Projectdb
The projectdb (alpha name) is a database designed to capture and track requests for XOs.
It is currently being used as our live contributors program/developers-program interface. Over the past week it has processed 100 applications from LinuxTag.
Current feature requests are listed below; AaronKaplan and Chris Hager are the two current project maintainers. The codebase is currently partly open; as git access is not generally available. The code is being cleaned up this week to accomodate important feature requests.
- SJ *believes* the code is only partly open . In fact it is. But I just have a roadmap for it for initial release. And that roadmap will be followed in order to assure some quality. 213.47.53.193 15:39, 7 June 2008 (EDT)
The first weekend in June, --Sj talk, femslade and erikgarrison are working on extending the curent data model to cover further aspects of projects (including the stage of completion, idea through implementation, initiatives and trials, as well as projects that require new hardware to get underway).
Contents
Roadmap & Feature Requests
Bugfixes
- Make all entries editable by admins
- Fix quote/slash escaping
- Show name of requestor on the requestor-view page
- Show requestor & project info side by side on same page
- Add status options : "submitted", "shipped", "received"
- Support post-to-wiki on submission
- Support creation of new accounts / granting admin status to accounts by admins
- Basic duplicate-catching for project submissions
Features
- Make tables sortable by column
- Support compressed view of main page, one line per entry
- Add export to csv (and xls?) of main tables/views
- Add filtering/viewing by country (for export by country)
- Add filtering/viewing by submitter / by duplicate submissions
- Separate list of all account-holders from list of project submitters
- Editing : add a "last edited/updated" field and track which admin last updated an entry.
- Backups : make a nightly backup; confirm backups b/t olpc.at and d.l.o.
roadmap
.
general db
A further extension of a shared database should cover information about all the data that is core to OLPC:
- Projects - ideas, development, creation; initiatives, trials, research
- Parts - XOs, spares, peripherals, power
- Places - countries, cities, schools
- People - individual teacher, developers; organizations, partners; chapters, interest groups
current Use
A half-dozen people are reading data from the current temporary interface and hand-parsing it into xls formats to satisfy a changing process; this will later be done via xml (with xml export to B* and xml import from them for confirmation of receipt and shipping). We are working on better export formats to support the interim process.
Data specification for processing projectdb requests
The process of requesting XOs currently is passed on to B* via spreadsheet. They are working on an xml solution that will accept and generate XML lists to automate the process.
- NOTE by Aaron: we have been pushing SJ for literally *months* to get a proper specification from B*. This sucks! Suddenly everything has to be done very very fast and the two main developers are not given any time to prepare a proper release. (unsigned by aaronKaplan)
- Yes, this is a problem. I still don't have those specifications; so we have to make do with what is currently required (csv --> xls, split up by country). Sorry for any perception of a rush; core feature requests (unrelated to xml formats) are roughly those from Feb, and the rest are not so important atm. We are hoping to use the db to process a # of requests similar to those @ linuxtag, and I'd like to avoid a repeat of the usability issues found there : particularly the dup submissions, lack of wikipresence, and lack of clear timestamping or admin-updating of reqs, which makes it hard to mark all LT08 reqs as coming from that group. --Sj talk 16:26, 7 June 2008 (EDT)
