Mr. Jones, thanks for the quick response.
> Someone else [[EMAIL PROTECTED]] reported this same bug just
> last week and there was some discussion about what to do -- the
> bug is, as you surmise, that update fails to detect a conflict
> when a locally-modified file has been removed by someone else.
> Check the archives (http://www.egroups.com or
> ftp://ftp-mailing-list-archives.gnu.org) for the thread titled
> "Possible bug with reviving dead files". The bottom line is
> that I'm not sure what should happen in this case and I've
> asked for input.
Ok, I'll volunteer some input:
In principle, I think it should not be possible to change the
status of a file from "Dead" to "Exp" simply by committing.
Something like a `cvs add' should be required.
In implementation, how about this. We make `cvs update' report a
conflict on files that are modified in the local tree, dead in
the repository, and have not been `cvs add'-ed. We make `cvs
commit' fail its up-to-date check when it sees such a file.
Thus, the user is forced to either `cvs add' or `rm' the file
before committing and is reminded of the need to do so on each
update.
How does that sound?
-Ken
