info-cvs
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Correct options for export and binary files


From: Bruce Hill
Subject: Re: Correct options for export and binary files
Date: Thu, 12 Jul 2001 16:00:39 -0700

> Eric Siegerman wrote:
> 
> The two worms I can think of are:
>   - The -kb mechanism is fragile, and when it breaks, the file
>     gets corrupted.  Well, that's what this thread is about
>     fixing.
> 
>   - No reserved checkouts.  That's clearly a problem for
>     unmergeable files (whether binary or the weird text formats
>     generated by some tools).
> 
>     Please remind us why adding this capabilty to CVS -- BUT NOT
>     FORCING PEOPLE TO USE IT ON TEXT FILES -- would be so evil.
> 

I think Eric's summarized this thread rather well.
It's also clear that with the exception of a few extremely vocal
individuals, most of those responding to this thread recognize
that these are indeed problems which need to be addressed.

Personally, I favor Eric's prior suggestion of -kvf to solve
the first problem.

> Better would be to add an "f" suboption to -k:
>     - "-kv" would be changed as we're discussing, to say
>       "override the stored keyword setting, except for 'b'"
>     - "-kvf" would do what "-kv" does now -- always
>       overrides the stored setting, even for 'b'
> Note that the "f" should NOT be stored in the ,v file; it should
> be strictly a modifier to the current "cvs" command.
> 
> "cvs admin" should honour this "-kf" suboption too.  This solves
> the problem of someone going "cvs admin -kkv ." and recursively
> clobbering any binary-file settings in their project.

The second problem, no reserved checkouts for non-mergeable file types,
is addressed nicely with Noel Yap's "cvs edit -c" patch.

> Try using the "cvs edit -c", "cvs edit -f"', and "cvs ci -c" patches
> available at www.sourceforge.com under project RCVS. 
> I think I named them reservation.diff.

Now all that's needed is someone with the ability, motivation, and 
time to submit a tested documented patch for each (and the willingness
of the CVS maintainers to approve a good patch).  I've got the
first two, but my boss won't authorize any time for this and I'm
to swamped to do it on my own time.

How about it, any volunteers out there with time to champion this?

- Bruce Hill



reply via email to

[Prev in Thread] Current Thread [Next in Thread]