GSoC proposal: Porting mercurial for py3k

[Dirkjan Ochtman]

On Fri, Apr 9, 2010 at 08:38, Nicolas Dumazet <nicdumz at gmail.com> wrote:
>> The first sketch of my proposal involves, obviously, using 2to3,
>> checking what fails and then iterativelly solving the problems until I
>> have a basic working version. Then work on the specifics of python 3
>> would begin. Yes, it is a very rough outline right now, but I can make
>> it better if you consider this as a valid approach.
>
> I'm afraid you will have to expand more. As you might expect it,
> porting a complex application to py3k *is* _very_ difficult. The fact
> that it has not been done yet particularly emphasizes this issue:
> throwing in 2to3 and "work on the specifics of python 3" is not going
> to be enough. It's not something that can be tackled without a lot of
> communication and an in-depth knowledge of Mercurial code base.
>
> Convince us that you know what you're talking of!

Indeed. Also check with the work that was done last year as part of
the Summer of Code.

Also, I think working full time on the SoC is more or less something
that we require; not a nice-to-have.

Cheers,

Dirkjan