Ultimate Debian Database Discussions
This page is meant as a log of the discussions that go on via mail/IRC.
There are two seperate categories of discussions: Those in which a consent was reached and those in which no consent was reached yet.
consented discussions
Meetings
No regular meetings will be held. Instead, I will log discussions here, and work done elsewhere.
open discussions
Programming language
Currently we are not sure which programming language will be used to write the scripts It seems that all involved people agree that only one language should be used as for retrieving the data from the sources. Furthermore we agree that all paricipating people should know this language well. The most likely choice at the moment seems to be python, mainly because of python-debian
final name
Nobody came up with a better idea, so it probably stays with "Ultimate Debian Database"
Database update
How should the database be updated? Possibilities are diffing the sources against the database and a complete rebuild.
Tests
How can we ensure database integrity?
Database
Which database should be used? Postgres seems most probable at the moment.
debbugs
While looking at bugs.debian.org, several questions arised concerning the harvesting of this information:
- There is a very convenient perl library to work with the data coming from this source. Should we use this library? This doesn't match that well with the most probable language. On the other hand, it would be much more convenient to use perl instead of reimplementing debbugs in python.
- Which information should be gathered? There are several things to consider:
- How valuable is this information?
- How expensive is it to gather this information hardware-wise?
Also see the matching article in data sources
Carnivore
Integrate carnivore?
License
Lucas suggested GPLv3, and I don't really care.
Key-type
Do we want multi-column keys or serials?