Re: different CVS_SERVER for different hosts

[Paul Edwards]

"Larry Jones" <lawrence.jones@eds.com> wrote in message 
news:mailman.2859.1067555624.21628.bug-cvs@gnu.org...
> > And in fact, what I would also like is to be able to have my
> > 32-bit DOS-extended version back again.  I used to have CVS
> > compiling under that environment.  Using EMX 0.9b.  I can't
> > remember which version I what I needed to do to get that
> > working.
>
> There is still EMX-specific code in the source distribution, but so far
> as I know no one has used it in a long time, so it probably would
> require some work to get a functional CVS built.

I really don't understand why Posix-compliant software
would require "some work", when a compile is attempted
under an arbitrary Posix-compliant environment.

That's what the Posix standard is there for!  So long as you
use PATH_MAX etc etc, it should all work.  I can understand
why, with the heavy directory traversal, it would require to
be targetted to only Posix-compliant operating systems or
"environments", rather than working on any ISO C environment,
but it beats me how it can't even compile on ANY 100%
Posix compliant programming environment.

> Feel free to submit patches.  :-)

Ok, I didn't know there was a specific EMX directory.  That is
actually probably what I did.  I was probably doing this around
the time that CVS worked on OS/2 and I would have just run
the same this with dos-extended EMX.

Anyway, the INSTALL file usefully told me that 1.9.22
apparently worked with EMX, so maybe I will start from a
known working base and see what has been destroyed since
then.  Maybe the output of that will actually be a Posix
compliant version of CVS.

Maybe CVS 1.11.x will be the version where we finally
get all those environments working again, instead of them
only working in ancient versions.  :-)

BFN.  Paul.