Re: Request for Ideas - Astronomy working group
Hi Ole,
at first thanks for all your great work on astronomy packages.
On Thu, Jan 09, 2014 at 10:41:42PM +0100, Ole Streicher wrote:
> Dear all (astrophysicists, amateurs, scientists, other),
>
> there are now quite some people around that are interested in using
> Debian for astronomy related tasks. It may be a good idea if we could
> somehow bring us together and see where we can coordinate our efforts.
+1
> Some days ago, I was contacted by the creator of the "DistroAstro"
> distribution <http://www.distroastro.org/>, which is dedicated to
> amateur and professional astronomers, asking for help with his
> distribution. Distroastro is a derivative of Ubuntu. I am now trying to
> get him interested in creating his packages for Debian instead of just
> his own distribution; but this example also shows that we are not
> visible enough to others when speaking about astronomy with Debian.
The situation you are facing in the field of astronomy is definitely not
new. If you would like to browse the mails on the OSGeoLive list (an
Ubuntu derivative dedicated at Graphical Information Systems) you see
the very same situation: Somehow people working on the same topic with
basically the same technique (Debian based) consider working together
more closely. I would recommend to check out
http://lists.osgeo.org/pipermail/ubuntu/2014-January/thread.html
and the mails I wrote there might apply 1:1 if you would replace Debian
GIS by (to be created) Debian Astronomy (or whatever cool name you might
like to choose) and OSGeoLive by DistroAstro. It is the situation I'm
propagating since the creation of this mailing list: Lets join forces
in a certain scientific field and Debian Science should be the umbrella
for these sciences as long there is not enough manpower to run an own
project == an on Blend. Considering the manpower we have inside Debian
(visibly like you and behind the scenes people who worked silently and
wrote private mails) and the additional fact that there is some
additional external manpower I'm pretty sure that this is a very good
point in time to start an own Blend right now.
> However, I don't have a good idea on how we could organize ourself. The
> debian-science mailing list is still probably the best place, and an
> additional mailing list would not make sense yet.
If you want to attract people who are not that interested in Debian
Science in general it might be that the signal noise ratio regarding
astronomy questions is not really comfortable for them. So if I were
you I would rather register an own alioth project with mailing list and
packaging repository. It is also a psychological effect to outsiders
newcomers: It makes a difference if they see: "Ohhh, there are people
discussing astronomy which is exactly the topic I'm interested in."
compared to "Well, there are some random scientist with a few
astronomers amongst them issuing some random mail about once a month."
IMHO this is the reason why some people stick to private mails rather
than simply writing to Debian Science list. A separate mailing list
would drastically lower the barrier for newcomers.
So I'd recommend having a look at the great work of Sebastiaan
Couwenberg who has just adapted and enhanced the Debian Med policy
to create policy which fits Debian GIS perfectly:
http://linuxminded.nl/tmp/pkg-grass-website/policy.html
IMHO, you would be really well served by doing s/GIS/Astronomy/ and
create the things which are needed technically on Alioth. I'd volunteer
to help you with all the steps if something remains unclear. If you
doubt the effect of this approach I have a good example that I was able
to convince an OSGeoLive activist (Ivan Mincik) to become a Debian GIS
member and the first package of him which I sponsored hit the archive
yesterday.
If you are short in sponsoring manpower I remind you about my Sponsoring
of Blends[1] effort. I'd also volunteer to lurk on the (to be created)
list to give input about project management as I'm doing also on Debian
GIS.
> There is a (really
> outdated) wiki page available at
> <https://wiki.debian.org/DebianScience/Astronomy> which could be used as
> a start.
As I frequently said the single fact that a page is a wiki is not
sufficient that it is really maintained. That's the reason why I keep
on advertising Blends tasks pages which are more probable to stay
up to date regarding the packages in the focus of the Blend. Well,
you also need to mention the packages and after checking the Wiki
page I did a
diff --git a/tasks/astronomy-dev b/tasks/astronomy-dev
index c5eb5f5..eb95f63 100644
--- a/tasks/astronomy-dev
+++ b/tasks/astronomy-dev
@@ -25,4 +25,6 @@ Depends: liberfa-dev
Depends: libfits-java
+Depends: libivoafits-java
+
Depends: tcl-fitstcl
which simply adds the prospective package to
http://blends.debian.org/science/tasks/astronomy-dev#libivoafits-java
I keep on hoping that it is obvious that with the same effort as editing
an entry to the Wiki page the effect on the tasks page is way more
powerfull since without any further action the tasks page will be
updated if the package description might change in Vcs and it will be
shown if the package was uploaded to new or arived inside main Debian
(with all the metainformation about available versions, popcon,
screenshot, translations etc.)
So please: If you want to present up to date information to your users
about the packages you are maintaining: Just stop wasting your time in
the Wiki but rather edit the tasks files. Since there are several other
Blends tools all dealing with these tasks files this is more than worth
the effort.
What really belongs to the Wiki is more general information about the
goals of Debian Astronomy, a link to the (to be written) policy document
(see above), a link to the Alioth project etc.
In case you agree with the strategy to start Debian Astronomy (or
whatever name you might choose) I'd recommend also to start more fine
grained tasks and I'd be happy to help creating these. I guess
astronomy has some certain subtopics you are covering with different
packages and this could be reflected in well designed tasks.
> What do you think?
I think that it would be a great idea to have a Debian Astronomy Blend,
that the conditions to start now are excellent and I'd be happy to
support this effort.
> What should be the topics we can deal with?
I have covered all topics I'm imagining from a technical perspective.
I can not comment on astronomy topics.
Kind regards and thanks for all you work on making Debian fit for
astronomers
Andreas.
[1] https://wiki.debian.org/DebianPureBlends/SoB
--
http://fam-tille.de
Reply to: