Re: Problems with RCS 5.7 on Windows XP (within emacs)

[Aaron S. Hawley]

On Sat, 15 Oct 2005, Gabriel R. Toro wrote:

> Hi,
>
> I use rcs 5.7 in windows (from within emacs 20.7.1), and I have two
> problems.  The first may be more emacs than rcs related.  My apologies if
> this is the case.
>
>
> The first one is that if I exit emacs before checking in the file I am
> editing, then the next time I try to edit the file, emacs tells me that
> another user (which happens to be myself) has the file locked for
> editing.  I have to go through the routine of notifying myself that I stole
> the lock.  Is there a way to avoid this nonsense?

I don't have access to a Windows XP(R) machine with both Emacs 20.7.1 and
RCS installed, so I can't concretely investigate the problem.  The Emacs
maintainers will simply ask you to upgrade to the CVS version before they
would test it.  I would report the problem to whoever prepared the Emacs
binaries.  If it isn't you, of course.

> The second question relates to the RCS_LF_ONLY environment setting.  About
> half of my editing of source files in Windows is with files that were
> originally created in Unix machines and still have only LF at the end of
> the line.  The other half are files created in Windows (with CR-LF).  I am
> currently using RCS_LF_ONLY=something and it appears that the *,v files for
> Windows source files contain copies of every version of the entire file,
> instead of just differences.  Is there a way to get rcs to handle both file
> types efficiently?  Also, I cannot seem to find the place where this option
> is documented.

Assuming the environment variable is behaving correctly, I imagine you
would need to set RCS_LF_ONLY when working with Unix files, and unset
RCS_LF_ONLY when working with Windows(TM) files.  Doesn't sound like there
is currently any better way to deal with it then that, besides dealing
with Windows files as binary data and relinquishing the ability to output
the line differences.

/a