Re: Bug#514903: ITP: pymca -- Python toolkit and application for X-ray fluorescence analysis

View: New views
2 Messages — Rating Filter:   Alert me  

Parent Message unknown Re: Bug#514903: ITP: pymca -- Python toolkit and application for X-ray fluorescence analysis

by Andreas Tille :: Rate this Message:

Reply to Author | View Threaded | Show Only this Message

On Thu, 12 Feb 2009, Chris Walker wrote:

>> * Package name    : pymca
>
> [snip]
>
> I've added it to the science-physics task.

Great.

>> Prelimary packages will be soon at the Debian Science repository
>> at git.debian.org.
>>
>
> When it is, if you let me know I'll add that information to the task too.

This would be specifically interesting if you set the Vcs-Git field
in the tasks file which is documented here:

    http://blends.alioth.debian.org/blends/ch-sentinel.en.html#s-packageslist

Currently the effect of using a Vcs-Git or Vcs-Svn field is that the
package is ranked in a higher section of the "yellow" packages called

    Debian packaging stuff in Vcs

You can see the effect of several Vcs-Svn fields inside the med/bio
task at

    http://debian-med.alioth.debian.org/tasks/bio.html.en#pkgsvn-debs

I admit that Vcs-Git is not yet tested at all because I had no example
(but will have soon once I commited cteddy to pkg-games for the Debian
Junior project).

The rationale behind this feature was to avoid confusion about the
different status of projects in the "yellow" section.  The differentiation
of packages in experimental has to be done.  (At this place I would
like to thank SkoleLinux people for enabling me to do this work at the
Debian Edu developer gathering!)

The plan for the future is to be able to add *only* a Vcs-* field and
just obtain the other information about prospective packages out of
debian/{control,changelog,copyright} inside the Vcs.  Isn't it fun to
be able to keep things simple and provide the same information only
one? ;-)
But I try to coordinate this with UDD people - perhaps we also write
the code needed for this only once.

I hope this makes the real power of tasks files even more clear for
those people who have their doubts.

Kind regards

          Andreas.

--
http://fam-tille.de


--
To UNSUBSCRIBE, email to debian-custom-REQUEST@...
with a subject of "unsubscribe". Trouble? Contact listmaster@...


Parent Message unknown Using UDD for Blends and Vcs (Was: Bug#514903: ITP: pymca ...)

by Andreas Tille :: Rate this Message:

Reply to Author | View Threaded | Show Only this Message

On Fri, 13 Feb 2009, Manuel Prinz wrote:

> This would be really nice! As for Git, there are two things to concider:
> First, we probably need to agree to the name of the branch in which the
> debian/ files exist: some use "master", some "debian", or whatever seems
> to fit. Second, one has to consider that Git repos may be bare, so you
> can not access the files directly. This is not really a problem, one can
> access them without the need to clone the repo. (I have a shell script
> to do so for use with mr.) I guess this can be done in Python as well.

Sounds good.

> Both of them are no show stopers, I just wanted to write it down so it's
> not forgotten. Hopefully I'll find the time to help implementing that
> soon.

I think we have some time.  I just became excited about UDD and I will
probably work on rewriting the methods I get package info to use UDD
instead of Packages.gz first which is much more clever and more reliable.
While thinking about UDD it might be reasonable to find a way to inject
the info we need for prospective packages to UDD or DDE.  This might
be interesting also for non-Blends issues and finding a general solution
seems to be more clever.  Later we might be easily able to select the
data we need for a Blend in question.

Kind regards

          Andreas.

--
http://fam-tille.de


--
To UNSUBSCRIBE, email to debian-custom-REQUEST@...
with a subject of "unsubscribe". Trouble? Contact listmaster@...