Timothy Murphy writes:
 > And the discussion about the internal semantics of new pdfTeX primitives,
 > most of which would be far better dealt with
 > as \special's in a DVI file,
 > leads me to conclude that pdfTeX and Knuth TeX
 > are rapidly diverging.
why is this such a bad thing? Knuth wrote TeX and said "here you go,
peeps, do with it what you will, but just make sure that `TeX' remains
`TeX'". he never said "if you use my code you must follow all my
misguided decisions as if they were tablets of stone"

 > While Thanh's pdfTeX is a marvellous piece of work, which satisfied
 > an urgent need, in my view it is fundamentally misconceived, and
 > will in time be replaced by a version based on possibly extended
 > DVI.

replace "will" with "may conceivably", and I might agree. however, the
history of attempts to get any development going in the TeX world is
awesomely depressing. leaving aside the truly impressive work done on
macro packages, and the solid work on TeX distribution, what have we
had in the last 10 years? well, we have eTeX, but i have never met
anyone who could get excited about it (maybe I don't move in the right
circles), and we have the endless talk about NTS, but that appears to
be vapourous. then we have Omega, a marvellous thing which is let down by
the apparent fact that its developers cannot work on it regularly. and
lastly we have pdfTeX, which is the nearest thing to a runaway success
that I at least have seen in TeX world.

I think the world is voting with their feet, Tim

Sebastian