Small reflexions on refactoring QA
These are few reflexions and ideas about what we can refactor in QA.
(Please add comments :-))
On one side we have a plenty of tasks to do to improve the
quality of our distribution.
On the other side, we have a large number of people who are
willing to help and don't know how to start and what to
start with.
What do we have now?
--------------------
- a mailbot to register tasks.
- a wnpp entry in the BTS to manage prospective packages
- a static web page which is not used at all
What can we do?
---------------
The task orientation of the QA seems a good approach to me.
I can see two kinds of tasks:
. QA tasks that are only done by QA people (i.e. MIA tracking, etc)
Work has already begun, some scripts for tracking MIA were made by Martin.
. QA tasks that can be bone by anyone.
All these kinds of tasks are already presented there :
http://qa.debian.org/howto.html
I have in mind, among others :
- per package page referencing all undocumented executables from
lintian reports
How?
---
. Mailbot
I personnaly love the idea of the mailbot : this mechanism works pretty
well (the best example is probably the BTS).
I don't know whether it meets our needs.
. Web page
What we have to do it to present tasks such a way that people can better
have a look to what is to be done. All BTS related tasks won't appear
on the web page (better presented?).
Everything has to be as much automated as possible.
This is not enough but a good start imho.
Cheers,
--------------
Jérôme Marant <jerome.marant@free.fr>
Reply to: