Page 1 of 1

LUA, footbot going directly at destroy()

Posted: Thu Mar 14, 2019 3:30 pm
by guiMaitre
Hello,

The problem:
Some of my footbot (not always the same) will directly go to the function destroy(). The proportion is between 2-10 footbot destroying themself out of 100 footbot. The screenshot show what is happening.
The code to write the log is in the destroy function and the only way to get there is the end of the experiment (50000 ticks normally ).

After spending some times looking for an answer or something wrong in my code that would run the destroy function, I don't see anything wrong and get 0 error of compilation. Is this normal or something like the 2 footbot spawning in each other and destroying each other?

It would be great to know how to avoid this problem because I need the right amount of footbot.

Thank a lot for the futur response,
Best regards.

Re: LUA, footbot going directly at destroy()

Posted: Thu Mar 14, 2019 3:35 pm
by pincy
Without your code it's impossible to tell. Please share your code and your argos file.

Re: LUA, footbot going directly at destroy()

Posted: Thu Mar 14, 2019 3:49 pm
by guiMaitre
Sorry here is the code.

Re: LUA, footbot going directly at destroy()

Posted: Thu Mar 14, 2019 3:54 pm
by pincy
I can't run your files because one_spot.png is missing.

Re: LUA, footbot going directly at destroy()

Posted: Thu Mar 14, 2019 3:59 pm
by guiMaitre
Everything is here this time

Re: LUA, footbot going directly at destroy()

Posted: Thu Mar 14, 2019 4:18 pm
by pincy
I tested your code and it works perfectly. I don't observe the issue you mention. What can I do to make it happen?

Re: LUA, footbot going directly at destroy()

Posted: Thu Mar 14, 2019 4:53 pm
by guiMaitre
Some more information: I runned the code with that problem on 2 computer : a MacBook air 2014 (macOS 10.14.3) and a desktop computer (latest Ubuntu available) with 12gb RAM and a coreI7 @ 4.8Ghz. Both of them had the problem.
The version of ARGoS I use is the last one, (started not long ago).
If you need more information on my setup just ask me.
I have an other footbot experiment that do something else working without this bug.
But it's strange since it occur before the simulation even start I think.