In certain cases, selecting the logfile generated by spikes from a WU component will cause the application to hard lock with high processor utilisation. This doesn't occur when selecting spike logfiles generated by neuron bodies or when selecting WU logfiles generated from analog outputs.
I get this behaviour 100% reliably in a model connecting a population of 250 neurons to a population of 340 neurons using an all-to-all connection scheme in an explicit list.
In certain cases, selecting the logfile generated by spikes from a WU component will cause the application to hard lock with high processor utilisation. This doesn't occur when selecting spike logfiles generated by neuron bodies or when selecting WU logfiles generated from analog outputs.
I get this behaviour 100% reliably in a model connecting a population of 250 neurons to a population of 340 neurons using an all-to-all connection scheme in an explicit list.