contributors 08 Jul 14:24
skvidal stickster: I'm sure you're not but it may feel like that 08 Jul 14:24
quaid spot: a nicer way of saying what I meant :) 08 Jul 14:24
f13 which essentially is the same thing. 08 Jul 14:24
spot if we simply present the statistics, i see no fault in that 08 Jul 14:25
quaid even if it were above the individual level, e.g. tracking how much 'work' a SIG is doing, people will feel it personally. 08 Jul 14:25
spot the fedora packaging reports do this, without assigning blame 08 Jul 14:25
spoleeba spevack, and there is the question of what do we plan to do with anything we collect... assuming for example we can trend the amount of work ambassadors as a group are doing...how would we use the data...do we have resources that could be allocated based on growth or need? 08 Jul 14:26
stickster f13: So letting a contributor know, "Hey, I noticed you're regularly taking care of 400 packages, and I just want to make sure you have what you need to be comfortable doing that" is bad? 08 Jul 14:26
skvidal stickster: it's not bad - it just won't stay that way 08 Jul 14:26
skvidal b/c it just won't be us being well-intentioned 08 Jul 14:26
skvidal it'll be some ass hat being an asshat 08 Jul 14:26
spot stickster: no one is actually, i'm the biggest at 266. but still. ;) 08 Jul 14:26
skvidal and brandishing the stats like a bludgeon 08 Jul 14:26
skvidal 'You don't do enough' 08 Jul 14:26
stickster spot: I just want to make sure you have what you need to be comfortable with that :-) 08 Jul 14:27
skvidal 'jeez, you didn't do shit last week, did you' 08 Jul 14:27
f13 stickster: I think that's targetted which might not be the rigth thing. Making a general announceement to provide anything that anybody is missing is likely more "correct" 08 Jul 14:27
f13 stickster: just because somebody is taking care of 400 packages doesn't mean that he deserves any more attention than the person managing 2 packages. 08 Jul 14:27
ctyler I think we need to consider that someone who organizes local installfests and is responsible for adding hundreds of people to the Fedora community might barely register using most metrics 08 Jul 14:27
f13 yes 08 Jul 14:27
spot i would much rather see energy going into quicker detection of absent maintainers 08 Jul 14:27
f13 metrics-- 08 Jul 14:27
stickster ctyler: Yes, spoleeba's point earlier 08 Jul 14:27
h\h ctyler, right 08 Jul 14:27
spot if someone hasn't committed in XX months, we gently checkin 08 Jul 14:28
skvidal stickster: information which can be used well-intentionedly can also be abused - hence privacy policies 08 Jul 14:28
f13 metrics merely present a way to game the system and morph data into ones personal vendetta. 08 Jul 14:28
stickster So the answer is peer-based 08 Jul 14:28
skvidal yes 08 Jul 14:28
spoleeba skvidal, i want aggregate data to trend. I want to be able to see if some area is exploding or imploding relative to the growth of the work being done in another area.. i dont want individual contributor stats from a project management standpoint 08 Jul 14:28
f13 spot: I'm perfectly OK with that 08 Jul 14:28
f13 spot: in fact, I'm highly for that. 08 Jul 14:28
quaid 08 Jul 14:28
* quaid cannot find himself on ohloh.net for example 08 Jul 14:28
spevack JonRob has a followup comment: if you highlight big, successful contributions, are you perhaps demotivating people for whom a small contribution is a major effort -- newbies, etc? 08 Jul 14:28
skvidal 08 Jul 14:28
* skvidal is fine with spots suggestion 08 Jul 14:28
f13 JonRob: that's what I'd like to avoid 08 Jul 14:29
stickster JonRob: good point, and that also points to a peer-based solution 08 Jul 14:29
f13 however at times it may be impossible to avoid that 08 Jul 14:29
spot JonRob: like ctaylor said, there are lots of contributions that are hard to track, and hard to "reward" 08 Jul 14:29
quaid peer rewards might be cool 08 Jul 14:29
f13 everybody's contribution means something, and it's all very dynamic and nearly impossible to 'metric' 08 Jul 14:29
spot i've had the basic idea for a karma system bouncing around in my head for months 08 Jul 14:29
mdomsch 08 Jul 14:29
* mdomsch has a real-life loud fire alarm blaring in his ear; bbiab 08 Jul 14:29
spot but the implementation details are... hard. 08 Jul 14:29
f13 but from time to time the board or some other sig may choose to highlight a particular individual for a particular job 08 Jul 14:30
stickster mdomsch should not be smoking in the server room. 08 Jul 14:30
spoleeba f13, if we dont stand up the metrics as the basis of any project wide reward system..i think we can avoid it 08 Jul 14:30
spot i'd be interested in hearing ideas on how to do such a karma system. 08 Jul 14:30
skvidal do it in fas 08 Jul 14:30
f13 spoleeba: again, it's not what 'we' would do with the data, it's what other people would do. 08 Jul 14:30
spot skvidal: not on that level. ;) 08 Jul 14:30
skvidal 'give this contributor karma' 08 Jul 14:30
skvidal no, I'm serious - in another db 08 Jul 14:30
notting spot: bugzilla points? (not great karma, but exists) 08 Jul 14:30
h\h spot, what about the ohloh kudu system? 08 Jul 14:30
f13 yeah, that fits in with my method to promote people to "uberpackager" group 08 Jul 14:30
ctyler you'd almost need each sig/subproject to define karma in their context 08 Jul 14:31
spot ctyler: exactly. 08 Jul 14:31
stickster notting: How do you promote ambassadors if it's BZ? 08 Jul 14:31
f13 nod 08 Jul 14:31
notting stickster: you don't. separate karma for separate subprojects, i suppose 08 Jul 14:31
f13 there likely isn't any one single way to do this. 08 Jul 14:31
quaid h\h: can you point at an explanation of the kudu system? 08 Jul 14:31
h\h kudo 08 Jul 14:31
f13 quaid: essentially it allows you to pat somebody on the back and write up a little nice note about them 08 Jul 14:32
f13 ohloh can show for any account how many and what the content of kudo's they've received 08 Jul 14:32
quaid ic 08 Jul 14:32
spevack i have always thought that every time we plan a fudcon, we need to see who the people are geographically close by that maintain huge numbers of packages for us and invite them with the fully-sponsored rockstar treatment. Same for other major contribs outside packaging. 08 Jul 14:32
f13 I could certainly see a Fedora kudo's board, maybe part of planet, that as peers give eachother kudos those are broadcasted on the kudo board 08 Jul 14:32
spevack as we continue to get more budget for fudcons, this becomes more of a reality 08 Jul 14:33
spoleeba f13, the problem with different groups having their own reward systems is...if we need to allocate any resources to underpin a reward process..how big of a group start to matter? A SIG with 3 people in it..does that get its own karma system? 08 Jul 14:33
f13 spoleeba: why shoudl the package count matter? 08 Jul 14:33
spoleeba f13, not package... people count 08 Jul 14:33
f13 er 08 Jul 14:33
notting spoleeba: perhaps sigs are responsible for their own metrics. delegate! 08 Jul 14:33
f13 sorry 08 Jul 14:33
f13 spevack: why should the package count matter? 08 Jul 14:33
notting f13: in the absence of better metrics... 08 Jul 14:34
f13 is a person who maintains 50 tiny packages any more of a rock star than a person who maintains 5 big packages with lots of updates? 08 Jul 14:34
spoleeba notting, no you miss my point... do we let a 3 person SIG have equal presence with a 20 person SIG in whatever system we use? 08 Jul 14:34
spevack f13: no, of course not. replace "package count" with some sort of better metric :) 08 Jul 14:34
h\h quaid, http://www.ohloh.net/about/kudos 08 Jul 14:34
quaid rockstar systems ... suck more than they are good 08 Jul 14:34
quaid h\h: thx 08 Jul 14:34
spot 08 Jul 14:34
* spot notes that this is a non-trivial problem to solve, which is why i've never solved it. ;) 08 Jul 14:34
f13 indeed 08 Jul 14:35
f13 I would much rather leave it purely on a peer to peer basis 08 Jul 14:35
f13 any peer can kudo or high five or whatever any other pper 08 Jul 14:35
f13 peer 08 Jul 14:35
stickster Maybe we should track meta-kudos. 08 Jul 14:35
f13 one's account could keep track of those, or it could not, and we could broadcast those on the kudowall, or not 08 Jul 14:35
spoleeba f13, which is why im just looking to expose aggregate metric trending..and not person-to-person info 08 Jul 14:35
f13 spoleeba: then I think you're going to have to come up with specific uses for those metrics 08 Jul 14:36
spevack 08 Jul 14:36
* spevack wonders if we should move on to the next question 08 Jul 14:36
quaid +1 08 Jul 14:36
f13 and not just leave them out there to be (ab)used by whomever. 08 Jul 14:36
spot yep. next Q? 08 Jul 14:36
stickster 1 08 Jul 14:36
h\h spoleeba, metrics for cool non-code contributions? 08 Jul 14:36
spevack NEXT QUESTION -- inode0 points us to http://www.lanl.gov/roadrunner/ -- it runs Fedora, and what marketing possibilities might exist due to this? 08 Jul 14:37
spoleeba h\h: we need to get a handle on what areas of "work" are growing relative to others 08 Jul 14:37
spoleeba h\h, including the non technical work 08 Jul 14:37
skvidal spevack: where does it say fedora? 08 Jul 14:37
spot skvidal: lanl has said that it does, in several random places 08 Jul 14:37
skvidal spot: okay. cool 08 Jul 14:38
spoleeba h\h, not on a person by person basis...but area by area...trending the growth rates. So we can start to figure out if we need to do a recruitment drive for one area specifically as part of say f11 propoganda 08 Jul 14:38
h\h spoleeba, ok 08 Jul 14:38
skvidal spoleeba: do you know someone at lanl? 08 Jul 14:38
spot sounds interesting. spevack should get right on it. ;) 08 Jul 14:38
skvidal if not I mentioned hitting up tchung 08 Jul 14:38
spoleeba skvidal, i probably know someone who knows someone 08 Jul 14:38
notting well, first step would be to get them to mention it in press releases and non-random places, if possible 08 Jul 14:38
skvidal and if not I can probably find some folks at the national labs 08 Jul 14:38
stickster http://www.top500.org/system/9485 08 Jul 14:38
spoleeba skvidal, the peopel I know..dont do the admin stuff that keeps the compute farms actually running 08 Jul 14:38
f13 jack's blog post about NASA using Fedora got lots of attention, we should certainly capatalize upon lanl, although I would hate to find out they're abusing Fedora in some way that we don't normally allow 08 Jul 14:39
stickster The problem I've found (and experienced) in the Federal space is that it's very difficult for govvies to talk about this stuff 08 Jul 14:39
spot i'm guessing its a heavily modified fedora, given that ibm is involved. :) 08 Jul 14:39
ctyler bounce this to fedora-marketing-list? 08 Jul 14:39
spoleeba the arctic supercomputer here runs fedora as well... i can reach into their little world and see if they can reach into lanl 08 Jul 14:39
stickster ctyler: +1. 08 Jul 14:39
spevack spot: i'm fine with adding this topic to our community architecture "big press stories we're hunting down" agenda 08 Jul 14:39
--- ChanServ (ChanServ@services.) changed mode: +v mdomsch_ 08 Jul 14:40
skvidal ctyler: nah 08 Jul 14:40
skvidal I'd say rather than delegating it out 08 Jul 14:40
skvidal find someone who knows someone in the org 08 Jul 14:40
spot this is so cool, i ssh into this system and the dell fire alarm goes off 08 Jul 14:40
skvidal it'll be easier of an approach 08 Jul 14:40
spot i mean, hi matt! :) 08 Jul 14:40
stickster skvidal: Usually that results in the same problem I mentioned above, only with more dead time in between. 08 Jul 14:40
spevack ctyler: give it to me, greg, and jack to chase down and see if we know anyone, like skvidal says. IMHO. 08 Jul 14:40
mdomsch_ 08 Jul 14:40
* mdomsch_ watches the firetruck drive up 08 Jul 14:40
skvidal spevack: +1 08 Jul 14:40
spoleeba skvidal, give me a couple of weeks 08 Jul 14:40
mdomsch_ from the parking log 08 Jul 14:40
stickster spoleeba: If you can get an actual govvie to talk about this in an interview, I'll give you a prize. 08 Jul 14:41
skvidal spevack: If you hit any walls, bug me - I still know enough physics people I can probably find a connector there 08 Jul 14:41
spoleeba skvidal, also HAARP runs on a mix of RHEL and Fedora... 08 Jul 14:41
skvidal actually, you know who I do know 08 Jul 14:41
skvidal most ofthe Scientific Linux people 08 Jul 14:41
skvidal 08 Jul 14:41
* skvidal wonders if connie is still doing that... 08 Jul 14:41
skvidal they'll know the sysadmins at the various labs 08 Jul 14:41
f13 I just want to make sure that before we make any loud noises about these cool uses of Fedora, that ti's uses we'd wish to promote, not Hey look they use Fedora; Oh, but they've replaced half the packages with propriatery stuff and... and... 08 Jul 14:41
skvidal f13: which is why we want to talk to them directly 08 Jul 14:42
skvidal so, agreed 08 Jul 14:42
ctyler skvidal: my point was that f-m-l handles marketing, and this is a marketing thing 08 Jul 14:42
f13 I'd be more than happy to let marketing handle that, if they keep the above in mind 08 Jul 14:43
stickster skvidal: ctyler: f13: Several of the marketeers are here listening in as we speak ;-) 08 Jul 14:43
f13 (don't spout without knowing the full details as to avoid any "whoops" incidences) 08 Jul 14:43
spoleeba stickster, i can get an interview with the guys here who run the artic supercomputer 08 Jul 14:43
stickster spoleeba: f-m-l away, then! 08 Jul 14:43
spoleeba stickster, i'll use the marketting list to compile a list of questions 08 Jul 14:43
stickster spoleeba: Brilliant. 08 Jul 14:43
spot next? 08 Jul 14:44
spevack f13: this particular case is something that would end up being vetted before any press goes out. 08 Jul 14:44
spevack f13: as we would for anything "noteworthy" 08 Jul 14:44
spevack next question is a simple one from former board member caillon 08 Jul 14:44
spevack he would like nickserv authentication to be removed for fedora-board-public and fedora-board-meeting 08 Jul 14:45
spevack 08 Jul 14:45
* spevack has a real question next :) 08 Jul 14:45
spot i'll take care of that after the meeting. 08 Jul 14:45
spevack ok, next: 08 Jul 14:45
stickster spot: Thanks 08 Jul 14:45
spevack There was a long thread on f-a-b that touched on this 08 Jul 14:45
mdomsch_ spot: stickster has ops 08 Jul 14:45
spevack but the question was: 08 Jul 14:45
stickster mdomsch_: Oh, can I do that from here? 08 Jul 14:46
spevack what is the board going to do, or thinking about doing, to promote more diversity in nominations/appointments for the top committees? 08 Jul 14:46
* stickster goes to look it up 08 Jul 14:46
spot i'm going to bribe less people next time. 08 Jul 14:46
spot i mean, uhh. 08 Jul 14:46
spoleeba "top" committees? 08 Jul 14:46
spot seriously though, are you asking us to influence the vote? 08 Jul 14:46
spoleeba are we talking Board and fesco specifically? or is there more 08 Jul 14:46
spot the best we can do is to try to get the word out more, encourage people to run 08 Jul 14:47
spoleeba i want some frelling feedback from the frelling community during the election cycle..for candidates to speak to directly 08 Jul 14:47
f13 making it easier to vote (thanks G!!!), making it easier to nominate ones self or somebody else, and making it obvious that there is a nomination process ongoing/voting ongoing. 08 Jul 14:47
stickster There are some other changes that have been suggested, like term limits. 08 Jul 14:48
stickster I think those got fairly broad support in the Board, fwiw. 08 Jul 14:48
spoleeba term limits... a nomination process so you can suggest and endorse other people into running 08 Jul 14:48
f13 ah yes, consecutive term limits with mandatory break periods 08 Jul 14:48
quaid spevack: was the question "top committees" or the "Board"? I read "Board" in the other channel, and that is a difference 08 Jul 14:48
notting spoleeba: i believe the word you're looking for is 'goad' 08 Jul 14:48
spevack i think in a follow up there was suggestion that it is a problem in other big steering committees (like Fesco) 08 Jul 14:49
skvidal notting: it's not 'shanghai'? 08 Jul 14:49
spevack but the first question was just Board 08 Jul 14:49
stickster spevack: I would like to see some proposals generated from outside the Board 08 Jul 14:49
skvidal notting: 'voluntold' 08 Jul 14:49
spoleeba why aren't more individuals endorsing candidates? Where's the labor and teamster union of Fedora land? 08 Jul 14:49
stickster spoleeba: Careful, I think spevack's got a guy 08 Jul 14:49
--- ChanServ (ChanServ@services.) changed mode: +o spot 08 Jul 14:50
spevack inode0 suggests that nominations to fill empty board seats are excellent ways of providing recognition 08 Jul 14:50
--- spot (i=spot@redhat/spot) changed mode: -r 08 Jul 14:50
--- ChanServ (ChanServ@services.) changed mode: +r 08 Jul 14:50
f13 I think I'd like to see an anonymous survey of why people didn't vote, and the results of that survey made public 08 Jul 14:50
f13 right now we're mostly guessing at why people didn't vote, or why people didn't run for election 08 Jul 14:50
h\h f13, survey from people, who don't vote .. hmm, how do you get them to fill out the form? :) 08 Jul 14:51
spoleeba "you are doing a heck of a job doing what you are doing..so to reward you..we are going to appoint you to this committee so you can get less of what you are doing well done and spend more time wringing your hands over unsolvable problems" 08 Jul 14:51
skvidal h\h: census workers :) 08 Jul 14:51
f13 h\h: great question, we won't get everybody. But I bet we'll get some people to comment. 08 Jul 14:51
skvidal let's be fair about a few things, though 08 Jul 14:52
--- ChanServ (ChanServ@services.) changed mode: -r 08 Jul 14:52
quaid more likely those who care and had a reason, v. pure apathy 08 Jul 14:52
skvidal 1. I bet a number of folks just forgot about the election 08 Jul 14:52
quaid (longer voting window would be OK) 08 Jul 14:52
skvidal 2. a number of folks probably didn't realize they could vote - b/c it was a little confusing as to who was allowed 08 Jul 14:52
skvidal so direct reminders might help 08 Jul 14:52
spoleeba skvidal, i can fix that... auto dialing of asterick numbers in our new call infrastructure before the election 08 Jul 14:52
skvidal and a bit more build up before it 08 Jul 14:52
stickster Let's also consider the part of the question that's more goal-oriented 08 Jul 14:53
skvidal spoleeba: that's nice - but I meant more seriously - we can do some emailing to people in fas 08 Jul 14:53
--- spot (i=spot@redhat/spot) changed mode: -o spot 08 Jul 14:53
skvidal and we should 08 Jul 14:53
quaid skvidal: I tried to do that, fwiw, but couldn't get the nail hammered in -- there really is no way right now to alert all contributors about something 08 Jul 14:53
spoleeba skvidal, why stop at emails...we have the power to call people 08 Jul 14:53
quaid or if there is, I still haven't heard about it, and I've been asking 08 Jul 14:53
skvidal quaid: there is - it's not trivial - we shall make it trivial 08 Jul 14:53
skvidal quaid: I'll take that task to drive it along 08 Jul 14:53
quaid spevack: if Fedora starts calling me on stuff like this, I'll drive to Server Beach and yank asterisk from the rack myself 08 Jul 14:53
quaid skvidal: thank you 08 Jul 14:54
spoleeba quaid, dont you want to hear from each and every candidate personally on the phone? 08 Jul 14:54
quaid skvidal: agreed it needs to be a heavily moderated comms channel, but we do need it. 08 Jul 14:54
quaid spoleeba: only if I call them myself 08 Jul 14:54
spoleeba quaid, doesnt that show significant commitment 08 Jul 14:54
f13 Hi, this is the topic over here, would be nice if ya'll came back to me. 08 Jul 14:54
spoleeba quaid, or would you rather not have the candidates bother trying to find out what you care about as a contributor? 08 Jul 14:54
spoleeba quaid, YOU have a duty! 08 Jul 14:54
stickster We'll get further down this road by limiting the reductio ad absurdum 08 Jul 14:54
skvidal stickster: +1 08 Jul 14:55
ctyler 08 Jul 14:55
* ctyler notes that sending N wav-attached-voicemail e-mails takes a lot more bandwidth than sending N text e-mails 08 Jul 14:55
spoleeba stickster, seriously... we need input from the community to guide candidates..before the election 08 Jul 14:55
spevack comments from caillon that i'd like to transcribe: 08 Jul 14:55
spoleeba stickster, community generated questions and issues..are the only really fair way for candidates to be directly compared 08 Jul 14:55
stickster As I said before, we should consider the part of the question that's goal-oriented: What is the desired end-state of the Board composition? 08 Jul 14:56
skvidal spoleeba: that's fine - quit harping on the asterisk thing - it's just silly 08 Jul 14:56
spoleeba stickster, everyone comes in with their own agenda..what matters is how the candidates speak to community issues... 08 Jul 14:56
spevack caillon, in summary, says: FESCOs scope is still to broad, the tasks need to happen, but should be going to smaller targetted groups, not one all-encompassing group 08 Jul 14:56
spot is there a question in there? 08 Jul 14:57
f13 I think that's a different topic all together 08 Jul 14:57
spoleeba how fesco wants to delegate out its work is completely up to fesco. 08 Jul 14:57
spevack spot: not really... caillon and f13 were having a longer talk in the -public room and i asked him to try to summarize his comments in something short. 08 Jul 14:57
spevack i guess the question is "does Fedora need a re-org, and how do we go about answering that and/or doing it transparently and smartly?" 08 Jul 14:58
spot he should really raise these issues with fesco. 08 Jul 14:58
quaid spot +1 08 Jul 14:58
notting because what fesco really needs is a discussion of its purpose :) 08 Jul 14:59
f13 spoleeba: although I don't think I'd much like to as the board appoint a task to a group of elected people only to have them send it off to a set of appointed people who may or may not have any bearing on community desires. 08 Jul 14:59
spot if fesco wants to subdivide into groups, then whatever. 08 Jul 14:59
spot i dont think i care how it gets done as long as it does, and we have a hierarchy to know whos supposed to be doing it. 08 Jul 14:59
stickster spot: spevack: FESCo claimed responsibility for a set of tasks, basically the technical execution of the platform 08 Jul 15:00
spevack stickster: caillon would like voice 08 Jul 15:00
spevack stickster: and i can't give it to him. i think he doesn't like my summary of his summary :) 08 Jul 15:00
--- ChanServ (ChanServ@services.) changed mode: +v caillon 08 Jul 15:00
spevack caillon: you have voice 08 Jul 15:01
stickster If their vision of doing that means appointing SIGs to handle specific issues, that ought to be acceptable 08 Jul 15:01
caillon fesco was sort of put in a "tell us what you do/want to do" and basically recited off what they are doing currently. there has already been grumbling that it's the wrong thing to do. on top of that, the baord has since gotten involved. I do not feel that FESCo is the right committee to make the call as to what they should do at the moment. 08 Jul 15:02
f13 stickster: if FESCo only wants to exist to offload their work to other people, why can't hte board just offload directly? 08 Jul 15:02
caillon the tasks still need to happen, though! 08 Jul 15:02
spot caillon: i love how "there has already been grumbling" wasn't at the FESCo meeting. 08 Jul 15:02
caillon but I think it would make more sense for them to happen by smaller, more targetted groups. 08 Jul 15:02
caillon I might be a docs/marketing person that is interested in participating in the feature process but don't care about packgae guidelines, conflict resolution, etc 08 Jul 15:03
spoleeba f13, i have absolutely no problem with appointments to delegate specific work..and the appointments retire after a set time...delegating work doesn't mean creating lifetime supreme court tenured positions 08 Jul 15:03
quaid caillon: I don't get why this is something the Board should dictate? 08 Jul 15:03
quaid heck 08 Jul 15:03
spot i'm not sure it would be appropriate for the board to force FESCo to subdivide 08 Jul 15:03
spot i really think you should take this up with FESCo 08 Jul 15:04
quaid propose an amendment to the feature process to allow subject-specific groups to handle them 08 Jul 15:04
f13 spoleeba: what happens when 'specific' work becomes 'all' work? 08 Jul 15:04
spoleeba f13, you have an election cycle 08 Jul 15:04
spoleeba f13, why are we stuck on the belief that elections = ability to handle the work necessary? 08 Jul 15:04
f13 FESCo already voted to allow specific groups weigh in on specific sections of the feature, like QA gets to ack/nack the test plan area. 08 Jul 15:05
spoleeba f13, elections at best provide oversight... 08 Jul 15:05
spoleeba f13, good for them 08 Jul 15:05
f13 spoleeba: I'm not saying that vote was wrong, I think you're misreading me 08 Jul 15:05
spoleeba f13, and if fesco doesnt like how QA is handling it..they can yank them out of the process 08 Jul 15:06
caillon quaid, because the board has already stepped in, various members of fesco had previous raised it to the board, and nothing has changed even after it was raised on fesco's list and to the board. i really don't think anything will change by raising it with fesco. 08 Jul 15:06
f13 what I'm saying is if FESCo wants the board to tell it what to do, and the only thing we've so far told them to do they want to offload to somebody else... 08 Jul 15:06
f13 what'st he point of FESCO? 08 Jul 15:06
caillon f13, FESCo already expressed that. 08 Jul 15:06
f13 caillon: /part/ of fesco. 08 Jul 15:06
notting huh? 08 Jul 15:06
caillon the board shrugged it off by saying "you tell us what we should tell you to do" 08 Jul 15:06
quaid no 08 Jul 15:06
spoleeba caillon, i dont tell anyone what they should do 08 Jul 15:06
notting f13: i'm not parsing what you're saying 08 Jul 15:06
f13 but hey, lets have this argument again, for another week or 3 08 Jul 15:06
f13 that should be fun. 08 Jul 15:06
quaid "you tell us what you want to do" 08 Jul 15:06
quaid caillon: it's different 08 Jul 15:06
spoleeba caillon, i need individual subprojects to claim an area of work 08 Jul 15:07
quaid caillon: then we can assign the stuff to someone else if they don't want to do soemething 08 Jul 15:07
spot fwiw, FESCo did tell the board what it wanted to do 08 Jul 15:07
notting and the board said 'that's ok' 08 Jul 15:07
spot caillon just thinks that is too broad. 08 Jul 15:07
spot i'm not about to override them. 08 Jul 15:07
spoleeba spot, if its too broad...fesco can come back and tell us its too broad and we can stand up anothe group to take on the segment they dont want 08 Jul 15:08
spoleeba spot, i dont think anyone expects this to be the end of the tuning 08 Jul 15:08
spot sure, but FESCo hasn't failed yet. 08 Jul 15:08
spot imho, of course. 08 Jul 15:08
spot nor have they come to us asking to lift the burden that they defined for themselves. 08 Jul 15:08
spoleeba spot, im not saying anyone's failed. I'm saying there is no sense in carving things up.. until we either know its too much work..or we have a another group who is working in the same area and there is conflict 08 Jul 15:09
caillon spot, not as a collective, no. some people have. 08 Jul 15:09
spoleeba spot, the key is that some group has to be willing to claim the area.. we cant mandate anyone do anything 08 Jul 15:09
spot caillon: now you're just name calling. 08 Jul 15:09
spot i dont think that's very productive and definitely off-topic. 08 Jul 15:10
* stickster would like people to take a breath before we continue here. 08 Jul 15:10
caillon spot, I am physically complaining as a member of fesco. 08 Jul 15:10
spot caillon: and yet, you haven't complained to fesco 08 Jul 15:10
spot unless i missed a meeting or 5. 08 Jul 15:10
ctyler stickster: I regret that I must step out. will read scrollback later. 08 Jul 15:11
quaid clearly, if someone in FESCo doesn't like their mission, they can work to change it or vote with their feet. 08 Jul 15:11
stickster We're getting to the point where this is unproductive as a Board meeting. 08 Jul 15:11
caillon I complained on the joint board/fesco meeting 08 Jul 15:11
quaid stickster: depends on your definition of productive :) 08 Jul 15:11
stickster zing! 08 Jul 15:11
quaid stickster: part of the reason for this Q&A is to air out opinions and personalities 08 Jul 15:11
quaid stickster: e.g. so people can see that the Board doesn't say shit in private that it doesn't in public, or whatever 08 Jul 15:12
spot are there other questions? i don't think there is an "answer" that the board can give here. 08 Jul 15:12
f13 inode0: has brought up a point to this topic. 08 Jul 15:12
f13 we don't currently have an official line that allows for open nominations, the current statement is self nomination. 08 Jul 15:12
spoleeba spot, i could be blunter 08 Jul 15:12
spevack spot: there were no other questions in the queue. Though i prepared a few of my own, in case there weren't any 08 Jul 15:12
spoleeba spot, but im not sure that's gonna help 08 Jul 15:12
spevack spot: i could ask them, or not.  :) 08 Jul 15:12
f13 Is there any reason why we couldn't allow open nomination? 08 Jul 15:12
quaid f13: I see no reason 08 Jul 15:13
f13 (in practice the nominee would have to accept the nomination anyway to get on the ballot I would think) 08 Jul 15:13
spevack 08 Jul 15:13
* spevack feels like Wolf Blitzer or something. I want to moderate the first ever Fedora Board Election Debate 08 Jul 15:13
quaid right 08 Jul 15:13
spot as long as the person nominated is willing. 08 Jul 15:13
f13 ok. 08 Jul 15:13
stickster It doesn't seem problematic to me. 08 Jul 15:13
spoleeba f13, open nominations are fine...as long as the people who end up being nominated accept it 08 Jul 15:13
quaid can we call that a consensus? :D 08 Jul 15:13
f13 awesome. 08 Jul 15:13
caillon spot, no, no answer expected. just saying it's a problem, and i do not have the confidence that fesco will magically fix itself, after bpepple and i started the discussion a few months ago, and the recent developments where the board got involved. 08 Jul 15:13
spoleeba f13, by doing the same sort of stuff the self-nominated people do 08 Jul 15:13
spoleeba f13, to indicate their candidacy 08 Jul 15:13
f13 inode0: if you noticed, we just approved open nomination. Can you take an action item to help us (the board) find places where we'd need to document that open nominiation as well as self nomination is acceptable? 08 Jul 15:14
f13 inode0: (you can answer in -public) 08 Jul 15:14
caillon anyway, i guess i'm done. someone can -v me 08 Jul 15:14
stickster spevack: Maybe you can ask a question? 08 Jul 15:15
spoleeba f13, in fact I would go further and add an area in the bio info for candidates which had a place for personal contributor endorsements 08 Jul 15:15
stickster We're over time, but I'm not going anywhere 08 Jul 15:15
spevack stickster: here's another question for the Board 08 Jul 15:15
stickster It's up to the other members whether they can stick around 08 Jul 15:15
spoleeba f13, if you nominate someone..surely you can add your name to their endorsement section 08 Jul 15:15
f13 f13: sure thing 08 Jul 15:15
spevack Aside from this current topic, which obviously needs more discussion (FESCO/engineering organization in general), what are some ot the other decisions that the Board will be publicizing in the next few weeks? in other words, what's currently brewing? 08 Jul 15:15
f13 inode0: thanks a ton! 08 Jul 15:16
f13 <crickets> 08 Jul 15:16
stickster We've started talking about trademark guidelines: http://fedoraproject.org/wiki/Board/Meetings/2008-07-01#Trademark_Guidelines 08 Jul 15:17
spevack i have other questions :) 08 Jul 15:17
spot 08 Jul 15:17
* spot has been here for a week now. i haven't had time to brew much. 08 Jul 15:17
spoleeba at the board level? I'm not actively persuing anything that has reached firestorm panic level 08 Jul 15:17
spoleeba i still have to watch the SPIN SIG process development 08 Jul 15:17
spoleeba I'm watching the development of irc help re-org 08 Jul 15:17
stickster I'd like to have trademark guidelines that put more use in the hands of the community 08 Jul 15:18
spoleeba i'm sucking up to JonRob and his miro channel 08 Jul 15:18
spevack spoleeba: to what extent are you driving the irc help re-org, or is it just passive? 08 Jul 15:18
f13 I'm watching FESCo and that whole ball of fun 08 Jul 15:18
spoleeba spevack, im not leading it...delibrately 08 Jul 15:18
spevack spoleeba: is it moving, though? 08 Jul 15:18
spoleeba spevack, they have draft guidance 08 Jul 15:18
spevack 08 Jul 15:19
* spevack doesn't know the answer, it's not a planted question :) 08 Jul 15:19
spevack spoleeba: great 08 Jul 15:19
stickster nirik is doing a lot of the organizational work for irc help revamp. 08 Jul 15:19
spot 08 Jul 15:19
* spot is afk 08 Jul 15:19
spoleeba spevack, basically im watching for process breakdown..and i'll wade in at that point 08 Jul 15:19
stickster He's got a bunch of people involved from throughout the community 08 Jul 15:19
spevack spoleeba: if someone wants to see/participate, what's the URL they should go to? 08 Jul 15:19
spoleeba spevack, that's a good question...they are doing irc meetings 08 Jul 15:19
stickster There's been threads on fedora-devel-list IIRC. 08 Jul 15:20
stickster spevack: I can find some info for you to follow up if desired. 08 Jul 15:21
spevack 08 Jul 15:21
* spevack doesn't have anything else -- we're 20 minutes over time anyway, and there were no other questions from the public channel. The last few I've just made up myself. 08 Jul 15:21
stickster spevack: Now's a good time to call it then. 08 Jul 15:21
stickster All right, thanks everyone;. 08 Jul 15:22
spoleeba User:Kevin/DRAFT-IRCSupportConduct 08 Jul 15:22
stickster spoleeba: Thank you 08 Jul 15:22
spevack thanks all 08 Jul 15:23
f13 thanks! 08 Jul 15:23
stickster Thanks for coming, everyone. We'll post a log shortly at the normal page: Board/Meetings 08 Jul 15:23

Generated by irclog2html.py 2.6 by Marius Gedminas - find it at mg.pov.lt!