Synaptic modification: STP, LTP, and STDP

One of the problems with working with lots of technologies under development is that things aren't always ready.
One problem we have is that the learning mechanisms that work in Nest, don't work on SpiNNaker.
Our Java parser did binding with STP. The Nest STP model is Tsodykys-Markram, which doesn't seem to be STP to me.
We could bind in a fashion with that, but that's not yet ready on SpiNNaker.
Similarly, any form of learning reduces the numbers of neurons per chip.
So we've not explored LTP (STDP) in PyNN.
Eventually, we should be able to put our own STP and LTP algorithms onto PyNN and then onto the chips. For that matter we should be able to put our neural model on.
We'll look into that in the new year.