by Graf Zahl » Sun Apr 30, 2017 6:32 pm
Sorry, but this is pretty much undebuggable.
The biggest roadblock are those ACS binaries which I cannot decompile or disassemble to rule out bad code generation as a culprit. To use dynamic strings from ACS libraries they need to follow very strict procedures of how to deal with strings. So far every time someone reported string corruption it was caused by badly generated byte code.
I need something smaller and more contained where I am capable of seeing what the scripts are actually doing. This is 1.8 MB of ACS binary code, all of which may indirectly be responsible for the problems.
Sorry, but this is pretty much undebuggable.
The biggest roadblock are those ACS binaries which I cannot decompile or disassemble to rule out bad code generation as a culprit. To use dynamic strings from ACS libraries they need to follow very strict procedures of how to deal with strings. So far every time someone reported string corruption it was caused by badly generated byte code.
I need something smaller and more contained where I am capable of seeing what the scripts are actually doing. This is 1.8 MB of ACS binary code, all of which may indirectly be responsible for the problems.