Re: Porting wptserve handlers to Python 3: next steps / code review

I've got some experience porting the eventsource handlers (although I
haven't submitted that PR yet due to a couple unfinished tricky handlers),
so I can review PRs as well.

Cheers,
Josh

On Tue, 12 May 2020 at 13:03, Stephen Mcgruer <smcgruer@chromium.org> wrote:

> Thanks for sending this email Robert; I'm excited to see us keep the ball
> rolling on Python 3 support.
>
> > In addition, we'd really appreciate a few more people to sign up for
> reviewing these changes to share the workload. Anyone volunteering?
>
> I'm happy to review PRs, albeit with no specific prior knowledge. +foolip,
> who volunteered to review as well.
>
> That'd bring us to 5 reviewers assuming jgraham and annevk are willing to
> review; do you think that is enough Robert?
>
> On Mon, 11 May 2020 at 18:19, Robert Ma <robertma@chromium.org> wrote:
>
>> Hello everyone,
>>
>> We're getting close to finalizing the plan for migrating close to 500
>> wptserve handlers we have in WPT. Now we have a few concrete steps to take:
>>
>> 1. Regarding the trial PR
>> <https://github.com/web-platform-tests/wpt/pull/23363>, James, Anne and
>> others who'd like to take a look, do you have any other comments on this
>> PR, especially high-level ones about the general approach? This would
>> unblock the following steps and we can address small issues in parallel.
>> 2. If we agree this approach is what we wanted by having consistent and
>> explicit semantics across Python 2 and 3, I'll update the RFC
>> <https://github.com/web-platform-tests/rfcs/pull/49> (essentially
>> swapping the currently "recommended" and "alternative" approaches and
>> filling in some more concrete guidelines), and kick off a new round of RFC
>> process (hopefully relatively quick since many people are already on board
>> with the new approach).
>> 3. Meanwhile, Ziran can start porting more handlers (we can wait until
>> the RFC is accepted to actually merge the PRs). We have hundreds of
>> handlers and we should expect lots of PRs. Reviewing them is a critical
>> task, too. Since we now have concrete guidelines and changes will be
>> largely mechanical, I'm proposing to adopt the "LGTM % nits" convention
>> widely used in Chromium: if a PR largely looks good but has some minor
>> issues, approve the PR with comments. In addition, we'd really appreciate a
>> few more people to sign up for reviewing these changes to share the
>> workload. Anyone volunteering?
>>
>> Best,
>> Robert
>>
>