Hello,
No, the message isn't very informative. It just verifies an assumption, that should be handled by linker code earlier in the process: uninitialized data should always come after initialized data.
The problem, as I understand it right now, is that the size of ".dydat" is *zero*, fooling the linker into thinking it's *uninitialized* data. Since the Microsoft linker can handle this, so should I. I will test some more, and then post a new version for you to test. Probably tomorrow - going to bed now...
I guess I havn't come across an empty section since I wrote the first version of the linker. Congratulations to you for that...
Pelle