format-patch --cc and send-mail --s-o-b-cc don't play nice together

format-patch --cc and send-mail --s-o-b-cc don't play nice together

[Bert Wesarg]

Hi,

the --cc option from format-patch adds a Cc: header to the mail where
multiple recipients are put into a new line, separated by commas.
Unfortunately send-mail with the --signed-off-by-cc option can't parse
these lines.

Any ideas how to solve this?

Regards
Bert Wesarg

Re: format-patch --cc and send-mail --s-o-b-cc don't play nice together

[Junio C Hamano]

"Bert Wesarg" <bert.wesarg@googlemail.com> writes:

> the --cc option from format-patch adds a Cc: header to the mail where
> multiple recipients are put into a new line, separated by commas.

I am not sure if --cc ever made sense to be given to format-patch.  It is
not a tool to send e-mails, which send-email is.

> Unfortunately send-mail with the --signed-off-by-cc option can't parse
> these lines.

Isn't --signed-off-by-cc a completely unrelated feature, where it gives
carbon copies to people who signed off the patch?  Why does the option
need to read from existing CC: header of the material the command is told
to send out?

Re: format-patch --cc and send-mail --s-o-b-cc don't play nice together

[Bert Wesarg]

On Tue, Apr 29, 2008 at 11:39 PM, Junio C Hamano <gitster@pobox.com> wrote:
> "Bert Wesarg" <bert.wesarg@googlemail.com> writes:
>
>  > the --cc option from format-patch adds a Cc: header to the mail where
>  > multiple recipients are put into a new line, separated by commas.
>
>  I am not sure if --cc ever made sense to be given to format-patch.  It is
>  not a tool to send e-mails, which send-email is.
AFAIK, this option is not in the master branch.

>
>  > Unfortunately send-mail with the --signed-off-by-cc option can't parse
>  > these lines.
>
>  Isn't --signed-off-by-cc a completely unrelated feature, where it gives
>  carbon copies to people who signed off the patch?  Why does the option
>  need to read from existing CC: header of the material the command is told
>  to send out?
No, it scans Cc: lines (header and body) too and add these to the cc
list (in the body since abec100c3382f7d7b759f915a86e9773277263b6).

Bert

Re: format-patch --cc and send-mail --s-o-b-cc don't play nice together

[Daniel Barkalow]

On Tue, 29 Apr 2008, Junio C Hamano wrote:

> "Bert Wesarg" <bert.wesarg@googlemail.com> writes:
> 
> > the --cc option from format-patch adds a Cc: header to the mail where
> > multiple recipients are put into a new line, separated by commas.
> 
> I am not sure if --cc ever made sense to be given to format-patch.  It is
> not a tool to send e-mails, which send-email is.

But part of preparing patches for e-mail submission is often figuring out 
where to send them. I think the common use of format-patch is to generate 
the series, import it into some MUA (I actually append it to my "postponed 
messages" mbox for Alpine), edit the messages to add any commentary needed 
outside of the commit message, and send them, going through the user's 
normal outgoing email path. For this purpose, it makes sense to be able to 
get git to produce anything that is standard for patch submission in the 
project, so that the user doesn't have to edit them more than necessary.

I think that it would make sense for format-patch to handle 
--signed-off-by-cc itself, since this is clearly generating useful 
information for composing the email out of git's information about the 
commits.

> > Unfortunately send-mail with the --signed-off-by-cc option can't parse
> > these lines.
> 
> Isn't --signed-off-by-cc a completely unrelated feature, where it gives
> carbon copies to people who signed off the patch?  Why does the option
> need to read from existing CC: header of the material the command is told
> to send out?

It also Cc's people listed in Cc: lines in the commit message, which I'd 
guess is where it's getting confused. But it also makes sense that people 
might want to determine who to send each patch to individually, put it in 
the patch in the RFC823-compliant fashion, additionally include people in 
Signed-off-by: lines, and send the whole batch off, in which case 
send-email would need to know how to handle existing Cc: headers which may 
be wrapped.

	-Daniel
*This .sig left intentionally blank*