[Grassroots-l] Project Management Tools meeting in #olpc

Bastien bastienguerry at googlemail.com
Sat Aug 16 08:32:27 EDT 2008


Sameer Verma <sverma at sfsu.edu> writes:

> Sorry, but I couldn't make it. Meeting conflict on campus. Does anyone
> have notes/irc logs?

Here is what I have:

[18:03]*** You have joined channel #olpc

       *** Topic for #olpc: Web support: http://forum.laptop.org | IRC
           support: #olpc-help | Deployment:
           http://radian.org/notebook/first-deployment | GUI,
           activities: #sugar | Content: #olpc-content

       *** #olpc: topic set by neuralis, 05:44:05 2008/02/04

       *** Users on #olpc: bz_g jdub johnf garycmartin nessy dsaxena
           wcohen_ CanoeBerry Hiro2 greebo ivazquez Lerc_tablet webar7
           Sargun homunq eightdot reynaldo_ J5 faw xuxa nteon behdad
           isforinsects scorche|sh unmadindu hpachas-PE rgs_ joern aa
           pschulz01 rennj kfm82 Neskaya nullwork nullman pepe__ nealmcb
           xobot mether Kevin` dermoth _sj_ Renault mstrobert dwmw2
           blisto1 lars zmedico _bernie philipl paulproteus sjoerd RP
           cjb tarbo_ toi mako eMBee blistov svu araujo

       *** Users on #olpc: h01ger Robot101 bluehole clajo04_ bluefoxicy
           Ian_Daniher_ keithy bbs mpr scorche par gnu{- xjjk jwhite pgf
           christel fragalot mburns saigo mtd YChao_ pepse dogmeat
           kristian-m dirkx_logon sm|CPU bigeasy lkundrak ThomasWaldmann
           danjared [knurds]adc fche cassidy Zdra coderanger orospakr
           besonen_mobile__ smithbone m_stone IDoNotKnow ffm|sh shenki
           PecisDarbs nickr tglx Blaketh Lerc bdale nrp erikg
           summatusmentis dgilmore 

       *** #olpc modes: +tnc

       *** #olpc was created on dimanche 2006/11/26 01:42:47

       <isforinsects> good morning/evening jdub

       <johnf> hi all

       <nessy> hi :)

[18:04]<isforinsects> garycmartin, I'm not ignoring your last email re
       SOM's, I just haven't had time to parse the health stuff.

       <isforinsects> hi johnf, nessy

       <garycmartin> isforinsects: no worry, it's a busy timejust nowfor
       all I'm sure!

       <greebo> let's start, isforinsects do you want to run it or me?
       :)

[18:05]<garycmartin> isforinsects: excuse my missing spaces, you can
       guess mt HW :-)

       <isforinsects> You know more about the requirements than me.

       <_sj_> hiya greebo

       <greebo> _sj_, heya!

       <bz_g> aloya

[18:06]<greebo> ok, how about we start with briefly isforinsects or _sj_
       recapping what hq are trying to do, then we can touch base on
       what we are tring to achieve and go from there?

[18:07]<isforinsects> Let's have _sj_ recap, I don't know most/any of
       the details of projectdb

[18:08]<greebo> isforinsects, I mean more about a recap on what hq are
        trying to do with getting consistency of what the regions are
        doing

       <greebo> let's talk about projectdb after we have a clear picture
       of what everyone wants :)

[18:10]<_sj_> sure

[18:11]<_sj_> olpc wants to provide a steady stream of hardware to
              contributors, esp. developers

       <_sj_> including XOs and active antennas

       <CanoeBerry> and serial adapters

       <_sj_> and to hand off the process of identifying good
       projects/conributors to the community

       *** jrb (n=jrb at 68-116-198-66.dhcp.oxfr.ma.charter.com) has joined
           channel #olpc

       *** dwmw2 (i=ctrlprox at baythorne.infradead.org) is now known as
           dwmw2_gone

[18:12]<_sj_> the current projectdb is too opaque

       <_sj_> it was set up to catch personal data from contributors

       <_sj_> but most of the project-definition process, including
       reviews and suggestions about where to get hardware

       <_sj_> (your local community SIG, another person at your
       institution, a national chapter, via the contributors program)

[18:13]<_sj_> happens before personal data is needed, and can/should be
       public

       <_sj_> a larger-scope projectsharing program to help people
       announce projects, find collaborators, and work on
       testing/outreach/development with others

       <_sj_> is related to this data (and sounds closely related to
       what you have in mind, pia)

[18:14]<_sj_> currently we have this set of db scripts that maintains
       submissions from projectdb.olpc.at, but the code is hard to hack
       on and stored on dev.olpc.at

       <_sj_> hard as in -- you have to send a privkey to one of the two
       machine maintainers, who are busy...

       <_sj_> it hasn't seen many people hack on it, mainly c. hager and
       a. kaplan to date

[18:15]<greebo> because it is so hard to find to hack on :)

       <_sj_> its other main prupose is to produce neat xml exports (and
       import shipping updates from brightstar once they confirm
       shpiment)

       <_sj_> so that we have a better-ordered set of email alerts as
       the process of getting hardware proceeds

[18:16]<_sj_> right now we have to confirm manually that machines have
       gone out, and that they've arrived; all of which should be
       scriptable (b* provides in the end the tracking #, fedex and ups
       provide scrapable data about shipment status on their websites)(

       <_sj_> an open question is how much info about the shipping state
       of various bits of hww should be published; I expect most
       everyone would be glad to have a public list of whether their
       gear is in transit, arrived, or not yet shipped, but we might
       spotcheck.

[18:17]<_sj_> (also : about 1% of shipments get bounced back, with the
       phone # provied not sufficient to get the contribuytor to come
       pick their gear up, something that a bit more process transprancy
       could cure)

[18:19]<greebo> _sj_, we'll just be a sec

[18:20]<nessy> discussing :)

[18:23]<nessy> Ok, let me recap

       *** J5 (n=quintice at c-76-24-17-105.hsd1.ma.comcast.net) has quit:
           "Ex-Chat"

       <nessy> what we understand is that your system is essentially a
       fully automated ordering system for OLPCs

       <nessy> once a project is started, the order can be tracked and
       all

       <isforinsects> s/OLPCs/XOs

       <nessy> sure :)

       <isforinsects> :) sorry

[18:24]<nessy> at the same time you'd like to have a simpler means for
       local groups to suggest worthwhile projects to you

       <nessy> so if there is a local organisation that can make a good
       decision on worthwhile projects and suggest them to you, that
       will make your life easier

[18:25]<nessy> is this correct so far?

       <isforinsects> The process really isn't automated is the issue.
       A database entry gets created, but it's still hand managed.  And
       it's only really for developer/contributors, not for the actual
       ordering of Give-Many machines.

       <nessy> I'm talking about what the goal is :)

       <isforinsects> As-stands

       <greebo> from what I understand from the phone call you also want
       transparency into projects once laptops are granted to get public
       access to documentation and knowledge, right?

       <isforinsects> Yes and yes

[18:26]<nessy> cool

       <nessy> that really fits well with what we have been discussing

       <nessy> because we were discussing the system that we need for
       local people to propose projects, get them reviewed and started

       <isforinsects> as an aside, I would really like to see tools for
       smaller groups to aggregate funds, track resources and do some
       private/public planning.

[18:27]<nessy> yes, that's what we need :)

       <isforinsects> But a lot of people have suggested that this is
       something that exists.

       <greebo> there are a few other things we need

       <isforinsects> via the email thread

       <_sj_> nessy : a part I didn't add is that local gropus are
       organize their own versions of XO sharing/tracking

       <nessy> yes

       <_sj_> and one of the reasons projectdb was started was to help
       identify who had machines that were for a fixed-length projet

       <_sj_> that would be done soon.

       <_sj_> so every project has a duration...

       <nessy> I have over-simplified things to make a point :)

       <_sj_> when you're done you agree to hand the machine off to
       others who might need it

[18:28]<greebo> we are also wanting a way to register, track and
       allocate volunteers to projects, a way to share laptops in an XO
       library style arrangements where owners of laptops can allocate
       them as shared resources for loaning to developers or for demos.

       <greebo> so the latter of those ideas fits well with what you are
       saying _sj_

[18:29]<greebo> so we have a need for 1) project requests (including XO
       orders, tracking and such), 2) volunteers rego and 3) xo library

       *** adricnet (n=adric at atlanta.national-net.com) has joined
           channel #olpc

       *** jg (n=jg at c-76-19-89-183.hsd1.ma.comcast.net) has joined
           channel #olpc

[18:30]<jdub> hey jg

       <greebo> but for the "projects" we don't mean the running of a
       project, we will encourage people to use our forums for planning,
       and wiki.laptop.org for documenting

       <greebo> jg, heya! *hugs*

       <_sj_> right

       <_sj_> hi jg, adric

       *** adricnet (n=adric at atlanta.national-net.com) has quit: Client
           Quit

       <jg> hi jdub, greebo...

       <jg> how's it down-under?

[18:31]<isforinsects> Did everyone see Sameer Verma's post to the
       grassroots list?

       <jg> wish the rain would go your way....

       <isforinsects> About tools

       <greebo> jg, great, had a week doing the Niue XO/XS rollout,
       loads of fun!

[18:32]<greebo> isforinsects, no, could you please recap for everyone?

       <nessy> _sj_: how much software for these goal exists?

       <greebo> isforinsects, forgot the confirmation email, doing now

[18:33]<greebo> isforinsects, can you post the link?

       <_sj_> nessy: not much.

       <johnf> this one?
       hTtp://lists.laptop.org/pipermail/grassroots/2008-August/000619.html

       *** adricnet (n=adric at atlanta.national-net.com) has joined
           channel #olpc

       <_sj_> in my mind, half of this is producing the right
       wikitemplates

       <isforinsects> That's the one

       <isforinsects> I was just about to get that

       <_sj_> since that's something everyone can update

[18:34]<_sj_> nessy: just a simple data model for project descriptions
       and associated people (which is actually far simpler than the
       real world of collaboratoin)

       <_sj_> since the model for collab, project duration, and how much
       access to (one another, materials, testers) will be quite
       different from project to project

       <_sj_> it might make sense to start with a very open fram where
       participants can describe what they are doing

[18:35]<_sj_> and what they need

       <_sj_> and move towards a solution that works for them

[18:36]*** mstrobert (n=mstrober at wkstn.wycliffe.ca) has quit: "Leaving"

       <greebo> _sj_, for project status, documentation, etc the wiki is
       great, especially with soe great wikitemplates

[18:37]<greebo> for local management of project requests, volunteer
       rego, and the xo library, probably not

       *** adricnet (n=adric at atlanta.national-net.com) is now known as
           adric_work

[18:39]<jdub> _sj_: so what we're thinking about doing is a "local olpc
in a box" system, which will handle these more local tasks, but also
talk to your backend

       <jdub> _sj_: so local orgs can share heaps and focus on
       scratching their itch

       *** hpachas-PE (n=hpachas at 200.37.120.18) has quit: "Saliendo"

       <jdub> _sj_: while you guys can define the way they interact with
       HQ

[18:40]<_sj_> greebo/jd: if you replace "HQ" with "global olpc orgs"
       then I agree with you :)

       <jdub> heh

[18:41]<_sj_> the important group to collab with is the global network
       of other local groups, individuals who don't have a local group
       yet, and others who may want to help develop this sort of infra

       <jdub> sorry, still haven't found a quicker way of saying that
       than HQ, so it has stuck ;-)

[18:42]<jdub> we're also solidly focused on pushing as much as we can
       towards central resources (we won't have a wiki or code repo at
       all, for instance)

       *** andresambrois
           (n=aa at r190-135-178-99.dialup.adsl.anteldata.net.uy) has
           joined channel #olpc

       <_sj_> well, the OLPC Office has its own needs.  the community of
       people trading XOs for "nonlocal olpc in a box" is separate.

       <nessy> _sj_: you mentioned above that you'd like to see a system
       in place that helps local groups get themselves organised

       <_sj_> jdub: sounds good

       <nessy> I guess what we are saying is that we will be developing
       such a project :)

[18:43]<nessy> s/project/system/

       <johnf> _sj_: I suppose it makes sense that while AU and larger
       organisations will run an instance of the "frontend" there should
       maybe be a global instance as well for much smaller communities t
       be able to make use of

       <greebo> _sj_, it isn't just about trading XOs, it is about
       managing regional project requests in a sane way

       <_sj_> nessy: so the overall db should include who has machines
       available where

       <greebo> and other regional tasks

       <_sj_> but some of the entities who have laptops could be local
       xo pools

       <_sj_> with their own contact info

       <_sj_> which you need to poke to find out about short tem borrows

       *** aa (n=aa at r190-135-178-99.dialup.adsl.anteldata.net.uy) has
           quit: Read error: 104 (Connection reset by peer)

[18:44]<_sj_> if you want something for 2 months, you can get it with a
       little lag through the mail.  if you want to test for 2 days, you
       may want to fin d a local group to visit

       <_sj_> greebo, sure

       <_sj_> do you have a set of examples of project requests?

       <jdub> basically, some local organisations will be able to run
       libraries, while there will be a global library/request system
       too

       <_sj_> do you just mean "projects with hardwrae needs" or also
       projects with needs for writers, developers, children,
       wireless-free cages, etc?

[18:45]<_sj_> jdub: that would be nice -- the loca llibraries publish
       overviews of their state -- what they have available -- and you
       can visit them for more details or to incubate a project idea
       until you /really/ have your own needs

       <greebo> _sj_, ok there are the sort of requests that might be a
       local hacker who wants to develop or test an app for a month, and
       they could submit a request and another person who owns and has
       registered their XO could allocate theirs to that person for a
       month

       <blistov> anyone know why koji.fedoraproject.org is down?

       <jdub> we're having issues with overloading of the term 'project'
       as well

[18:46]<blistov> should my yum repos.d be changed to reflect a new
       server?

       <_sj_> there's also the common issues of a) wanting to test
       coillaboration, briefly -- without having a few machines
       yourself;  b) wanting to test something in a sugar-lab with a
       particular build (can be on any sort of machine)

       <greebo> there is a second project request which is a primary
       school wants to trial 50 laptops and purchase them for the trial
       which we (as the regional org) would approve if appropriate and
       then submit the order as part of a bulk order to hq

       <_sj_> c) wanting to test something with a network mesh; d)
       wanting to test something in a classroom

       <_sj_> all things that work better when you have a local pool
       with some contributors poermanently at a uni or research grp,
       some working in a school, &c

[18:47]<greebo> yeah, we can register a lab of xos as a resource people
       could use

[18:48]<_sj_> greebo: helping a school trial 50 laptops is an excellent
       example of something I'd like us to start supporting, which is
       easier to do with such a support-chain is built in

       <greebo> _sj_, but I just want to check that you understand we
       are looking at a multi-part system that includes the "library"
       (loaning system) as well as the project registration (namely
       deployments/trials) and a volunteer rego system?

       <_sj_> at which point more details matter -- site survey and
       assessment of loca l needs; specific point people external to the
       project group who are signed up to help them, &c.

[18:49]<_sj_> yes

       <_sj_> well, volunteer repo is something different again

       <_sj_> rego

       <jdub> _sj_: our definition of library includes individuals
       loaning their resources to the pool as well, btw

       <_sj_> greebo: say more about that?

       <greebo> cool, yes and having the people and laptop registration,
       so people can request and allocate those as resources will help
       with both library requests and project requests

       <jdub> (lots of aussie floss dudes happy to share their olpcs for
       good projects)

[18:50]<_sj_> the number of people who want to contribute in a way that
       involves a large project or may need hardware is in the thousands

       <_sj_> the number of volunteers who may want to register to help
       in some way could be 100 times that

       <_sj_> different systems.  I don't know what the solution is for
       the latter.

       <blistov> what is the supported yum repo for xo?

[18:51]<blistov> repo's aren't included in any of the recent builds.

[18:52]<_sj_> greebo: I would keep that discussion open, but separate it
       from something specifically geared towards
       library/project-registration

       <greebo> yes, and one of the components of what we want to build
       would be the ability for volunteers to register their details,
       what they can do, how much time they can commit, and then we (the
       local org) can allocate volunteers according to regional
       deployment requests (which could request volunteers and/or
       laptops). This is in particularly important for the Pacific OLPC
       deployments happening at the moment.

       <greebo> I'm currently doing some volunteer coordination for the
       Pacific manually :)

[18:54]<nessy> it's to stop greebo from going crazy ;-)

       <greebo> _sj_, what do you mean? We want to build something to
       meet our needs (as a regional project) to scratch our own itch,
       aka 1) deployment requests for volutneers/laptops, 2) a way for
       volunteers to reigster and a 3) a way to share laptops between
       owners and people who need laptops.

       <greebo> we want to ensure that we push people to commit their
       information about projects upstream to wiki.laptop.org where
       possible

[18:55]<greebo> and we want to provide a consolidated way for HQ to deal
       with regional laptop requests

       *** jg (n=jg at c-76-19-89-183.hsd1.ma.comcast.net) has quit:
           "Ex-Chat"

[18:56]<greebo> does this make sense? we don't want to extend your
       system to do everything, but rather create a system for local
       orgs to use and integrate with yours

[18:57]<_sj_> why are you tying volunteer reg to deployment requests?

       <_sj_> that's my question :)

       <greebo> because a lot of deployment requests need volunteers :)

[18:58]<greebo> so that a project in Solomon Islands can register to get
       some people to volunteer time to help them, as they already have
       their laptops involved.

[18:59]<_sj_> that makes sense

       <jdub> well, also they're not one-to-one tied

       <blistov> cjb: you around?

       <jdub> but there is a relationship there :-)

       <_sj_> so, my comment about keeping separate is : data about
       volunteers, which are numerous, and which are 1-to-many with
       projects

       <_sj_> is different data than what you want to gather about
       deployment-leads

       <jdub> yeah

       <_sj_> fewer in number, needed info is often more private, tied
       very closely to location/hardware

[19:00]<_sj_> I'd like to help the discussion and development of ways to
       support both of those issues for local groups similarly distinct

       <_sj_> precisely to avoid trying to build one system that does
       everything

       <nessy> so we have actually three different issues to deal with
       around "projects": 1) project proposals, 2) deciding on support
       from headquarters, 3) running the project (with volunteers,
       borrowed laptops, bought laptops etc)

       <_sj_> yes

       <_sj_> 1) needs some bot-assisted way to update wiki pages from
       data that exists elsewhere

       <nessy> we are locally interested mostly in 3)

       <jdub> i think part of the confusion here is related to the word
       'project'

[19:01]<_sj_> I would refactor 2 and 3

       <jdub> which we have grappled with on this end (even in high
       bandwidth voice/face environment) :-)

       [Got disconnected...]

       <_sj_> and potentially, support needed to direct project requests
       to olpc

[19:03]*** kalila (n=kalila at 84.13.75.52) has joined channel #olpc

       <_sj_> (say, a group that wants HQ to do something that is
       time-consuming for OLPC staff; working out all the kinks so that
       the scope and duration of the request is clear)

       <_sj_> and 3) involves updates on the project, facilitating
       communication with volunteers signed up, tracking local support
       (perhaps also budgeting, timelines, other shared tools)

[19:04]<jdub> right, i think (2) is where we have an expression issue

       <jdub> and slightly different perspectives

       <_sj_> yes

       <jdub> we're thinking about local delivery of laptops and
       services

       <jdub> you're thinking about granting responsibility to local
       orgs

       <jdub> so we're talking about different angles on the same thing

[19:05]<jdub> (seems to me)

       <_sj_> let's take a project as an example.

       <_sj_> do you have a local  one handy?

       <jdub> which suggests to me that us going ahead and scratching
       the local itch, while keeping you guys in the loop, is the right
       way to go

       <jdub> ok, greebo's one above...

[19:07]<_sj_> is there a specific SI ;project?

       <greebo> SI?

       <nessy> SI ?

       <_sj_> solomon islands?

       <greebo> oh, yeah

[19:08]<_sj_> otherwise, nessy, you say you are locally interested
       omstly in running a project, with volunters, borrowed laptops,
       bouhgt, &c.

       <greebo> so an example would be Ian Thomson puts in a request for
       a Solomon Island deployment that they are planning for 4
       volunteers and 50 laptops from the gifted laptops that OLPC
       Australia have on hand (from OLPC).

[19:09]<greebo> They would register the project, we could look at it and
       go "hell yeah" and allocate laptops to send, and volunteers to
       coordinate

       <nessy> _sj_: we have different types of projects here

       <greebo> it is a pretty simple use case :)

       <greebo> the _other_ kind of "project" is the "library" idea

[19:10]<_sj_> ok

       <greebo> where someone says I have this great education
       conference coming up and want to show off the XO to 10,000
       teachers, can I borrow 5 XOs?

       <nessy> plus some volunteers to demo them

       <_sj_> take the first case

       <greebo> and we can say "cool" and either allocate from our own
       ppool, or owners of laptops who own XOs can allocate their to the
       project for a limited time

[19:11]<_sj_> is the request from Ian specifically to you?

       <_sj_> or is it more public?

       <_sj_> I see the request as going out to a dispatch

       <_sj_> that notes that ian is close to greebo's library

       <_sj_> it announces the request to the world

[19:12]<jdub> in this case

       <_sj_> open for general review

       <jdub> because we're likely to be substantial enough to do
       warehousing on behalf of HQ

       <_sj_> and includes it in projects available to olpc-oz

       <jdub> it might be pretty closely held (or entirely public),
       depending on the appropriateness

[19:13]<_sj_> if 4 volunteers and someone with 50 XOs elsewhere in the
       world really wants to send them to Ian, in theory they could try
       to get in touch

       <nessy> the way in which we envisage it is: project proposals
       like Ian's would come into our local system and we would decide
       to support it and run it fully local - no need to interact with
       HQ - but we would recommend to publish it to the HQ wiki

       <_sj_> o-oz might claim the request and try to fulfill it

[19:14]<jdub> _sj_: this is very much the first case (deployment) rather
       than the second (library)

       <_sj_> the olpcwiki isn't an HQ wiki

       <jdub> so this is all about aggregation and warehousing in this
       region

[19:15]<jdub> (sorry, until we have a better term for HQ, i'm sure that
       term is goign to stick around)

       <nessy> sorry, what I meant with HQ wiki is the one you described
       above with published projects

       <_sj_> you keep weirding me out with that phrase :)

       <_sj_> hehe

       <_sj_> no prob

       <_sj_> jdub: interesting

[19:16]<_sj_> jdub: how do you see that working?

       <_sj_> jdub: if you're using HQ to mean "global website/database"

       <_sj_> you can't use it to also mean "the OLPC Office"

       <_sj_> as in "on behalf of HQ"

       <_sj_> I don't know what that means for olpc australia

       <_sj_> or what sort of warehousing you have in mind.

[19:17]*** dsaxena (n=dsaxena at 72.11.87.44) has quit: Read error: 110
       (Connection timed out)

       <_sj_> let's talk about library scope projects for now

       <_sj_> to avoid that complication

       <_sj_> nessy: what if you decide /not/ to support it.

       <_sj_> do you let others decide otherwise?

       <jdub> ok, in that case i'll explain the other side of things in
       /msg

[19:18]<nessy> so, what we're trying to do here is to sort out small
       local issues locally without having to bother HQ, but also filter
       out the ones that need HQ input to push up

       <_sj_> jdub: sure!  I just want to figure out what parts of this
       conv. are applicable to all local groups

       <_sj_> and most of them don't have those issues

[19:19]<_sj_> brb: phone

       <_sj_> nessy: again, where by "HQ" you mean "a global community
       site" ?

[19:20]<nessy> I guess we're trying to figure out if somebody else is
       doing a system like the one we'll be setting up to avoid
       duplicating development effort

       <_sj_> the global input that I think is relevant is what the
       community members reviewing submissions to the contributors
       program think should be supported this week

[19:21]<nessy> _sj_: yes, the global community site

       <_sj_> to the extent that you are taking part in that review
       process, the discussion involves that larger group

       <nessy> is it run by HQ?

       <_sj_> it is run by the community.

       <nessy> ok

       <_sj_> to the extent that you have your own local pool, you can
       make internal decisions

       <_sj_> and then you might be making requests /of/ the global pool
       for local olpc-australia projects

[19:22]<_sj_> that would be a long-term request

       <_sj_> if you expect an active flow of short-term local requests

       <nessy> agreed

       <jdub> our local pool is made of laptops owned by the local
       organisation *plus* those on loan from local developers

       <_sj_> jdub: +1

       <jdub> that's our library

       <_sj_> that's definitely the case for other hardware
       libraries/pools

[19:23]<_sj_> sometimes they also have the fruits of a repair center

       <jdub> yeah

       <jdub> we're likely to have that for local deployments soon too

       <jdub> so perhaps "library" vs. "deployment" are good terms for
       these scopes

       <jdub> and for various reasons, we'll be focusing on both

[19:24]<jdub> while most local orgs might only focus on library

[19:25]<nessy> the library stuff should be automated and self-organising
       as much as possible to reduce overhead, while deployments need
       much personal communication

[19:26]<_sj_> nessy: yes

       <_sj_> nessy: have you thought about the tools that deployments
       need?

[19:27]<_sj_> I think that we can share a lot of interfaces and code for
       dealing with libraries

       <_sj_> the needs of a local library are similar to the needs of
       the global group prioritizing requests

       <nessy> we're mostly considering using a forum software to
       support deployments

       <_sj_> In an ideal world, individual requests can always be
       handled by a local group

       <_sj_> or library

       <nessy> yes

[19:28]<jdub> thus we scratch our own itch to make an
       olpc-local-org-in-a-box system

       <_sj_> right now we're doing it with the projectdb because that's
       only true in small numbers in  austria, germany, france, boston.

[19:29]<nessy> when you say "it", what do you mean?

[19:30]<johnf> I think the core difference is in that in projectdb,
       there is one pool of laptops in the library. Where as we are
       talking about volunteers being able to add their XO that is
       gathering dust on their shelf into the library

[19:31]<_sj_> johnf: there is one pool in which library?

[19:32]<_sj_> projectdb is intended to be a project matching service.

       <_sj_> someone with a project proposal enters it

       <_sj_> it gets revised over time, matched with a group or person
       that has available hardware

       <johnf> so are their two groups of people in olpc. People with
       projects and people with XOs and they can get married up?

       <_sj_> (including the regular visitor 'Brighstar' who seems to
       have a couple hundred machines available every month)

[19:33]<_sj_> yes, and people get married up.

       <_sj_> with a good system, matching could be p2p

       <_sj_> right now people say "I have a laptop" and send it to a
       library; and the people organizng the library review public
       proposals every so often and pick recipients

[19:34]<_sj_> which is slower

       <greebo> _sj_, so in a way projectdb's goal is our goal, so the
       real next step is to a) get a copy of the source code so we can
       start to look, and b) get it more openly developed. Any ideas on
       b? :)

       <_sj_> greebo, let's just start a repo :)

[19:35]<nessy> yes, make it public

       <_sj_> sourceforge page?  or do you have another preference?

       <nessy> is a good way to build a devloper community around it,
       too

       <_sj_> I'm pretty nonplussed that the current code remains in a
       closed repo

       <_sj_> it's not a lot of code, either.

       <jdub> for a start, we just want to check it out

       <jdub> then we'll figure out the repo foo later

[19:36]<nessy> make it public somewhere - soureforge is fine

       <greebo> let's decide on where to put it after we check it out
       _sj_

[19:37]* _sj_ is inclined to rewrite it...

[19:38]<greebo> _sj_, heh, can you send us a tar.gz?

[19:41]<jdub> so i reckon we're on the same page now

[19:42]<_sj_> uploading some images...

       <_sj_> greebo: yes, I'll email you with details

       <_sj_> if I can't upload the file

[19:43]<_sj_> are you pia@ and jeff@ laptop?

       <_sj_> hmm, not jeff@ surely :)

       <jdub> naw

       <greebo> _sj_, no, pia.waugh at olpc.org.au is fine

       <greebo> thanks!

       <jdub> jdub at bethesignal.org

       <_sj_> better

       <greebo> then we can come back and chat online in a few days or
       so :)

[19:44]<_sj_> johnf, where are you?

[19:45]*** jg (n=jg at c-76-19-89-183.hsd1.ma.comcast.net) has joined
       channel #olpc

       <johnf> _sj_: 4 feet from nessy, greebo and jdub

       <johnf> :)

[19:47]<_sj_> :-)

[19:49]<greebo> ok, cool, thank you heaps _sj_ and isforinsects. I think
       we all better understand what we are all trying to achieve, which
       hopefully means we'll be able to better integrate, avoid
       duplication of effort, and collaborate where we can. I think we
       can create something generically useful for regional/local groups
       to manage their needs, and to fit with what OLPC themselves are
       doing.

[19:50]<nessy> yes, it seems we have overlapping needs and we should see
       that we can work on the same codebase, whichever that will be :)

[19:51]<greebo> yeah, that too :)


More information about the Grassroots mailing list