Re: Python code to retrieve all open wnpp bugs
On 22/03/09 at 19:52 +0100, Sandro Tosi wrote:
> I would like fresh data, so I take
> http://qa.debian.org/data/orphaned.txt and packages maintained by "QA
> Group" on the local Packages file (updated frequently), so here are
> some questions:
For reference, most of the info about UDD is available on
http://wiki.debian.org/UDD
or http://udd.debian.org/
> - is UDD querable by any host (my pc for example) or only by *.debian.org ?
no, only from merkel and alioth, but more (*.debian.org) could be added
if really needed.
> - how fresh is data on it? hours, days, weeks?
depends on the data source. Updated:
- several times a day:
sources/packages data for testing/unstable/experimental
bugs
testing migrations
- once a day:
sources/packages data for other releases and Ubuntu
popcon, upload-history, list of orphaned packages, lintian, debtags,
carnivore
ubuntu bugs (broken currently, should be fixed soon)
- every two days:
archived bugs
full DB dump on http://udd.d.o
> - what's the policy for reports from UDD? If i'll follow udd-way, I
> consider moving all the logic of this report on it, so:
> - are views (materialized view also) preferred to live query (from
> web page, f.e.)?
> - how do you store logic code? prefer procedure/functions on the db,
> or extracted to a "processing" layer, be it directly web pages or a 2
> layer (programm connects to udd, elaborate, and "outputs" a web page)?
It depends. There are no strict rules, just do something sane, and don't
D.O.S UDD.
> - is it a transactional db or more a datawarehouse (with staging
> area and consolidated data generated from it)?
it's postgresql. Updates happen in transactions, so even during updates,
you can still get the full data (except it's the old set, not the new
set)
--
| Lucas Nussbaum
| lucas@lucas-nussbaum.net http://www.lucas-nussbaum.net/ |
| jabber: lucas@nussbaum.fr GPG: 1024D/023B3F4F |
Reply to: