Brian meeting, 25 Oct 2011
--------------------------
1. Debriefing of Brian survey
-----------------------------
Docs and examples:
* More about using C
* Clarify the use of freeze (does it cause errors or wrong results?)
* Example without magic; more about how to turn it off in the docs
* More docs about model fitting (how to set parameters etc)
* Examples of extending Brian

Other comments:
* Memory problems: optimize creation, use floats, have lighter StateMonitors
  (e.g. with only population-averaged statistics).
  Here's an idea for StateMonitor: how about having a
  user-defined function that takes a vector of state variables and returns
  a single value (or tuple of values)? (e.g. mean or variance) - this would be
  a PopulationStateMonitor)
* More flexibility for synapses and parallel computing are high on the list
* Stiff integration methods
* Parameters in noise and multiplicative noise

User-contributed examples:
  Let's do a competition! "Brian twister"?
  We could announce it at the SfN social, on a Brian poster (oops we have to
  do one).
  There could be prizes (maybe 3?), with some Brian merchandise (tee-shirt etc).
  People have to contribute short scripts that do something cool (possibly an
  implementation of a model in a paper).

2. Plan for GPU
---------------
We might have 1 student in Jan/Feb (from Vilnius). He would work on general
stuff (spike queues etc).
I will also ask for 1 more student from March to June (ENS Lyon M1).
Let's do it by next summer!

3. Next releases
----------------
* Version 1.3.1 (10 November)
	* New SpikeGeneratorGroup: tests
	* Metrics: syntax/tests/examples
	* Brian Hears updates (Tan/Carney)
	* Little bit of docs
	+ new features (see txt file about this)
	[DONE] * Brian examples for our papers

* Version 1.4 (January)
	* New model fitting
	* MultiLinear...
	* Refractoriness
	* Victor's spatializer: syntax
	[DONE] * Brian examples for our papers

4. Synapses: 2012?
------------------
(Dan will have a look at Victor's code)
