Signs on the Sand

I've got an invitation from Mono guys to consider contributing to Mono Managed XSLT Processor implemenation (now they have only a wrapper around libxslt engine). Well, we at Multiconn have designed and built one XSLT processor couple of years ago (it's quite specialized and highly (may be even too) optimized XSLT 1.0 engine for mainframe OLTP environment, written in ANSI C). It was great opportunity to learn XPath and XSLT thoroughly from cover to cover and it was actually cool (apart from programming in ANSI C nowadays). I definitely have some ideas about XSLT implementation, especially if you have C#, .NET plumbing available and considering rumors around next Microsoft .NET XSLT impl :). So it's really awfully tempting.
But it's quite big open source project and I'm too busy unfortunately... I have long TODO list of not-less-interesing (but smaller) projects, I've started another article, I'm in MCAD cert preparation curve after all. And day work of course - we've just finished one trivial Web Service project, but another BizTalk-related one is on the horizon already. So I'm not sure...