Sunday, November 11, 2012

Sun Nov 11 11:51:33 CET 2012

Armistice Day, Remembrance Day or Veterans Day -- just choose the one you prefer ...  a new update ... after two years of silence ...

Things have happened -- for the good and for the bad.

Briefly: (1) there is now a new biochemical pathway solver, Chemesis-3 -- working but immature, (2) the network modelling functionality is now available from the gshell -- many thanks to Dave Beeman for his constant commitment for making this functionality useful and his understanding of the design of the system's infrastructure, (3) the python bindings have been further developed, thanks Mando, but they also have problems that need to be resolved -- we are using the gshell interface to implement a better and improved python API, (4) we organized a first workshop in Luebeck a couple of weeks ago -- some users tried interfacing with NeuroConstruct, with some success, (5) we got two papers accepted and published, and one in the pipeline (but unfortunately rejected for silly reasons so far -- a couple of years ago a reviewer of the same group gave us a 4/10 for English writing and figures ... guess what ... -- ah ... let's hope he leaves a comment here :-) science is about open discussion, conversation and convergence).

After pretty all of the core functionality has been finished, interfacing -- in the general sense of the word -- is now the highest priority for me.  Interfacing outward means graphics -- we have already done some work there: (1) Mando and Dave implemented a couple of python widgets that comply with the user workflow, (2) just recently I implemented a developer GUI that supports the developer workflows.  For sure, there is more to come here.  Horizontal interfacing is about interoperability with NeuroML / NineML, NeuroConstruct, PyNN and others.  So that's already started by user demand, see above.  Inward interfacing??  That's the essence of the CBI architecture, and, for instance, includes multi-scale modeling ... you can read about it here The CBI Federated Software Architecture.

Now, the bad news: my own personal situation and relation with G-3 development ...  Between October 2010 and November 2011 I was not being paid -- a complicated story, trials, blocked funding etc.  Then since November 2011, I am working for about 25% on G-3.  That is something, but not much.  Development has been very slow during the last two years and will continue to be slow if the situation persists the way it is today.  That's bad.

We hope that the competitive renewal that we just submitted will help -- but uncertainty is the master in sciences nowadays and those who master uncertainty have always the advantage.  I still hope that the uncertainty will never become a certainty -- that would require manager skills.