|Well! I’ve a collection of link those could be usefull for start and go inside hurd matter.
I think that currently the main stream development is to have and release a stable hurd running on GNU Mach. As you see can in last link there are a number of problem in porting hurd to L4, actually Fresco (an L4 branch). L4 is released in GPL, My opinion is that porting the Hurd to L4 is good because it’s really a Microkernel (even in term of space), which could make a good advance when porting OS to limited devices, also it already exists port of L4 to other platform.
First of all follow the link, read documentation (tecnical and even user’s), subscribe debian-hurd ml for solving preliminary doubt. Next, if you are still interested, install the Hurd and start to play with user interface (translator, et all) …. ask for what is needed and start to code!
Following this iter is for do not have a repulsive reaction to Hurd concepts to which you could not be familiar.
Hurd on L4 is good enought: IBM funded a project called Sawmill, this is an OS very similar to the Hurd on L4 both in respect of concept and interface. Sawmill was developed outside IBM, many hackers have the hope to see the source released with GPL terms.
Unfortunally I can’t contribuite to the Hurd, but I’m happy to know there are people interested in.
Emacs, gcc run, networking is ok at most (I don’t know if ppp is available). Xfree should run. The greatest missing part should still be POSIX thread support, thus things like Gnome couldn’t run or run without threading. From a user point of view the Hurd currently is very slow, orribly slow