Content meetings/Summer 2007
19 Aug 2007
- next: Monday 1500 EST
- SoCon -- need updates and help needed
- Bundles -- targets, wikibook creation, atlas, bundles
- c&c and lists -- update [dates & lists], teamwiki updates
- repository services -- web interface, rh/other discussion
- school server requirements
- other school server requirements... registration, uploading, aggregation?
14 Aug 2007
- next meets: Mondays
- socon - updates and help needed
- t3 - targets, wikibook creation, atlas, bundles
- coordinators - updates
- repository services
- school server requirements
- misc: offline wp, &c.
7 Aug 2007
Tues Aug 7, 2007, 2pm EST, on #olpc-content. add to agenda. to be present: paur, lauren, sj, mel, (kim?)
- next meeting :
- report from Wikimania - wiki, community, loc updates
- check-in on Activity & Library grid, XO and server collections
- reconnect re repository/server dev, PM, wiki farming
- Renaud Gaudin, Mali (offline editing/synch internship)
- collection sizes & C drop
- l10n priorities
1. curation structure
- pe,pn,ur,eth ---> spanish-language curation, model a spreadsheet for them
- compare br v ng
- scripted curation from repositories, suggest priorities
2. server structure ---> specs & processes.
- repository definition & use case ---> how to add to it? how to partition?
- download and distribute locally.
- making something available to a full class
- how to queue requests?
3. upload/download library content
- web form for creating cats
- web intern / UROP / NE
link/feature freeze M ===> code freeze 8/27 ==> 9/xx? country dep A : no mats provided; multiling bundle. country dep B : 9/30 extended bundle activity; update via net; v-chat:? mesh UI? assume we're not getting a server build to burn - how much should we do to extract some of it? calendars & schedules devs: detail testing. others: do large testing. general test plans, cf. test group release notes on [[test issues]] smoke testing -- browser tests (video, audio, library)
30 July 2007
Monday July 31, 2007, 2pm EST, 2AM Taipei, on #olpc-content. add to agenda. to be present: paur, lauren, sj, mel, (kim?)
- setting a time for regular meetings
- quick report on Wikimania; any requests from SJ and Mel (who will not be sleeping, apparently)
- check-in to make sure we have an easy place for Taipei Jam contributors (& others) to upload content - Library Grid not the most friendly thing in the world at present
- touch base re hiring (specifically re PM and server developer)
have a lot of disk space; useful as supernodes; shouldn't be required in general for services to work
- weekly discussions of implementation
- currently: no readily accessibly crypto signing
- libraries: preshipped, modular components added by hand
working on this: Adir Abraham
- use cases: world at large
5 July 2007
Who gets to write to the ]]data store]]?
Some of this is still up in the air. Interesting use cases:
- an activity wants to store something to the data store that is public and sharable. How can the store something in the data store and tag it as sharable?
- There's currently a "share activity" / "collaborate" button that shares a current document. Implications need to be worked out.
- We need an indicator of 'state' -- whether one is working in a shared or private state at a given moment.
- mesh board (distributed craigslist) -- as soon as you 'share' or 'publish' a new announcment, it is stored as something shared.
- Can something on the level of the presence service offer distributed content or other pervasive services that store data using the tag and permission features of the data store? should there be something similar but separat with different security implications for scratch-type store and forwarding? (see below)
Store and forward : Whether or not you use store and forward, there are times when you want to communicate with someone who isn't online.
- Imagine sharing distributed content across a loosely connected network.
- Can this use scratch storage on the laptops? Or another 'directory' in the data store?
Asynchronous notification: We need a style guideline for notifying someone that things are happening related to their interests outside the scope of the current full-frame activity
- notify people when they get responses to their efforts in a particular arena (say, board postings, or shared items further edited by collaborators)
- this can also be useful for doing semi-persistent sharing of files and other assets across the mesh.
Activity-level daemons : There's no spec yet for defining user-level activity daemons, for something persistent.
- Does a data sharing service need to be at the presence level? If so, we may need to provide a UI for turning this on and off.
- Do we need sth like a python libevent dbus loop to allow cheap addition of persistent service daemons?
12 June 2007
- Present: Ryan, [[User:Lauren|Lauren], Kathy, Mel, SJ
- Lightweight index library metadata spec called Ad Libris - think a little virtual "bookplate" that can be attached to (or pointing to) different files and content; these "bookplates" are the items that can be tagged/extended/indexed/have reviews attached, etc.
- Most of the discussion is detailed in the Ad Libris Spec (under construction).
- Note the immediate need for instructions on how to make a bundle (in the generic sense w/ content, not just software activities).
Monday 9 April, 2007
1. making a library activity that's readily customizable and skinnable
(with sugar and html views),
+ working with country partners to develop a local image in four
2. coordination of look and search with the World Digital Library and
+ html and other; design and usability guidelines 3. default views: by topic, by media, by language, by geography + existing projects and interfaces with view selections. 4. content guidelines: writing for children, for translation, other.
Saturday 14 April, 2007
Proposed schema for a [world open library exchange]: This needs defined interchange formats and collection-level style and feed guidelines. Most collections have limited collection or transparency of the fact that new additions or changes have been made (metametadata?).
Classification into groups and collections, needing standard "breadcrumb" metadata format for all content providers to reference object locations in the [OLE]. This draws upon the ideas in Content_Stamping and the structure of the current library as well as desires to promote multilingual cooperation.
Country materials and existing initiatives
Working through a few countries; looking for university and creative nexus collections. Feel free to add yours here: Library and learning initiatives (notes on Argentina, Alegre, Colombia, Nigeria, Pakistan, Paraguay, Uruguay, Nepal, Panama, Ethiopia, Mexico moved).
format and display issues
- screen resolution questions
- efforts to fix open font solutions
- What fonts are needed for interested countries, by priority?
Multilingual chat opportunities
- How does this impact development?
Models of content/levels/scale
(small to big)
- Leaves (whatever)
- Branches (.xols)
- Trees (web/wiki pages)
- Forests (indexes)
- Sky (views, overlays, lenses)
- Individual pieces of content: photos, stories, articles, etc.
- Grouped collections: a newspaper issue, a photo gallery
- Sources: wikipedia, this atlas site, this artist's homepage, that comment (large-scale visualizations of a set of materials)
- Reference, community, classes, search engines
Things to focus on:
- featured images
- wp snapshots
- rss feeds
- tpical categories
- audio, video, books
- Donald in Mathmagic Land
- "Surreal Numbers: How Two Ex-Students Turned on to Pure Mathematics and Found Total Happiness"
For bundling and metadata
- books, games, reference, free software for them
- childrens' books 
- math texts and resources (NLVM, teacher)
- not really: hdn, zeen
A catalog should be searchable, dynamic, with caching of recent downloads and the ability to check links and update, aware of restrictions (size, format, licensing)
We can minimize the need for each of these in a first pass; by setting the bar for restrictions low, making updates sporadically and in bulk, and using the web and simple fts engines for search.
The web app version: let anyone register to submit collections; these have metadata, index pages, and links to originals for verification. let others come review, verify, and update collection data. any compilation via scripts or otherwise should be shared for recompilation after small changes are made. APIs to this will include a way to add many groups or reviews at once; a way to synch distributed offline submissions and comments; and a way for bots to efficiently add or change metadata.
The content browser: at first, the browser itself. Eventually, like the activty browser: a way to find, remove, and archive things in packages. APIs to this will include ways to upload content all at once; for instance a default collection or a default view into a school library.
- calendar sanity check
- recruiting for positions & internships & volunteer needs
- heads-up that Mel will be sending out requests for reviews, later tonight, of:
- jam cookbook (finally)
- mildly more formal jam program
- university program
- summer of content stuff
- request (by Mel) to move 3pm Monday meetings to #olpc-content-mtg in the future, so we can enforce staying in-agenda & on-topic better without being distracted by other convos & passersby
- music - new contribs.
Visibility and stamping. 19:30
- Specific topics needed : bootstrapping materials, images for using each application
- Visual and playful introductions
- Module/topic priorities by country
- Curators and topical/language coordination
- Online v. school servers v. laptops -- finding & using material
- Exposing created works : documents, photos, videos, projects
- Finding project collections online
- Trial library testing
- laptop test materials (weekend)
- server materials, School Server Specification, et al.
- Content review and stamping
- microformats, Ratepoint, Connexions' del.icio.us model
- Game development, weekly discussions
- Content review and stamping