> >
> > With a proposed fix.  I have tested it on my test system and it
seems
> > to work OK.
> >
> >
> >
> > It is nothing fancy, but I thought that it would at lease plug a
whole
> > where data may currently be lost.
> >
> >
> >
> > I suppose one thing it does raise is the possibility of tool that
can
> > be run that will re-generate the given post commit command and
update
> > the Bug if required.
> 
> I am not happy with the approach taken here. If there is an error with
a
> backend, the error should be reported and the commit should be
aborted.
> That is the fundamental premise of all the verification checks. I
won't
> apply the fix. I believe bug 1014 is invalid.

The issue arises that the adding to the comments is done in a post
commit hook.  A post commit hook does not return to the client any error
code or output.

Ref:
http://svnbook.red-bean.com/nightly/en/svn.ref.reposhooks.post-commit.ht
ml

It is effectively a Notification of Successful commit so at this point
the commit can not be rolled back.

The only way I can think of informing the user that the bug has not been
successfully updated is to E-mail them. (As there is no way of getting
the error displayed on the screen).

Thanks

Rob


_______________________________________________
scmbug-users mailing list
[email protected]
http://lists.mkgnu.net/cgi-bin/mailman/listinfo/scmbug-users

Reply via email to