Re: PATCH: backport of symlink fix for issue 142 to cvs 1.11.9.1

[Paul Edwards]

"Derek Robert Price" <derek@ximbiot.com> wrote in message 
news:mailman.2596.1067352558.21628.bug-cvs@gnu.org...
> In addition to the destabilizing potential of the patch, I was inclined
> to consider this a new feature and not a bug fix since it has been
> broken for so long that few long-time users of CVS expected it to work.

I'm a long term user of CVS, and I had no idea that such a
problem existed.

> Run the question by info-cvs and be sure to point out the destabilizing
> potential (which we do not generally inflict _intentionally_ on the
> "stable" branch, Paul - regardless of how much one or two users would
> like to see a feature in the stable branch, I don't like intentionally
> violating the contract we created when we advertised the branch as
> "stable"

Stable to me means bug fixes only.

I'd be equally happy if there was a check in there for Lockdir
which says "your repository is symlinked, you can't use lockdir",
should I ever have the misfortune of trying to activate this
feature unwittingly.

In the absence of someone willing to write such code, but
in the presence of someone who is willing to write code that
fixes that by making lockdir work with symbolic links, as I
would have expected at the time it was written, I'd rather
have the latter.

This is speaking as an end-user who may unwittingly enable
Lockdir, expecting it to work.

BFN.  Paul.