D pull request review process -- strawman formal definition, query for tools

[Lars T. Kyllingstad]

On Thursday, 9 May 2013 at 21:10:55 UTC, Steven Schveighoffer 
wrote:
> 4) Reviewers' roles and responsibilities.  I am not defining 
> how a reviewer gets assigned, not sure how that should work, 
> and it likely depends on the tool we use.

I think each Phobos module should have an official maintainer.  
The modules' documentation have "Author" sections, but those 
typically list everyone that have ever made a significant 
contribution to a module, and many of those people have long 
since withdrawn from the D community.

I propose that there be additional "Maintainer" sections which, 
for each module, specify *one* person who has the primary 
responsibility for that module.  If and when that person 
disappears from the D community, she or he must be replaced by 
someone else.

Whenever someone makes a pull request, it should be assigned to 
the maintainer for the module which is most affected by the 
request.  That person also has the primary review responsibility 
for the request, in the manner described by Steve, but may of 
course reassign it to someone else if necessary or appropriate.

For example, I would be happy to be the official maintainer of 
std.complex, std.path and std.process.  I feel a certain 
ownership towards those modules, and I very much want to review 
changes made to them.  Unfortunately, time currently does not 
permit me to scan the forums, Github and Bugzilla every day for 
discussions, pull requests and bug reports pertaining to these 
modules.  If, however, someone would assign them to me, I would 
be automatically notified via e-mail, and then I would definitely 
take the time to deal with it.

There will of course be requests that have a large impact on 
several modules, and there should also be someone that takes care 
of coordinating the reviews of these.

Lars