RE: How well does CVS handle other types of data?

[Greg A. Woods]

[ On Wednesday, July 11, 2001 at 16:07:14 (-0400), Jeff King wrote: ]
> Subject: RE: How well does CVS handle other types of data?
>
> I think there's a major difference between datatypes that are occasionally
> difficult to automatically merge (source code) and datatypes that are
> fundamentally unmergable (bitmaps).

But bitmaps are not fundamentaly unmergable.  There are all kinds of
tools to mix images in various ways.  The problem is that the intent of
a change in an image may be orthogonal to the intent of a change in
code.  Normally conflicting code changes are either fundamentally in
conflict (as a change in an image may be), or completely unrelated and
just accidentally (by way of the diff algorithm) in conflict (just any
change in an image may be too).

Noel's description of gross changes to source code that can cause the
overly simplistic and overly optimistic text merge algorithm used by RCS
and CVS to fail was right on the money.  When drastic all-encompasing
changes are to be made there are major planning issues to be resolved
first (eg. all branches that will be merged have to be merged first,
etc.).  This suggests that special measures need to be taken any time
merging becomes impossible for any reason.  I think keeping binary files
completely outside of CVS is simply a special measure that has to be
taken to maintain CVS' ability to offer its full feature set.

-- 
                                                        Greg A. Woods

+1 416 218-0098      VE3TCP      <address@hidden>     <address@hidden>
Planix, Inc. <address@hidden>;   Secrets of the Weird <address@hidden>