feature tips and small bugs.

[Bernd Gehrmann]

On Fri, 18 Jun 1999 Ansley, Michael wrote:
>I think the point here is that the kdevprj file is updated needlessly in
>certain situations, which will cause any source control mechanism to think
>that there are real changes to it, which require a commit.  This would be a
>problem.  Of course, this may not actually be the case, but I think that was
>the intention of the mail.

Aha. In principle, there is no reason to actually commit a file
that cvs reports as 'modified'. But I see, there is a certain
psychological effect that one usually wants 'tabula rasa' after
a commit and no files marked as modified. A naive solution would
be to open the file read-only by default and re-open it read-write
if it must be actually modified. Don't know if this is feasible
without too much pain.
Sandy?

Bernd