[SOLVED] Simulation deletes brain file


Good afternoon,

I just wanted to report that I have been running the simulations with SpiNNaker and several times, when I was trying to make some modifications to the brain file, once I pressed the save button, it deleted the file. I have learnt the lesson and now I am constantly making copies, but I do think that it is concerning. Not too sure if it is on the NRP or SpiNNaker side (or somewhere in between the communication).

Best regards


Dear Gabi,

I’m really sorry to read that you lost your work and had to rewrite it! That’s for sure a serious bug. Though, can you tell me which save button you mean? Because brain files are auto-saved and the brain editor has no such button (see image attached).
Do you mean the save button in the Object library pane? (see other attchment).

Best regards,


Thank you for your reply. Apologies for unclear wording, I quickly wrote the post without proof reading. I was referring to the ‘Apply’ button in the brain editor.

I am not too sure if it the right place to report another problem: the platform quite often is not able to detect ROS data types used in the transfer functions. Explicit import does not help and I get messages as:

NoneType’ object has no attribute

Best regards


Dear Gabi,

Few clarifications.
there is no guarantee that when a TF is run, a ROS message be available on a subscribed topic. Therefore, before accessing the message, you should check it test against None.

if camera.value is not None:
    # do stuff
    # do something else, possibly return immediately skipping one timestep

Moreover, any module used in the body of the TF, shall be imported within the body. Importing it before the decorators (outside the body of the TF) will make the module available only to the code outside of the TF.

import math
@MapVariable("pi", initial_value=math.pi)
def t_fun(t, pi):
    clientLogger.info(pi.value) # ok
    clientLogger.info(math.pi) # ERROR, math not available

def t_fun(t):
    import math
    clientLogger.info(math.pi) # OK

I hope it helps,



Thank you for your message. Regarding checking for None, I will try it, I am checking for this condition in all of my other transfer functions, but except the one that was throwing this error, so that might have been the cause, thank you!


Hi Gabi,

I filed your bug for the dev team to look at it:
If you could attach to this ticket your experiment files, it could be easier for them to reproduce the issue.




The bug (deletion of brain simulation file) could not be reproduced by the dev team and has been closed. Please post here if you still encounter it.

Best regards