On Mon, Jun 17, 2013 at 06:49:19PM -0700, Walter Bright wrote: > On 6/17/2013 6:28 PM, Brad Roberts wrote: > >Don't symbol names from dmd/win32 get compressed if they're too long, > >resulting > >in essentially arbitrary random binary data being used as symbol names? > >Assuming my memory on that is correct then it's already demonstrated that > >optlink doesn't care what the data is. > > Optlink doesn't care what the symbol byte contents are.
It seems ld on Linux doesn't, either. I just tested separate compilation on some code containing functions and modules with Cyrillic names, and it worked fine. But my system locale is UTF-8; I'm not sure if there may be a problem on other system locales (not that modern systems would actually use anything else, though!). Might this cause a problem with the VS linker? T -- It only takes one twig to burn down a forest.
