I believe the most recent version of Henry's son-of-RFC1036 draft is
zoo.toronto.edu:pub/news.txt.Z -- there's also a postscript version.
I really need to update it somewhat based on recent comments... I've
got a bunch of good comments (including some from John) that I haven't
had a chance to look at and think about properly yet.
For the record, I think that document's comments about See-Also
vs. References and about message/news are way off the mark.
My current take on the former is that I'm right if you believe 822 and
John's right if you look at actual practice, although I need to take a
second look at it all. As for the latter, it's a deeper issue than I
first thought and I don't want to comment until I've had a chance to
sort it all out, but John has managed to raise some doubts in my mind.