Re: Debconf 6

[Thomas Walter <t.walter@nefkom.net>]

Hello,

On Sun, 2005-12-04 at 19:03, Frederic Lehobey wrote:
> Hi,
> 
> Helen's talk at debconf 5 led to the creation of this mailing list and
> to this wiki: http://wiki.debian.org/DebianScience
> 

I have read through some topics, especially ocatave and related stuff.
For plotting I think 'plplot' is the best as it supports 2D and 3D,
maybe the latter needs some more "high level" user functions for simple
use.  Octave itself is/was a very good program for me when I did some
research in physics.  But one has to use always the current version.
It is easy to write octave functions for the octave interpreter and also
to write C++ functions to new functions or wrapper to call external
libraries (I did that for fftw and gsl some years ago).
Another candidate to do similar things may be 'pdl' (perl data
language). But I have not used that much.
All in all, what I have learned and still learn by using complete
releases from debian:
It is good to have a stable release over a longer period with respect to
basics, but not for "end applications".
I think "End applications" need a much shorter release cycle to get new
features to the end user.  Or it happens what was mentioned about
'octave' -- it's far behind for example 'matlab'.  That's only because
the 'octave version' in 'woody' was stone old.  Octave has very good and
supportive mailing- and developing lists with a short but careful code
change loop.  Thus using current release is not that risky but very
helpful.
>From my opinion, similar applies to all similar applications.
Thus either install from recent source or think in debian about making
"end applications" independent from the basic system.


> Are there any plans to have some followup at debconf 6 with respect to
> the scientific use of Debian?
> 
> Are people on this list interested in settling some gathering there?
> (Who already knows s/he will go?)
> 
> Best regards,
> Frederic
> 
> https://www.debconf.org/comas/general/proposals

Kind Regards
Thomas