Re: [PATCH 2/2] rebase -i: use some kind of config file to save author information

Christian Couder [off-list ref] writes:

This patch would use the message from HEAD instead of "$DOTEST"/message, but 
it looks like we are changing "$DOTEST"/message sometimes with 
the "make_squash_message" function.

Heh, that is why it was "something along this line" patch ;-)

Regarding the C rewrite vs rebase--i.sh update, I tend to agree with Dscho
that changing the scripted Porcelain is not worth it if we are rewriting
the whole thing in C soon.  But perhaps we can allow combinations of the
two options ("-[cC] commit" and "-F file") given to "git commit"?  The
intent of the caller in such a case is quite clear---use the authorship
but do use the message from the other source (if we do this, it would
probably make sense to do that also for "-m message").  The version
entirely written in C obviously does not even need such an option (it can
read authorship from HEAD and use its own message), but the point is if
going that route would eliminate the need to store "which commit were we
dealing with when we gave the control back" information on disk.  I
suspect that the sequencing information is already on disk (i.e. $TODO
file) and author-script may be redundant information.

On Tuesday 23 June 2009, Junio C Hamano wrote:

Christian Couder [off-list ref] writes:

quoted

This patch would use the message from HEAD instead of
"$DOTEST"/message, but it looks like we are changing "$DOTEST"/message
sometimes with the "make_squash_message" function.

Heh, that is why it was "something along this line" patch ;-)

Regarding the C rewrite vs rebase--i.sh update, I tend to agree with
Dscho that changing the scripted Porcelain is not worth it if we are
rewriting the whole thing in C soon.  But perhaps we can allow
combinations of the two options ("-[cC] commit" and "-F file") given to
"git commit"?

I agree that it would be nice.

The intent of the caller in such a case is quite 
clear---use the authorship but do use the message from the other source
(if we do this, it would probably make sense to do that also for "-m
message").  The version entirely written in C obviously does not even
need such an option (it can read authorship from HEAD and use its own
message), but the point is if going that route would eliminate the need
to store "which commit were we dealing with when we gave the control
back" information on disk.  I suspect that the sequencing information is
already on disk (i.e. $TODO file) and author-script may be redundant
information.

Yeah, perhaps we could ge this route, but I'd rather just port what I can 
step by step first and then polish and/or rewrite some parts than the other 
way around.

Thanks,
Christian.