Mini-conference on Apr 3/4!

Edward Cherlin echerlin at gmail.com
Tue Apr 1 14:07:31 EDT 2008


On Fri, Mar 28, 2008 at 1:58 PM, C. Scott Ananian <cscott at laptop.org> wrote:
> We've gotten 9 proposals for our mini-conference, and I know of two
>  and a half more which haven't yet hit devel@ (Walter/Kim, Michael, and
>  Martin, on "build process", Bitfrost, and school servers).

If I could get to the mini-conference, I would offer a talk on the
wretched state of planning for language support. I know of several
dozen languages in target ("green") countries for which there is no
Pootle project. I just sent out a list of the languages needed for the
Philippines in response to an announcement of projects there. Since
nobody else stepped up, I have become the administrator for Haitian
Kreyól and Cambodian Khmer, languages that I do not know (although I
can sort of read Kreyól due to its French roots).

Bottom line: OLPC needs somebody to do this full time, paid, in Cambridge.

If you like, I can send you a slide show (OpenOffice Impress) and give
a talk over the phone, or we might be able to set up a video
conference. There are plenty of facilities in Silicon Valley.

I could talk about some other development management failings at OLPC,
and suggest some measures that should help, but one is enough per
e-mail. There are other management failings outside the area of
development which I will not bring up on this list, except to give you
two words--premature optimization.

>  Since most of the participants are local, we'll be having our
>  conference at the OLPC offices at 1 Cambridge Center next Thursday and
>  Friday (April 3 and 4).  Tentative plan is for talks from 12-6pm, then
>  dinner, and then informal discussion in the evenings. The community is
>  welcome!
>
>  One of my goals with this mini-conference is to encourage
>  cross-pollination between the teams working on various technologies,
>  and between developers and the community.  To that end, we'll schedule
>  each proposal in a ~1 hour window: a 30 minute talk, followed by no
>  more than 30 minutes of discussion, then on to the next topic.  The
>  goal isn't to nail down every open question during the hour, but
>  instead to start discussions which will continue in the evening and
>  online.  People who aren't terribly interested in a topic should be
>  able to enjoy and learn from the structured talk without having to
>  suffer through too much "vigorous debate".

That's a shame. I love vigorous debate.

>  I'd like to have webcasts of the talks for posting later, and/or for
>  live streaming, if possible.  Can anyone volunteer to help out with
>  this?

I'll see whether my friends in the business can help out. They set up
live conferencing for a UNESCO/Club of Rome event in a number of
countries over the World Bank network.

>  For proposers: I'd appreciate if you could organize your talks roughly
>  as follows:
>   * background material ("what is a mesh network")
>   * history & current state of affairs ("what currently works? what doesn't?")
>   * open tasks, both "simple matters of programming" and "hard problems
>  no one has a clue about"
>   * concrete proposal for forward progress.

OLPC's localization and translation planning needs

Background
* Localization: strings in software
* Translation: Wiki, textbooks, content
* Multilingual text to speech for literacy and accessibility

History and state of affairs
* 60+ languages so far on Pootle
* Strings per module: Etoys, XO-Core, etc.
* Wiki translations
* Library
* No library repository
* No mechanism for tracking permissions and licenses
* No visible planning
* No outreach, no recruitment
* No public information on language communities to be served

For example, in Afghanistan we know we need Pashto and Dari, and both
are in Pootle. But what about Hazaragi and Aimaq? Where are the G1G1
XOs going in Afghanistan? Are we just going to leave these
historically persecuted minorities out?

Open tasks
* Dozens of languages needed but not begun
* Lists of countries
* Lists of languages
* Resources: NGOs, Universities, LUGs, children...

Proposal
* Hire me, or somebody

>  The first two bullet points encourage the dissemmination of the bits
>  of specialized knowledge that live in a few developers' heads, and
>  insisting on a concrete proposal at the end will hopefully encourage
>  productive discussion.
>
>  More details as they're planned!  Hope to see you all there!
>   --scott
>
>  --
>   ( http://cscott.net/ )
>  _______________________________________________
>  Devel mailing list
>  Devel at lists.laptop.org
>  http://lists.laptop.org/listinfo/devel

-- 
Edward Cherlin
End Poverty at a Profit by teaching children business
http://www.EarthTreasury.org/
"The best way to predict the future is to invent it."--Alan Kay


More information about the Devel mailing list