a way to submit a patch to be merged without going through code review

[John A Meinel]

You might want to mark this as Won't Fix or Invalid, but it is at least something to consider.

Occasionally it is useful to submit trivial work directly to the mainline without going through a formal review process. This is either because it really is trivial (simple typo fixes, etc), or because you are tweaking a patch that is already reviewed.

As tarmac only uses the launchpad code review for its queue-of-things-to-merge you have to go through the full process.

One possibility to at least make it lighter weight is if you could propose a merge request and mark it as 'approved' in the same step. (Or at least as approved before the diff email gets sent out....)

Just some thoughts.

[Michael Hudson-Doyle (mwhudson)]

My instinctive response to this is that it would be better to make the full workflow so slick and awesome that we don't feel the need for something like this.

[John A Meinel (jameinel)]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Michael Hudson wrote:
> My instinctive response to this is that it would be better to make the
> full workflow so slick and awesome that we don't feel the need for
> something like this.
>

Absolutely, as long as you can define what such a slick-and-awesome
workflow would look like. :)

Certainly at the moment resubmitting for a trivial fix generates 3
emails, 4 if you want to include a comment. Which is 2 more than "submit
this, someone else reviews and approves". Certainly IMO resubmitting
seems like it should be less overhead than an original submission.

John
=:->

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (Cygwin)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iEYEARECAAYFAkpeR0wACgkQJdeBCYSNAAMnMACeLVg8UFjzOpu7LWq+OSOanDJ0
GAsAoNE0FpSrRrPOryubx0LMrm4b1YT3
=n+Vn
-----END PGP SIGNATURE-----

[Aaron Bentley (abentley)]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

John A Meinel wrote:
> Michael Hudson wrote:
>> My instinctive response to this is that it would be better to make the
>> full workflow so slick and awesome that we don't feel the need for
>> something like this.
>
>
> Absolutely, as long as you can define what such a slick-and-awesome
> workflow would look like. :)
>
> Certainly at the moment resubmitting for a trivial fix generates 3
> emails, 4 if you want to include a comment. Which is 2 more than "submit
> this, someone else reviews and approves". Certainly IMO resubmitting
> seems like it should be less overhead than an original submission.

I think resubmit can be improved beyond the improvements we could make
to improve the whole workflow.

Aaron
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iEYEARECAAYFAkpeSsYACgkQ0F+nu1YWqI0IeQCeJlqNKoa8JR2EURSrNO25ay4W
A2wAn0Cgd0LxT2eP9xoJ6ufbenCKwEWp
=OQeH
-----END PGP SIGNATURE-----