[Cvs-dev] Re: cvs-chpass.patch (former: cvs-passwd.patch)

[Prasad J Pandit]

On Tue, 20 Jun 2006, Mark D. Baushke wrote:

 Hello Mark,

> I got your message, but I will not be able to do
> anything with the patch right now as real work
> deadlines rush up against my holiday plans
> (starting next week).
>
> It would be great if another member of the cvs-dev
> could do something to aid you in vetting the code
> and seeing how it looks in the top-of-tree
> sources, but failing that I won't be able to do
> much of anything with this stuff until probably
> late July.
>
> I hope and expect that Derek is likely to release
> cvs 1.12.14 fairly soon (a few finishing touches
> to get windows-NT building again as well as the
> SGI systems), so your patch will probably not make
> it into that release.

 Okay.


> > with following major changes...
> >
> >     + First of all, the 'passwd' command is now renamed as 'chpass'; As for
> >       the interoperability issues(with cvsNT passwd), indicated earlier, by
> >       Mark.
>
> Did you raise the issue of using 'chpass' with
> address@hidden and "cvsnt-dev"
> <address@hidden> in the client/server
> protocol yet (if so, I missed it)?

 No, I did not. But what would I ask there? I mean, it was just that, you 
told me, cvsNT also has the command named 'passwd', which does some more 
things beside changing the user password; And that, it would be better on 
my part, if I change command name 'passwd' to something else. So I did it.

> A copy of the snapshot of the ccvs FEATURE branch
> may be found here:
>
>  http://www.baushke.org/cvs/ccvs.20060620174602Z.tar.bz2
>  http://www.baushke.org/cvs/ccvs.20060620174602Z.tar.bz2.sig
>
> do let me know after you fetch the files as my
> quota on that system is very near full.

 done; I've got the file. Thank you!

> A few quick points... regarding
> http://www.cdacbangalore.in/~prasad/tools/cvs-chpass-patch.tar.bz2
>
>  1) patches that contain just whitespace changes
>     are frowned upon (see your server.c patch
>     where you added a tab or your cvs.h file
>     where you removed a blank line).
>
>  2) It actually is possible to write better tests
>     for passwd that do not need to assume that
>     the system has been configured for :pserver:
>     by effectively replaying the client side
>     responses at a "$testcvs pserver" invocation.
>     Examples may be found in the current
>     sanity.sh, but they are fairly opaque, so
>     maybe someone can help you.

 Yeh I know! I'll see to it.

Thanks!
--
regards
   -Prasad
PS: Please don't send me html/attachment/Fwd mails