I have no issues compiling/building Blender in Windows 10, it’s running fine. Let’s say I compile it in C:\Users\ruyman\Documents\my_blender.
Then I copy my compilation and distribute it to other computers, pasting them in completely different locations in every machine. For instance \bin\custom_blender\
However, when running Blender, the error messages it produces at runtime still point to where it was originally compiled. That’s confusing my users a lot. Is there a way to fix that?
Current situation ERROR (bke.modifier): C:\Users\ruyman\Documents\my_blender\source\blender\blenkernel\intern\modifier.c:409 modifier_setError: Target polygons changed from 4751 to 75968
Goal: ERROR (bke.modifier): \bin\custom_blender\source\blender\blenkernel\intern\modifier.c:409 modifier_setError: Target polygons changed from 4751 to 75968
Or at least removing the original path where blender was built, so the users don’t think Blender is pointing to non-existing paths.
The path is produced by the compiler by expansion of a macro and is referring to a source file which only exists on the build system. this issue will show across all platforms (with slightly different paths naturally)
Not sure what this would be solving, your goal path would not exist either, leaving the user as confused as he was before.
I’d probably look at solving the root cause and try to stop the errors emitting in the first place.
The path is produced by the compiler by expansion of a macro and is referring to a source file which only exists on the build system. this issue will show across all platforms (with slightly different paths naturally)
Thank you, that was my question. So the build path is embedded somewhere in the compiled code, and there’s no way to workaround this. The moment I move my custom build into a different location, the path printed in the runtime error messages will still point to the original source file, regardless of whether it still exists or not, and the location of the actual code being wun
Not sure what this would be solving, your goal path would not exist either, leaving the user as confused as he was before.
My goal path would exist because it is where Blender is actually located in the user’s computer. The user runs my custom Blender that I copy/pasted from my computer into his computer. I buid Blender in my computer in a path, but the user runs my custom Blender in his computer. Blender got there by means of copy/pasting from my computer to his computer (via a pendrive for instance). The user pastes my custom Blender it into any location of his computer, which might or might not match the path from where this custom Blender was actually built. Maybe it’s more clear with this verbosity level?
So I wish the error message would point to where Blender is actually running from (in the machine of the user) instead of the path of my computer (which, exists, only, as you say, in my computer).
I’d probably look at solving the root cause and try to stop the errors emitting in the first place.
I’m sorry if my original question was not clear enough. I do not wish to get rid of the error message. That’s not the purpose of this question, that would be a different question.
This is where you went wrong, the source file referenced here will not exist on the end users computer it will only exist on the computer blender is build on. The file is not shipped to end uesrs
Ah, I see your point, you mean that .c file is no longer part of the build after compilation and it has become some DLL or something? Then yes, the error message maybe should point to where that code ended up being compiled. Or maybe just the relative part blenkernel\intern\modifier.c:409 for debugging purposes.
But anyways, thanks again for confirming that’s what was expected to happen.