> (2) CVS "tries to be really nice" is not nearly the same as "succeeds."  I
> have other local patches that regularly cause conflicts in one of my
> trees, and CVS doesn't have the faintest idea how to deal with them.  One
> of these is a PR I originated aeons ago, and virtually every time someone
> has touched the affected file, I've had to manually reintegrate my patch.

When cvs discovers a conflict, it attempts a "rcsmerge" operation. If
this fails, it lets you know that a conflict has happened.  The result
is <<<<< and ===== and >>>>>> merge directives in the file.  You
simply need to fix these up.

Compared to the work of dealing with sup deleting changes in files,
this is an easy task.

*REALLY*.  IT IS THAT EASY!  cvs never messes up; an rcs merge conflict
is just it's way of dealing with a conflict -- to give you a chance to
help it.


Secondly, I can't find any open high-impact PR's from you. Can you
give me the number so that I can check whether we have integrated it?
If it is correct, I will merge it into OpenBSD.


> > Now, when looked at this way, I'm not surprised that Jukka is
> > complaining, or that Scott has missed the point.
> 
> Once again, Theo has pronounced judgement after asserting his speculations
> as fact.  This astounding lack of integrity is disheartening; I honestly
> hope this doesn't bode ill for the innocent folks using OpenBSD.

Nothing I say affects OpenBSD users much, as long as they can use their
machines.