Disable emacs-diffs mails for scratch branches

Disable emacs-diffs mails for scratch branches

[David Engster]

Subject says it all. Could we please do that? I think it's pretty normal
for scratch-Branches being deleted and pushed anew, and for large
branches like cedet-merge or cedet-tests this needlessly creates
hundreds of mails each time.

-David

Re: Disable emacs-diffs mails for scratch branches

[Stefan Monnier]

> Subject says it all. Could we please do that? I think it's pretty normal
> for scratch-Branches being deleted and pushed anew, and for large
> branches like cedet-merge or cedet-tests this needlessly creates
> hundreds of mails each time.

I agree that those 300 emails dumps are more annoying than useful.
But FWIW, I see two problems with your suggestion:
1- I don't know how to do it.
2- I do find it very useful to see commits on scratch branches when
   those branches are used for gradual development (e.g. the byte-switch
   branch).

Maybe the problem can be solved by sending one email per push instead of
one email per commit.


        Stefan

Re: Disable emacs-diffs mails for scratch branches

[David Engster]

Stefan Monnier writes:
>> Subject says it all. Could we please do that? I think it's pretty normal
>> for scratch-Branches being deleted and pushed anew, and for large
>> branches like cedet-merge or cedet-tests this needlessly creates
>> hundreds of mails each time.
>
> I agree that those 300 emails dumps are more annoying than useful.
> But FWIW, I see two problems with your suggestion:
> 1- I don't know how to do it.

We'd have to adapt the post-receive hook, or whatever else is triggering
the mails.

> 2- I do find it very useful to see commits on scratch branches when
>    those branches are used for gradual development (e.g. the byte-switch
>    branch).

How about simply not naming those branches with the 'scratch' prefix,
then?  I was under the impression that 'scratch' means "can be
force-pushed/deleted any time, don't base your work onto it".

-David

Re: Disable emacs-diffs mails for scratch branches

[Stefan Monnier]

>>> Subject says it all. Could we please do that? I think it's pretty normal
>>> for scratch-Branches being deleted and pushed anew, and for large
>>> branches like cedet-merge or cedet-tests this needlessly creates
>>> hundreds of mails each time.
>> I agree that those 300 emails dumps are more annoying than useful.
>> But FWIW, I see two problems with your suggestion:
>> 1- I don't know how to do it.
> We'd have to adapt the post-receive hook, or whatever else is triggering
> the mails.

Yes, that much I know, but I also know that we don't have direct access
to it.  So it depends on Savannah admins, and if I was them I wouldn't
just accept random patches, since AFAIK they use an existing mail-hook
and so they probably want to stay close to the upstream version to be
able to incorporate future changes.

So, we should figure out exactly what choices we have available.

>> 2- I do find it very useful to see commits on scratch branches when
>> those branches are used for gradual development (e.g. the byte-switch
>> branch).
> How about simply not naming those branches with the 'scratch' prefix,
> then?  I was under the impression that 'scratch' means "can be
> force-pushed/deleted any time, don't base your work onto it".

AFAICT the byte-switch does belong to scratch.  Being able/allowed to
force-push/delete doesn't mean that you have to do it all the time.


        Stefan

Re: Disable emacs-diffs mails for scratch branches

[Lars Ingebrigtsen]

David Engster <deng@randomsample.de> writes:

> Subject says it all. Could we please do that? 

Can't you just kill-file the messages you're not interested in?  I mean
score down.  That's what I do with the scratch/ branches...

-- 
(domestic pets only, the antidote for overdose, milk.)
   bloggy blog: http://lars.ingebrigtsen.no

Re: Disable emacs-diffs mails for scratch branches

[David Engster]

Lars Ingebrigtsen writes:
> David Engster <deng@randomsample.de> writes:
>
>> Subject says it all. Could we please do that? 
>
> Can't you just kill-file the messages you're not interested in?  I mean
> score down.  That's what I do with the scratch/ branches...

Yeah sure, I do that as well. It's just that I'm kinda old-school and
think that force-pushing an amended commit shouldn't trigger 300
mails...

But since this doesn't seem to be easily changed, I guess we just live
with it.

-David