Re: Kernel SCM: When does CVS fall down where it REALLY matters?

[Mike Fedyk <mfedyk@matchmail.com>]

On Fri, Mar 08, 2002 at 06:52:38PM -0700, Val Henson wrote:
> On Thu, Mar 07, 2002 at 05:38:27PM -0700, Erik Andersen wrote:
> >
> > 6) Ability to do sane archival and renaming of directories.
> >     CVS doesn't even know what a directory is.
> 
> How about sane renaming of plain old files?
> 
> For a laugh, read the instructions on how to "rename" CVS files.
> Hint: "Rename" is not the correct word.
> 
> $ mv old new
> $ cvs remove old
> $ cvs add new
> $ cvs commit -m "Renamed old to new" old new
> 
> Gee, that looks like adding a new file to me.  Upon reading further,
> that is exactly what this "rename" operation is doing.  There are two
> other ways to rename a file in CVS, one of which is described as
> "dangerous" and the other as having "drawbacks."  References:
> 
> http://www.gnu.org/manual/cvs-1.9/html_node/cvs_66.html
> 
> Note that the way to rename a file in in BitKeeper is:
> 
> $ bk mv old new
> 
> No danger, no drawbacks, no hand editing of history files.
> 
> I strongly recommend that anyone attempting to make CVS a viable
> replacement for BitKeeper start out by actually using BitKeeper.
> You're so used to being crippled by CVS that you don't even know what
> you're missing.
> 

No.

They're not trying to make cvs fit into the space bk lives in now, they're
trying to take the cvs *replacements* (arch, subversion, etc) and make them
usable (they probably are close now, but not as good as bk) for kernel
development requirements.