Strange redefinition error upon Brian upgrade to version 2.4+git: handling of multiple runs in standalone not consistent with previous Brian versions

About the C/C++ top cap of multiple runs: @mstimberg I am ok with this limit in general, so far, but perhaps having an option to lift it, issue a warning that nonetheless that could result in very high consumption of memory resources could also be useful. Brian would be delivered with the option disabled. But then the user can enable it at her/his own risk. After all, the potential issues that you mention, are bound to the user’s computational infrastructure, and say, a hard limit of N multiple runs on one platform may be constraining on another one with very large resources…

Now I’m confused. What do you mean with “this limit” @mdepitta? As I explained above, Brian does not impose any limit itself, the only limits are due to memory and/or limits of the operating system (maximum length of command line, for example).

@mstimberg Sorry, drop my suggestion then: I misunderstood your reply: I thought that Brian was implementing some internal locking mechanism itself on multiple runs in standalone to prevent exhaustion of memory resources or max length of command-line errors. Since this is not the case then, the proper question escapes Brian’s realm and is how to estimate such limits on someone’s machine/compiler options.