[BioC] Database for Microarray Data - request for user stories

[Ido M. Tamir]

Hi all,
I hope you don't consider this question completely off topic.
(I didn't find a thread on this in the archive)

How are your experiences with some of the publicly available
database systems for microarray data (including commercial ones)?

I don't want to roll my own unless I have to.

Did you create your own and why?

----
My requirements are:
LIMS ( probe, spotting, annotation )
microarray data storage
accessibility by GUI and API esp. for analysis with
bioconductor (so the question is not off topic ;-))
It does not necessarily need a GUI for bioconductor.
Spitting out gal, gpr and target files is enough.
And for the nimblegen data I would use SQL.

If source code is available it would be a plus
java is also a plus (for bug fixing, enhancements)

Analysis facilities are not that important.
Important is efficient and reliable retireval of 
huge data sets (in house two color and
nimblegen datasets, maybe affymetrix one color).

Ability to easily migrate from our current
database would also be important.

----
MARS seems very interesting but provides no source
http://genome.tugraz.at/Software/MARS/MARS.html

I saw longhorn which is still in development based on
SMD (which is tied to Solaris/Oracle)
http://www.longhornarraydatabase.org/


GEOSS/GeneX2/GeneXl ite
which already confuses me, but some people
say choice is good.
2003-09-20 seems to be the latest news
on genex 2 so the project is dead.

Another dead project seems to be the
yale microarray database:
http://info.med.yale.edu/microarray/status.htm
also 2003/9 latest entry with no db software
download link.

Thank you very much for your answers.

Ido Tamir