CHICKEN Scheme hacking sprint

On our last meeting at FrOSCon the idea of having another hacking sprint (aka CHICKEN weekend) has come up. I am willing to organize logistics for a meetup at Nuremberg, Germany. Since some foreign guests might want to do a little sightseeing as well and we are already busy in autumn with T-DOSE in Eindhoven, I propose one of the first three december weekends.

Location

U43 youth club Uhlandstraße 43, Nuremberg, Germany, Planet Earth 2011

Google Maps: http://g.co/maps/pv8vg

Date

09th - 11th Dec 2011 The sprint will start at Friday evening on the 9th and we will wind down sometime in the afternoon on Sunday.

Accomodation

Most CHICKEN Hackers plan to stay at the "Hotel ibis Nuernberg City" http://www.ibishotel.com/gb/hotel-1608-ibis-nuernberg-city/index.shtml

Located at: Koenigstorgraben 9 90402 - NUERNBERG GERMANY Tel.: (+49)911/24090 Fax.: (+49)911/2409413

Who is allowed to join?

This is an open invitation for anyone that wants to help the CHICKEN Scheme project improve.

What will be provided?

I will organize the venue and come up with a plan to get food and entertainment in time without the hassle to reinvent democracy. I might be able to come up with some interesting accommodation venues but that is still in a planning stage.

What will we be doing there?

Hacking on CHICKEN, socialising and having fun of course. An initial proposal for possible topics has been made by Felix (feel free to add your topic!):

Optimization and refactoring of the core libraries. Peter suggested this once (for a particular case in srfi-1.scm) and I'm sure there can be lots of improvements everywhere. Now, with specialization, it would be worthwhile to have unsafe entry points that typesafe code can specialize into.
Cleaning up of the mess in setup-api/setup-download/chicken-install.
Modularization and cleanup of the compiler - simplify knowledge transfer for core tools.
Should we distribute responsibility for core subsystems to other developers?
A COMPREHENSIVE test suite, probably (at least partially) machine- generated. That means tests for each and every library function.
A "battery included" distribution, sort of like the "Haskell Platform", with portable binaries for Windows, Mac and Linux (to make it easy to give it a try, to grab new users by the neck).
Distributed, turnkey salmonella runs ("Salmonella @ HOME" - Yuck!), or some sort of "testing probe".
Clarify perspective for future releases. Should there be some "focus"? What is the general opinion of R7RS? Should there be a roadmap?
Target users? Niche to be filled by CHICKEN? Or should this just be a universal hacking tool?
Think about the GUI situation.
The Book.
Refactor the SDL egg (and possibly some others).
Fully automate the release build.
Rewrite the scheduler from scratch.
Quicklisp like scheme interface to henrietta: see what eggs are available, their versions, compute dependencies etc.