Re: Results of egrep -l '^<<<<<<< |^=======$|^>>>>>>> |^\|\|\|\|\\|\| '

[Paul Sander]

>--- Forwarded mail from address@hidden

>[ On Friday, July 20, 2001 at 14:43:22 (-0700), Paul Sander wrote: ]
>> Subject: Re: Results of egrep -l '^<<<<<<< |^=======$|^>>>>>>> 
>> |^\|\|\|\|\\|\| '
>>
>> That is not at all what I conveyed.  My question was:  What if the files
>> that contain the conflict mark-ups are supplied by the vendor, to be
>> checked in on the vendor branch?  Do you now expect CVS to be able to
>> tell the difference between vendor branches versus any other branches,
>> and allow the user to commit files containing arbitrary text on vendor
>> branches while refusing commits elsewhere if files contain conflict
>> mark-ups?

>Well, that's pretty easy, isn't it.  You never ever do any commits to
>the vendor branch in the first place!  Obviously I'm not proposing that
>"cvs import" reject files containing constructs that appear to be merge
>conflict markers.  After all "cvs import" doesn't work with potentially
>merged files and so it can safely assume that anything in the files it
>handles is exactly verbatim what should be in them.

Fine.  I've imported the vendor's sources.  Now I merge them into the trunk.
CVS now refuses to commit my merge because the vendor has supplied conflict
markers which appear in the output of the merge even though CVS did not
introduce them.

Now, I want my code to be as close to the vendor's as possible.  That means
I don't want to escape, quote, or otherwise hide the conflict markers.  I
know I want to do this because I know that the vendor will reject changes
like this that appear gratuitous (i.e. they have no technical purpose to
be introduced to the source, from their point of view).

On the other hand, I MUST escape, quote, or otherwise hide the conflict
markers just to get the file under source control.  Now, when I do my
"cvs diff | mail", the vendor sends me two things:  A nastygram reprimanding
me for sending in a gratuitous change (for the umteenth time, because we
exchange sources on a regular basis), and a new update with the conflict
markers restored.  And the new update must again be merged, which removes
the escape/quote/hide mechanism that I must once again re-add after the merge.
And the process repeats.

This is bogus in a big way.

>--- End of forwarded message from address@hidden