This should be integrated into contributing.
Every now and then, people show up who have an inward urge to contribute to the GNU Hurd, but have some difficulties about how to do that.
For example, IRC, #hurd, 2010-10-06:
<rah> I find it difficult to find the will to contribute to the hurd while hurd != hurd-ng <pochu> hurd-ng? <pochu> ah, http://www.gnu.org/software/hurd/hurd/ng.html <pochu> rah: you may want to work on achieving that then <rah> pochu: I'm not in a position to do OS research <antrik> rah: if you are not into OS research, why do you need it to be ngHurd? :-) <rah> antrik: I don't want to work on software which I know is already obsolete <tschwinge> rah: My position on that can be found here; you may want to think about it. http://lists.gnu.org/archive/html/bug-hurd/2007-07/msg00111.html <antrik> rah: the existing Hurd implementation is not any more obsolete than any other large software project <antrik> there are always things that could be redone in a better way some time in the future <antrik> but we have to start somewhere <antrik> software development is a dynamic process <antrik> trying to come up with a perfect design before you write any code will never lead anywhere, ever <rah> antrik: of course, but when you know your start is wrong, have identified its problems, and are in the process of designing a second attempt, working on the first seems pointless <antrik> rah: well, do you know all these things? because I do not <antrik> what the experiments with new Hurd designs proved so far is that nobody is in a position to claim, "I have a better design" <antrik> it's not hard to come up with a design that is better in some points -- but it's damn hard to come up with one that's not lacking in others <antrik> the existing Hurd design is actually the only one which we *know* to work <antrik> while research on improving the design is certainly beneficial, it's not like there is something new ready to replace the existing design at any moment <antrik> and frankly, I'm more and more convinced that only iterative changes can ever result in any real improvement <antrik> (and doing these changes requires a certain momentum, which we will never gain unless we actually have something usable first) <LarstiQ> rah: afaik, not much is being done of designing another attempt <rah> antrik: yes, I know all these things