GSoC project suggestions?

[Matt Mackall]

On Sun, 2011-01-30 at 11:03 +0100, Gilles Moris wrote:
> On Friday 28 January 2011 11:05:56 pm Matt Mackall wrote:
> > All of our projects last year (py3k, lightweight copies, history
> > punching) were way too hard.
> 
> I think I remember that lightweight copies and parent delta  were on a good 
> way (with respectively a patch queue and an option to activate in 1.7). May 
> be still missing a new wire protocol.

Yes, but the wire protocol is/was/will always be the hard part. In other
words, neither project got within 20% of being finished.

If I can't do it in a week, it's definitely not suitable as a GSoC
project. 

In my view, the goal of GSoc is to get reasonably competent beginners
involved with open source communities and usefully contributing code. 
It's more important that you learn how to get over the social hurdles
than the technical ones.

-- 
Mathematics is the supreme nostalgia of our time.