GT.M for Debian (Was: [gnumed-devel] Debian Packaging Drive)

[Andreas Tille]

On Sun, 3 May 2009, Allan MacKinnon wrote:

> ---------- Forwarded message ----------
> From: K.S. Bhaskar <address@hidden>
> Date: Sun, May 3, 2009 at 8:43 PM
> Subject: Re: Debian Packaging Drive
> To: Allan MacKinnon <address@hidden>
>
>
> Thanks for writing, Allan and thanks for your interest.
>
> I did have a discussion with some Debian folks a couple of years ago.

This was probably me. ;-)

>  It takes an existing GT.M
> binary to build GT.M - i.e., it is bootstrapped (not unlike gcc- you have to 
> have a gcc in order to
> compile a gcc).  They understood the issue technically, but didn't seem 
> comfortable with the idea.

I would like to clarify this: It is not that I'm not comfortable with this idea 
- it
is just the only chance to do so and it just needs some energy (=time) to push 
this
trough.  So *I* feel not comfortable to grab another time consuming tasks on my 
own
shoulders when not having dealt with a few others finally and it makes no sense 
for
me to let rott several tasks in an undefined state.  As you can see at

   http://debian-med.alioth.debian.org/tasks/his#gt-m

GT.M is on our todo list and I also have given a link to the debian-devel 
mailing list
where people discussing similarities to gcc and mono.  So the issue can be 
technically
solved under two circumstances:

  1. There is somebody doing the work (as always).
  2. There is enough support of users to convince ftpmaster to accept a
     first binary upload because it is important for our users

>  So, GT.M probably can't be pushed up to the Debian repositories.

If 1. and 2. are solved chances are not as bad.

> In any case, I do plan to package GT.M as a .deb file myself, but I'm probably 
> about a month away from
> having the bandwidth to do so.

I'd consider it a very good idea if you would coordinate on the Debian Med 
mailing
list and perhaps even push your work to the Debian Med SVN repository.  This 
might
gain some more interest and despite the small user base of people in health care
(compared to the large amount of "general" users) Debian Med has gained the 
status
of an "interesting" project - so we could push item 2. of above if we can argue
like this: GT.M was developed and tested as an inofficial package in the Debian 
Med
scope and has x active users.  We would like to start a migration path via 
Debian
experimental - or something like this.

I would really like to support GT.M in Debian but we are just lacking manpower
(and knowledge!) to do so.  If we would join forces GT.M knowledge + Debian
packaging experience we might deal with the issue.  Just keep me informed if you
are doing anything Debian related - just let us (address@hidden - you
might like to subscribe at http://lists.debian.org/debian-med/) know.  We are
really interested.

Kind regards

       Andreas.

--
http://fam-tille.de