Mail Archives: djgpp-workers/2000/12/12/12:34:56
> I only mentioned other ports because, when I suggested to replace a
> colon with a semi-colon, you said that compatibility with other CVS
> ports will take a blow.
Point taken.
> I don't see how text file format is a problem: I expect CVS either to
> use text-mode reads or remove CR before LF after reading the files.
It uses text mode reads. This means that if a sandbox is shared
between a DJGPP CVS and a Linux one, text files would get dossified(tm)
frequently, causing potential trouble.
I guess the only thing I'd want is to make sure that CVSes print
sensible error messages if used on an incompatible sandbox.
> If compatibility with other ports _is_ important, converting the
> result of _fixpath to use backslashes is a trivial loop, so it would
> be easy to add.
True. The backslash thing wasn't the problem; the requirement of the
:local: prefix might be (i.e. should we require/prepend a :local:
prefix just because the Win32 CVS does?).
> ??? _fixpath does not resolve SUBST and JOIN. Perhaps you were
> thinking about _truepath, which I didn't suggest for this very
> reason (and also because _truepath can return a UNC).
My bad. In that case, using /dev/ internally, and fixpathing the root
whenever its shown may be an elegant workaround.
> If you find the drive support too painful to add, I'm willing to
> volunteer to add it and send you patches.
No need - I'm quite willing to do it - I just don't want to go
through the trouble if there's no real need to do it.
I'll see where the internal-/dev-external-fixpath(/dev) changes
take me.
To Be Continued
- Raw text -