I'd like to participate in GSoC as a mentor. The projects I'm particularly interested in are:

- 3.4 Quick compilation - here I have a direct interest, because I'm really hitting the wall with the current compiler in my projects that involve sophisticated DSLs - it seems like space requirements for the compiler grow exponentially (or as a high order polinomial) with single function code size increase, and I get into situations where the compiler memory usage jumps to multiple gigabytes. I would slightly rephrase the name of the goal as: tunable compilation, i.e. make the compiler behave differently depending on the speed and compilation-speed declarations

- Implementing coroutine support - some brief remarks may be found here: http://orthecreedence.github.io/cl-async/2012/11/07/missing-coroutines-in-common-lisp.html. It is, I think, at the same time a promising and complicated subject, which should, obviously, be discussed in more detail. The interesting part is that, as far as I understand, a huge chunk of code for implementing coroutines can be reused from the condition system implementation. In fact, I've experimented with implementing coroutines on top of the signal protocol, and there's just one piece missing to be able to do it

At the same time, I don't have a lot of practical experience with the SBCL codebase: I neer made any contributions, and only look at the code from times to times. So, I can bring to the table only my overall Lisp and system design experience, a demanding real-world environment to test some of the use cases, and a vision on how the things should work ideally. Given this I don't know if I can be a helpful mentor...

Any comments on the above topics welcome,


Vsevolod Dyomkin
skype, twitter: vseloved