[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: [DebianGIS] Bug#588812: [DebianGIS-dev] Debian GIS metapackages available [Was: debian-gis_0.0.1_i386.changes ACCEPTED]



On Mon, Sep 27, 2010 at 12:13:31AM -0700, Hamish wrote:
> Andreas:
> > So what do you think about my suggestions to make Debian GIS more
> > popular than it currently is?
> 
> come on now, we are all intelligent people worthy of mutual respect.

I'm sorry if I have made an impression which might have questioned this.
It was not intended to disrespect your work.  Rather the contrary:
Because I highly respect your work I think it deserves more popularity.

> So what are the suggestion(s) we are talking about??  I take it you just
> mean to rename pkg-grass-general@lists.alioth.d.o to debiangis@lists.d.o
> and pkg-grass-devel@lists.alioth.d.o to debiangis-commit@lists.d.o?
> 
> anything else?

My suggestion is on the one side less on the other side more.  IMHO it
is enough to rename the general list because I would consider this as
the main entry point for newcomers.  (Perhaps bug #588812 is not really
clear about which list to rename.)

The more part is that I'm having the hope that the list will not only
get a new name (and home) but turns into a medium where we actively
can attract users (and later developers) better than before.
 
> > I think it is a shame that Debian GIS is quite unknown to potentially
> > interested persons and please trust me:  I asked a lot of people when I
> > did talks about Blends and mentioned Debian GIS as a
> > potential Blend - there was nobody aware of this project whereever I
> > asked.  And there where people around which are working with GIS
> > software.
> 
> any interested person looking at the apt-cache info of any of our packages
> will clearly see who the maintainer is.

Because you do not seem to trust my experiences in general I would like
to quote a woman who was engaged in the scientific track of Libre
Software Meeting (main annual Free Software event in France) for several
years, a geographer by profession, doing OSM work, etc. - in short a
potential and interested user of Debian GIS software.  She is using
Debian and Ubuntu.  She has heard and my talk in Amiens in 2008 and
Nantes 2009 about Blends and Debian Science that there is a Debian GIS
project and pronounced that she is an interested user but no developer.
This year at Bordeaux LSM she told me:

   I have heard that Debian has decided to package Qgis.  That would
   be really cool.

So what to say?  My first point is that she did not understand the
Do-O-cracy principle in Debian.  Not "Debian has decided" but one (or
more) developers just have done something (packaging Qgis).  If she is
using stable she might not have noticed that qgis is in testing and
obviosely has not checked the Deban website or any other medium to
verify that there are just packages (which might be easy to backport
or not).  So she simply failed to get the information she was seeking.
If there would be a list which is obviosely accepting questions like:
"How can I run Qgis under Debian?"  she might have gotten help.

> I would not think that renaming
> the mailing lists would harm our exposure, but I think it's somewhat
> optimistic to think that that it will bring us too many new committers.

I don't think that it is optimistic - I just can prove that it worked in
the field of medicine and I would say that it is in several ways harder
in the field of medicine to find users and developers than for GIS.  If
you are keen on some reasoning about this I might elaborate on this but
it becomes off topic.  If you compare

   http://blends.debian.net/liststats/authorstat_med.png

against

   http://blends.debian.net/liststats/authorstat_pkg-grass-general.png

you will see a contrary development of active users on the list.  Debian
Med has gathered active users over time and I can say that we do not
just gathered commiters to the list but we have *at least* six new DDs
*because* the Debian Med project exists (and some more are in the
queue).  For all of them the Debian Med project and the mailing list was
the main entry point.  It also serves as contact point for upstream
developers.

The tendency on the pkg-grass-general list is rather contrary.  I do not
want to say that it is just a question of the name of the list.  But I
know for sure from the example above (and others not mentioned here) that
a pure packaging list on Alioth is not a good entry point to start.
 
> i.e. people heavily using the packages are the best dev pool, and they
> already know about us since looking at a package status or filing a bug.

That's for sure.  But did you gave users a reason to use Debian packages
and not Ubuntu, Fedora, OpenSuse, etc.?  Do you think that you have given
a random GIS user an explicite reason why he should use Debian?
 
> > I also continue to think that we should try to involve OSM into
> > the Debian GIS project.
> 
> the door is always open to anyone who wants to show up.. all are welcome.

There are so many open doors in Open Source - have you ever asked why so
many people do not care about passing these doors?  I'm quite convinced
that's because they do not really know what is behind the door and it
really helps if there is some leading hand to help them through.

I'm also convinved that from a technical packaging point of view Debian
is really great and every maintainer in Debian GIS does a great job.  My
main issue in all this Blend stuff is that we fail to make the step from
single package maintenance to providing a complete set of work tools and
advertising it to the users.  Some steps are done by creating some live
CDs.  But how many live CDs a user will really try excessively?  IMHO
the fact that a user can get *support* for potential problems in GIS
software and becomes involved in supporting other users is stronger and
the experience from Debian Med has somehow proven this concept.

> In my experience working with both sides, there exists a slightly
> different community between the DebianGIS/OSGeo crowd and the OSM crowd.
> Not that it is good or bad, and many of us know people from both so it
> is easy to cc an email when needed, but just my 2c observation.
 
This seems similar to the decision to cover not only health record
applications into Debian Med but include also microbiological research
into the project.  The common set is quite small (probably smaller than
DebianGIS/OSGeo and OSM) but you need a critical mass for a team.  The
microbiological software was quite strong and we needed this strength to
grow as a project.  So IMHO it is really clever to actively involve a
strong community into a common project.  Some health care application
upstream authors call Debian "reference distribution" and recommend
their users using it.  I would really love to hear that GIS and OSM
people would say the same in this field.

I do not say that this happens simply because of renaming a mailing
list.  IMHO we should dedicate a mailing list to these tasks and it
should have a proper name and location.

> For my part, I do not object to renaming the mailing list to something
> more relevant, but with 3 conditions:
> 
> *** the mailing list history is preserved
> 
> Alexander Wirt wrote:
> > > Its no problem to import old mails if the come in a common format like
> > > mbox. 
> 
> (I assume the Alioth admins are willing and able to provide the mbox file?)

I think so.
 
> *** Right now this list sends me _zero_ spam. I would be rather unhappy
> if that were no longer the case and

The list policy on lists.debian.org is more relaxed (=not moderated and
not subscribed people can post).  Perhaps we can ask for a specific
policy but I'm not sure about this.  Several people consider the policy
rather a feature than a bug.

> I was forced to set up a local
> spamassassin instance to deal with it,

I wonder whether there is anybody out there without a spam filtering
mechanism - it would be interesting to hear that it would work for you
these days.

> and have to dig through spam when searching through the archives.

The SPAM can be actively deleted from the archives and when I'm calculating
the stats I refered to above I'm providing data to the listmaster which
helps fighting SPAM. 
 
But I admit this condition of yours is probably not fulfilled.
 
> *** the pkg-grass-devel list be renamed debiangis-commit (i.e. what it is)
> to stop the confusion there of humans posting to that list.

This list can definitely remain on Alioth and I do not really consider
this name as to confusing for the target audience which are defintely
package developers who are really interested and educated enough to
accept this historical name.  So if you prefer changing both for the
sake of completeness it is fine.  I'm personally in favour of
completeness and clarity, but this is technically different because it
would only be a rename and no move to a different host and is conceptual
different because it remains for issues which are definitely packaging
related while I want to widen the topic of a general debian-gis list.
Moreover the renaming is more complicated because it has the consequence
of a need to change packages maintained by the Debian GIS team (in all
releases which is not possible).
 
Hamish, I hope this lengthy mail explains enouth that I do not
disrespect anybodies work or intelligence.  I just intend to share
experiences with others for the good of the project.

Kind regards

     Andreas.

-- 
http://fam-tille.de



Reply to: