Oversight Board/2010/Meeting Log-2010-01-29

<cjb> morning
<tomeu> hi all!
<bernie> 'morning
I hope people are not late as usual!
I can't stand late comers
<SeanDaly> hey bernie :D
<cjb> :)
<SeanDaly> hi cjb, tomeu
<walterbender> Hi everyone. Shall we get started?
<meeting> Meeting started at 11:04 UTC. The chair is walterbender.
Commands Available: #TOPIC, #IDEA, #ACTION, #AGREED, #LINK
<walterbender> how about a quick rollcall?
  • walterbender:waves
  • cjb:here
<walterbender> cjb: did everyone else suddenly leave?
<SeanDaly> hi walterbender
<walterbender> almost a quorum :)
<bernie> here too
<walterbender> Here is the agenda:
<bernie> CanoeBerry, mchua: ping
  • tomeu:here
<walterbender> 1. review of TM discussion status
2. report from the OLPC meeting in Miami
  • mchua:here
<walterbender> 3. update from the infrastructure team?
anything else?
<mchua> not fromme.
<walterbender> #Topic TM
  • mchua:wants to thank SeanDaly for all the work he's done on the TM stuff this week, case studies page looks *awesome*
<walterbender> Sean, do you want to summarize where we are at the moment?
<cjb> we seem to finally have got people talking about trademarks on the list :) thanks for the case studies page!
<SeanDaly> thanks mchua sorry I only got to it last night
  • walterbender:just made some cosmetic changes to the case-study text
<SeanDaly> I feel the case studies page has indeed helped us to reflect on the policy
and, revealed differing points of view cf. today's interesting exchanges with cjb, Sascha and myself
I feel it is important to explain how I have arrived at my point of view
otherwise, we could vote in sthing others in the community could find they are in disagreement with
of course, not saying everyone has to agree,
but my vision is not necessarily the best one :D :D
at the risk of being strung up on the yardarm,
<cjb> SeanDaly: I think your vision is pretty understandable, for a marketing guy. It's normal that the free software hippies and the marketing folks don't see entirely eye to eye. :)
<SeanDaly> I feel it may be necessary to debate further, then to refine language before a vote
<walterbender> SeanDaly: it seemed to me that there are two threads: (1) written vs implicit permission and (2) are we being too restrictive
<cjb> SeanDaly: yeah, nowhere near a vote yet, let's keep the discussion going.
walterbender: that sounds right
<SeanDaly> cjb: the marketing people often think of me as geeky hippie
<cjb> SeanDaly: :)
<SeanDaly> walterbender: yes 2 issues
<walterbender> SeanDaly: maybe the way to deal with (1) is review language...
if we ask that they inform us on some schedule, like 1x per year, we can audit...
<SeanDaly> yes, my concept is a period, with renewal procedure
  • mchua:would like to hear more about where SeanDaly's perspective comes from
<bernie> cjb: we're lucky to have SeanDaly. he's the hippiest marketing guy we could have hoped for :-)
<walterbender> To me it seems that the important thing is we find out so we can ensure that the rules are being adhered to... but the ping back of written permission need not be a necessary step...
<tomeu> :D
<SeanDaly> an audit can be quick, for active community participants we know; or take more time to verify companies playing fair; or some research to find if somebody who has disappeared is elsewhere
<walterbender> of course, if they ask for written permission, we could give it.
<SeanDaly> walterbender: my concern is, someone ill-intentioned claims a mail was sent when it never was... end of story
<walterbender> so if they ask, then they get implicit permission for (fill in the blank)...
<sdziallas> walterbender: just wondering, has anybody already mentioned a "fair use" clause as Gnome's doing it?
<walterbender> sdziallas: no se
<SeanDaly> sure - boilerplate license agreement as Mel suggested, PDF
  • SeanDaly:greets sdziallas
<cjb> sdziallas: I believe the presence of trademark fair use exists outside of any trademark policy
the trademark policy can't override it; it's just there.
<sdziallas> (IANAL, but one could argue that "fair use" works as we define it - and once somebody comes up with bad intentions, it's no "fair use" anymore and they need permission.)
<SeanDaly> cjb: sure - any news outlet can publish a logo
<cjb> sdziallas: no, you can't do that.
<sdziallas> cjb: well, but you can define what "fair use" is, no?
<cjb> no.
<sdziallas> (hi SeanDaly at el, btw) ;)
<cjb> you can't define what fair use for copyright is, either.
dirakx (n=rafael@ has joined #sugar-meeting
<SeanDaly> sdziallas: I will look at that, however, I remain more interested in successful brands than unknown ones (to nongeeks)
<sdziallas> 0x01ACTION goes quoting: "You may make fair use of GNOME word marks to make true factual statements GNOME, or to truthfully communicate that your product or service is compatible with, designed for use with, or was designed using, GNOME. Assertions of compatibility, design, must, of course, be accurate. You cannot use GNOME logos or stylizations of GNOME word or other GNOME Trademarks unless you have explicit written permission to do so. You shou
  • mchua:would still like to hear where SeanDaly's perspective is coming from - as a marketer, and someone who deals with lawyers, what are the things that are usually concerning?
<mchua> (some of these are already written out, but... just to kinda sum them up in one place)
  • sdziallas:shuts up. (was just wondering) ;)
<SeanDaly> re fair use: lawyers & courts at work on that all the time
<mchua> 'cause I think many of the rest of us are hackers, and prioritize low-barrier-to-entry stuff over... almost everything else. "ask forgiveness, not permission" mentality.
<cjb> sdziallas: So, I think the ability to make factual statements comes from trademark law itself, not from that text of GNOME's that says that you have the ability; it would be there either way.
<mchua> which, obviously, doesn't really work the same way here.
<SeanDaly> mchua: well... former developer, current marketer, with a whiff of legal after years at Groklaw
<walterbender> sdziallas, SeanDaly: in any case, GNOME asks that explicit permission be obtained (see 5.)
<sdziallas> cjb: heh, that makes some sense, indeed. but I'd bet that both Gnome and Fedora had some lawyers thinking through what they were doing.
<SeanDaly> let's put it this way: how do you deal with abuse
<cjb> sdziallas: sure. it's helpful of GNOME to show that they're adhering to the common understandings of fair use.
<SeanDaly> any lawyer will tell you is much easier when explicit deal broken
<sdziallas> walterbender: yup, right.
<bernie> SeanDaly: you send a cease & desist letter?
<mchua> SeanDaly: moving from developer to marketer (with a whiff of legal), sounds like something that needs to be learned to make that mindset switch is how to do the hypothetical "what if there were Bad People Everywhere?" thought experiment?
<SeanDaly> bernie: hee hee I don't, the SFC does
<CanoeBerry> SeanDaly: a web page can be set asking people to agree to the legal agreement. Period.
They click "I Agree"
<cjb> (my answer to how do you enforce -- by encoding your constraints in the trademark policy, and sending C&D if they aren't followed.)
<CanoeBerry> Done
<bernie> CanoeBerry: I agree
CanoeBerry: sorry for the pun
CanoeBerry: Ubuntu does it by asking you to send the agreement back by email, signed with gpg
<cjb> bernie: ooh, that's a nice step
<bernie> CanoeBerry: it's much more legally binding, I guess. and only slightly more annoying
<SeanDaly> CanoeBerry: that's fine if we don't care about checking them out... but we do
<cjb> so you send a mail to automatic-trademarks@, it sends you back a copy of the policy, which you gpg-sign and e-mail back
<bernie> I'm totally against sending papers by snail mail like the fsf does. they're leaving in the 20th century.
<walterbender> It seems like we are converging on a closed-loop process, regardless of the details of how the loop is closed.
<sdziallas> cjb: Fedora does that for the CLA, too - interesting.
<cjb> yeah, I think this mostly needs time to digest on the lists. We probably shouldn't argue much here, it's better to have the conversation on the lists anyway.
<SeanDaly> bernie: I wouldn't second-guess the FSF's methods, they have been extremely effective guardians of the GPL and its freedoms
<bernie> now that we all agree on the process, we'll probably want to move on arguing on the actually contents of the agreement :-)
<SeanDaly> Perhaps one way to put it is: should license be a mere formality or not
<cjb> :)
SeanDaly: I don't think that's the issue at all
<SeanDaly> In my view, it should not
<cjb> you can have a strong set of trademark constraints, just as the GPL has a strong set of copyright constraints
<SeanDaly> because we want to grow ecosystem
<cjb> and they can be available automatically
just because the GPL is automatically transferred, that doesn't mean it's "a formality"
you just said you think it's being successfully enforced :)
<SeanDaly> if licensing a clickthrough formality, discussion is discouraged
I'd much rather we learn from & exchange with potential partners
<cjb> I'd like us to do that too
<SeanDaly> in particular classroom feedback
<bernie> SeanDaly: the nice thing about granting a license is that, if it is void for any reason, then the third party has no license whatsoever and is using the work in violation of copyright.
<cjb> do we have to threaten them with legal stuff to get them to talk to us?
bernie: for a copyright license, at least, yeah.
it's the nice thing about a license that encodes its constraints and revokes itself if they fail to be met.
<SeanDaly> bernie: sure, but a trademark much easier to damage
that's why copyright & trademark law distinct
<bernie> cjb: I thought (maybe incorrectly) that a trademark license would work pretty much the same.
<cjb> bernie: yeah; I just sent a mail out to the lists that makes the same claim
<bernie> cjb: that is, you retain all rights unless you grant a valid license to someone on conditions you decide.
ah ok
<SeanDaly> no, what is commonly called "intellectual property" (term hated by FSF) is meant to cover very different legal contexts of copyrights, trademarks, patents, and trade secrets
<bernie> SeanDaly: yep
<SeanDaly> trademarks are a special animal: very hard to build a brand, very easy to break it
usual brand building is done with massive advertising & promotion expense
we have to be smarter since we're poorer
since brand easily breakable, much more caution necessary; very different from free sofware copyright context
<walterbender> are we stalled out for the moment?
<SeanDaly> our balancing act needs to be: spread awareness, build up brand values, avoid confusion
<cjb> SeanDaly: we weren't trying to draw a parallel between trademark and copyrights in general, just in that it is possible to write licenses for both that are automatic, declare what needs to be done for the license to be/stay valid, and revoke itself if the conditions are no longer met.
<SeanDaly> walterbender: well, I believe there is still discussion to be had on the lists
<walterbender> can I suggest an action to be continuing for another week on the lists?
<cjb> walterbender: probably more than, even. seconded.
<walterbender> but with a focus on the two issues: (1) process and (2) scope?
<SeanDaly> cjb: again, what works for copyright won't necessarily work for trademark protection
walterbender: ok
<bernie> walterbender: +1
cjb: what you said
<walterbender> OK. I'll try to summarize today's discussion for the lists...
<cjb> walterbender: thanks
what's up next?
<walterbender> #ACTION: Walter to summarize latest TM discussion for the lists...
#TOPIC: trip report: OLPC meeting in Miami
<SeanDaly> here's an analogy: a famous golfer spent years cultivating an image of prefection which was wrecked in three days...
<bernie> I guess we all agree on the "gpg signed agreement filed by email or web form" part
<cjb> bernie: yeah, I just mailed that idea out too for discussion
<SeanDaly> walterbender: was unaware you were down there! great
<walterbender> I spent 3 days in the OLPC office in Miami.
It was a meeting with representatives from many of their major deployments.
<bernie> walterbender: yeah I hope you had a good time with cecilia
she should be back soon
<walterbender> I got a chance to talk about Sugar and Sugar Labs and listen to reports from the deployments, including a report on the amazing work in Paraguay
There is great enthusiasm for Sugar throughout the organization and with all of the deployments.
And they want to work more closely with us.
<SeanDaly> great
<walterbender> A few concrete things came out of the meetings:
(1) They reiterated their desire to work with us on marketing. Sean, perhaps it would be timely to ping Rodrigo and Chuck again, while it is fresh in their minds.
<SeanDaly> walterbender: will do (Rodrigo had never answered my mails)
<walterbender> (2) We discussed getting more direct feedback through OLPC channels, not just hit-or-miss with the deployments themselves.
Claudia agreed to join our Design meetings, which will add important on-the-ground perspectives.
<cjb> this sounds pretty promising
let me know if I can help poke anyone to follow through :)
<walterbender> (3) We agreed in principle (but need to work out the details) of a model where by OLPC would offer up to deployments some Sugar developer time to help est. "local lab" teams for contributing to Sugar and FOSS.
Sort of like developer training to complement teacher training, to be partially subsidized by OLPC and partially subsidized by the deployments.
<SeanDaly> fabulous
<walterbender> In other words, they will help us get Sugar teams est. as part of the sales and deployment process.
(4) They offered to host the next Sugar camp in Miami.
<bernie> things are coming back together nicely
<walterbender> (I have access to several apartments there with couch and sleeping bag space, and beach access :) )
and they have a great work space we can use.
<mchua> walterbender: any notes on timeline for that? this summer, etc?
<bernie> walterbender: are there public minutes or slides from the meeting?
<SeanDaly> wow great Miami
<walterbender> We need to get back to them with when we want the next camp...
<bernie> walterbender: and the list of deployments and other participants?
<walterbender> bernie: They took extensive notes... I don't have copies yet. I'll see what I can dig up.
<tomeu> perhaps we need to set our HQ there, instead of a sugar camp :p
<SeanDaly> in alligator country
<walterbender> tomeu: it is very nice there... and very easy access for Latin America.
I have a lot of follow up to do, but all in all, it was a very productive 3 days.
<bernie> walterbender: did someone explained why olpc is moving there from boston, btw?
<CanoeBerry> Moving ain't quite the word :)
<walterbender> #Action: schedule the Miami Sugar Camp (after discussions on the list)
<bernie> CanoeBerry: ah, it's a branch?
<walterbender> here is my understanding:
<CanoeBerry> P2P network ;)
<walterbender> the Association has moved to Miami.
the Foundation remains in Cambridge
the Association is primarily dedicated to deployments
the Foundation is primarily dedicated to fund-raising and engineering
CanoeBerry, cjb: did I get that right?
<CanoeBerry> Good enough.
You prob saw the OLPCorps and Interns announcements in the last week.
<cjb> yep
<SeanDaly> walterbender: still a mystery to me how press relations & marketing works with two OLPC structures
<CanoeBerry> Many such "deployment" stuff is spread between Miami and Cambridge.
<walterbender> There seem to be a few deployments run from the foundation, but mostly it is association
<bernie> CanoeBerry: url?
<cjb> bernie: I'm not moving anywhere :)
<CanoeBerry> http://blog.laptop.org
<SeanDaly> OLPC Europe does PR too aside from sales, cf. press interviews
<cjb> walterbender: I guess the ones being run with donated laptops would be foundation, association is handling all finances for deployments I think.
<walterbender> I think OLPC Europe reports to the Association (at least loosely)
In any case, the point is that there is a deployment-facing team in Miami that wants to work closely with us.
<bernie> walterbender: YAY
<mchua> Any decisions that need to be made (or that will need to be made) SLOBs-wise for that?
like "what does a partnership mean, do we need to sign any pieces of paper" etc?
<walterbender> No decisions... just a trip report.
<CanoeBerry> (Of course, it's convenient that Claudia actually in Cambridge ;)
<walterbender> We may have to make some decisions as the details of how a local Sugar Labs agreement might come into exisitance.
<tomeu> ah, claudia is in cambridge?
<CanoeBerry> Yes, Claudia lives in the Boston area.
<tomeu> walterbender: maybe she would like to get involved in the experience testing in ny?
<walterbender> but she works closely with the Miami team...
Next up, infrastructure? Since Bernie is here?
<bernie> good
<walterbender> #topic infrastructure
bernie: can you just give us a quick status report?
<bernie> We're working with RIT to get 3 old servers racked there
the servers are being donated by the Wikimedia foundation
lfaraone is working on getting the machines shipped to RIT
  • dogi:waves bernie
<bernie> RIT's Prof. Stephens got us the hosting offer
our contact at the CSH is decause, who lurks often on #sugar. we met him at FUDcon.
the machines may turn out to be really bad, we don't know yet
in which case, we'd need a contingency plan
such as acquiring the ~$3000 machine we were planning before this opportunity even came up
the cluster at RIT will run ASLO, which is probably our most important service.
reliability is very important to us
<SeanDaly> may I ask a silly question?
<bernie> dfarning has a good plan to spread the load on these 3 machines in such a way that any one of them could die without stopping the service
SeanDaly: sure
<SeanDaly> is there an advantage to racking & maintaining directly as opposed to a service? just cost?
<bernie> dogi will be on site to coordinate the project: racking of machines, OS installation, etc.
<dogi> if somebody wants to know more details, here are the logs: http://me.etin.gs/treehouse/treehouse.log.20100126_1601.html
  • mchua:appreciates this infrastructure report - very helpful to know what's going on, and it helps us make budget decisions (like the one last week)
<bernie> SeanDaly: gregdek of red hat offered some free amazon EC access for us... but it's not clear when it will become available
SeanDaly: I'm lurking on a mailing list where people are coordinating to develop a good Fedora image for us to use
<tomeu> virtual machines hosted on someone else's computers has caused problems in the past to us, right?
<mchua> bernie: I thought the blocker on that was "when we can get F12 images on amazon"
[scs] (n=scodas@mail.paraguayeduca.org) has joined #sugar-meeting
[scs] has left #sugar-meeting
<bernie> SeanDaly: if it weren't for Red Hat, I'd disregard any existing cloud service out there due to costs.
[scs] (n=scodas@mail.paraguayeduca.org) has joined #sugar-meeting
<mchua> bernie: I think (I don't know for sure, but we could check) that requirement came from your side of things, not RH's.
<bernie> SeanDaly: our services are heavyweight. the cloud stuff is meant for small websites
<SeanDaly> ok clear thanks
<dogi> I think the most important thing is the relationship with RIT/CSH and the students which will help us there ...
<bernie> mchua: exactly. F12 images are not yet available... it seems to be an Amazon-side issue as well as a Fedora-side issue
<mchua> Ok. So bernie, that's basically the blocker for... no matter where it comes from, that's the only blocker?
<bernie> mchua: RH is of course also interested in getting F12 on EC2. Now they have F8, which... sucks.
<mchua> oh geez that's old.
<bernie> mchua: no. there's also the cost issue. I don't think greg can set aside so much money for us...
<mchua> bernie: the cloud SIG in Fedora is probably the place where that work will happen, if you want to watch. http://gregdekspeaks.wordpress.com/2010/01/20/announcing-the-fedora-cloud-sig/
<tomeu> now that we are so offtopic, is it know on which fedora relesae will be based the next rhel?
<bernie> mchua: unless red hat has a special contract with EC2
<mchua> tomeu: you're right, we're offtopic... sorry, back to agenda
<walterbender> we should prob. end the meeting...
<bernie> mchua: I'm expecting ASLO alone to cost some $3000-4000 per year.
<mchua> walterbender: any notes for what we want to do next week?
TM again, I'm guessing, after the community gets another feedback round in?
<walterbender> mchua: next week, I'd like to discuss our 2010 goals
<bernie> mchua: yes I'm subscribed to the Cloud SIG mailing-list, but I'm skeptical that we can ready in two weeks or so
<SeanDaly> TM list discussion continues even as we speak ;-)
<walterbender> bernie: any action you need from us at the moment?
<mchua> walterbender: +1
<bernie> It's quite important to be ready in 2-3 weeks because SCHOOLS WILL OPEN IN LATAM!
  • mchua:is awed by the TM list thread, it's AWESOME
<SeanDaly> 2010 goals very important
<bernie> If we're not prepared to deal with the traffic, infrastructure will collapse
<mchua> does everyone have homework to come with their individual list and try to get others in the community to write theirs?
(individual thoughts on individual 2010 goals, sl 2010 goals, and how the two might map?)
<tomeu> bernie: I don't understand yet why we don't cache more of aslo
<walterbender> (when the .uy kids come back from vacation, they'll all want to download the new version of Turtle Art :)
<tomeu> do we need to have it so dynamic?
<bernie> Dfarning thinks that working with RIT would also help develop important partnerships for the future. They want to get students involved.
<walterbender> (but the new TA is only 10% of the footprint of the old one :) )
<bernie> tomeu: bandwidth is not an issue (and downloads are cached anyway)
<davidfarning> I would appreciate feed make on my working with Bozano
<tomeu> bernie: no, caching processed pages
<bernie> tomeu: or did you mean caching web pages? not really feasible
<tomeu> bernie: so they don't hit the php engine
  • SeanDaly:greets davidfarning
<bernie> tomeu: it's not a wiki, it's more of a search engine
tomeu: caching works best for applications servers, where most of the hits are actually to static content (js, png, css...)
<tomeu> bernie: ok, but the same url gives the same content 90% of the time, right?
<bernie> tomeu: in the case of php, those are *already* being served quickly by apache without bothering php
tomeu: so, the load we see is really caused by web pages we'd have toserve
<tomeu> bernie: not sure, I think most of the people go to a few urls, and those present the same info to all of them
<bernie> tomeu: we can descend into tech details later on #sugar, but caching the search results, etc is extremely hard. links would break randomly, etc.
<cjb> we use a cache for olpc's wiki
<tomeu> ok, I trust you
<bernie> tomeu: it also really doesn't make sense because a normal $3000 machine can easily do 8-10x the current load
<cjb> the wiki proper sends a cache invalidation request to the slaveswhen a page changes
<bernie> cjb: yes, mediawiki is easy and profitable to cache
cjb: but remora is a different kind of application, I don't believe it's equally easy to cache. people are welcome to try, though.
so, from an exclusively practical PoV (reliability, simplicity, cost), I'd recommend buying a machine
from a strategical pov, RIT is interesting too
<dogi> bernie, think the RIT students will try :P
<SeanDaly> I'd be in favor of buying a machine and contributing C-notes for it
<bernie> dogi: yeah, I have faith in their commitment, especially since prof stephen jacobs is backing them
<SeanDaly> experience has taught me that whenever you think you have enough capacity, a tsunami of visitors arrive
<mchua> walterbender: we're getting kinda late... assign goals homework, #endmeeting?
<bernie> SeanDaly: lol
I was going to ask the board to pre-approve a budget of $3000 for the infrastructure, which I'll use only in case the current plan fails.
if we set aside the money now, I'll be able to move quickly in case of need
<SeanDaly> I am favorable
  • mchua:as well.
<bernie> btw, we also need to start thinking on replacements for Solarsail. Ivan and I think that it should be dismissed soon or later.
<walterbender> Ok. it is a wrap.
<cjb> bernie: generally favorable, but we don't have a treasurer so it's hard to know how to prioritize. this is maybe 50% of our finances.
<walterbender> thanks everyone for attending today !
<cjb> (I mean, we don't have a process for comparing this request against any others we might get.)
<bernie> cjb: oh
<SeanDaly> thanks everyone
<bernie> cjb: I frankly though we had more non-encumbered money
<walterbender> I think the #actions speak for themselves...
<bernie> btw, maybe next week we should discuss the topic of fundrising
<walterbender> next week, more TM, goals, and perhaps a discussion of some of dfarning's questions.
<mchua> bernie: add to agenda queue :)
<cjb> bernie: It looked like about $5000 to my reading, $25k - $20k for the pilot, and Walter said $7k at one point I think.
<SeanDaly> bernie: I have been mulling over a plan for that
<mchua> walterbender: and we're all supposed to come with goals drafts to compare?
<bernie> SeanDaly: great!
<walterbender> mchua: or comments on the goals I drafted
<mchua> okeydokey.
<walterbender> cjb: the 7K figure is about right... but we may ask the Gould project for infrastructure help...
  • mchua:thinks we're getting some good momentum going here
<SeanDaly> wish me luck for Wednesday's presentation
<walterbender> SeanDaly: you'll do great...
<SeanDaly> Foundations and CSR execs from France's top corporations
with OLPC France
<walterbender> SeanDaly: shout if you need any copy-edit help :)
<bernie> dogi: in case we're stuck with no RIT hosting and no new machines, can we rely on treehouse/housetree to cope with the extra load?
  • mchua:has to pop out now - thanks for a good meeting, folks.
<walterbender> #endmeeting
<meeting> Meeting finished at 12:22.
Logs available at http://meeting.olpcorps.net/sugar-meeting/
<SeanDaly> walterbender: will do thanks
<tomeu> SeanDaly: good luck!
<bernie> dogi: it will not require extra IPs, but it will sure put a *lot* of load on those machins
<dogi> bernie, sure ... but remember treehouse/housetree was designed to be construction ground ... and since they are built your own servers, this would be only a temporary solution