> OK, I've bashed it around a fair bit and can't find any regressions over
> the previous code, so I'll commit the code. I don't have any threaded
> test code to properly test SQLCancel with at the moment though, so
> perhaps the original reporter can test that.
Ok, I'll include it in next development snapshot.
Thare is a lot of problems with SQLCancel. It needs huge changes.
I think I clean the code at first. There is a lot of dead ends.
Thanks,
Luf