Re: Debian for Systems Biology
On Mon, 30 Jan 2006, Charles Plessy wrote:
only critisism I have on it is its name: I guess that it contains
molecular biology software because it started with gene discovery on
microbial genomes, ans I think that when seeing the URL, a lot of
persons will think that it is related to medical bacteriology, and wil
overlook it. On the other hand, on the main page, the link name is
molecular biology, which makes more sense. So if you modify the
debian-med structure one day, I think it would be worthwile try to solve
this, but I suppose that changing the name now would break links and
make this page lose readers instead of winning some.
Well, several months ago I suggested that we should start building such
web pages automatically and started some very basic code (which is far from
beeing usable and no chance of continuing this for some further months).
The idea is as follows:
1) Obtain the contents of the first part of the web page by parsing
Packages.gz file for a list of packages that should be listed on
the page.
2) Use ddtp translations for translated pages. (Well, unfortunately
the ddtp project is currently in a quite bad state. It was much
more usable before the break into the Debian servers in the end
of 2003. :-()
3) Build a "fake Package.gz" file containing the projects that should
be included into Debian and into the CDD in question to build the
list of projects with inofficial packages or are not yet packaged
and use the algorithm from 1) to create the other part of the web
page.
If we would have such a system every project could easily build the
web needed web pages by just compiling a list of projects that should
be included. So if the only thing you do not like is the name of the
page you could compile your own one by just renaming the list name
(that might lead to a page name) and perhaps add or remove some projects.
That way we will not double any work but gain maximum flexibility.
The only problem is that someone has to write the code described
roghly above ...
Thanks for the link (http://people.debian.org/~tille/cdd/). I guess I
have to read it before answering. Anyway, I think that most of the work
is not the meta-packaging, but to gather the software into debian.
Sure. Getting the software packaged is the precondition and this is
the main goal. The meta-packages are just a nice tool to let users
install what we gathered.
Also, I wonder how some existing packages could be tweaked. For instance, the
search box of firefox can be used to query very relevant databases,
such as pubmed.gov or genbank (http://www.ncbi.nlm.nih.gov/Genbank/),
You mean mozilla-biofox? I have to include this into the Debian-Med
packages but this is more or less the only current pending change and
seems to small for inclreasing the packages version for the moment.
and shipping this configuration by default would make senes on a CDD.
Figuring out how to do this is definitely one of my goals.
If I did not misunderstood you installing mozilla-biofox package
would be enough.
So I will subscribe to the debian-med list to avoid duplication of the
work.
This might be a good idea.
Kind regards and thanks for your input
Andreas.
--
http://fam-tille.de
Reply to: