[Freeciv-Dev] Re: Introduction and Patch Approval Process

[Raimar Falke <hawk@xxxxxxxxxxxxxxxxxxxxxxx>]

On Wed, Aug 15, 2001 at 04:56:40PM -0700, Roy Tate wrote:
> Hi,
>   I've been playing FreeCiv for a few months, and monitoring
> the development mailing list for a few weeks now.  I just
> downloaded release 1.12.0 and did a make on TurboLinux 6.5.
> I'll probably read the source code over the next few months,
> and then start helping with patches and patch reviews.
> 
> I have been programming in C, C++ and Java for 5 to 15 years,
> but up to now I have written for DOS/Windows most of the time.
> 
> I like the idea of a patch approval process:
> 
> 1) Developer submits a patch
> 2) Someone is assigned as a reviewer
> 3) Reviewer offers feedback on the patch
> 4) If the patch is recommended by the Reviewer,
>    it goes the list for discussion.
> 5) If the patch is approved by one or more "core" developers
>    with experience with the areas of code affected by the
>    patch, it becomes a candidate.
> 6) If 3 core developers approve the patch, it is submitted
>    to CVS.
> 
> Reviewer's responsibility:
> - Check code formatting and style, portability and correctness
> - Apply the patch and test the code
> - Respond with suggestions, corrections, rejection or approval
> 
> The "core" developer is an experienced developer who is familiar
> with the code-base, and particularly with the area affected by
> the submitted patch.  His responsibilities:
> - Check the code just as a Reviewer would
> - Apply the patch
> - Walk through the code in a debugger or execute the code in
>   a test program.
> - Recommend approval, rejection or on-hold status
> 
> Note that the reviewer doesn't reject the patch, although they
> might pass on such a recommendation to the developer.  Also,
> please understand that the reviewer is also a developer, but
> they don't have to be as familiar with the FreeCiv codebase. 
> This isn't the whole process, but it is a start.

I'm not sure if a more formal approach is needed. IMHO just a bit more
dedication of the readers of freeciv-dev and other people is
needed. Look how many emails yesterday had come. If all the readers
continue commenting like yesterday freeciv wouldn't have a problem.

However one thing you mention above would be helpful: each "core"
developer should have an area of interest. This information should be
published. This isn't to cut the mails to freeciv by sending the patch
before to such a "part maintainer" but to have at least one person
from which the patch submitter can expect a comment.

        Raimar

-- 
 email: rf13@xxxxxxxxxxxxxxxxx
 "USENET is *not* the non-clickable part of WWW!"