Yes, Martin Richards BCPL has been nice since the 60s. It was used as 
an implementation language at Lincoln Labs on the TX-2 (via Jim 
Curry) and then Jim put it on the Novas and then the Altos at PARC. 
One of the operating systems (adapted from Strachey) and BRAVO (the 
first version of  MS Word) were implemented in BCPL. It also was the 
precursor to C at Bell Labs (and was nicer in my opinion).

The BCPL compiler was written in BCPL and translated into bytecodes. 
To port BCPL one wrote a simple interpreter to get the compiler 
going, and then one could produce optimized code for the new target 
machine, etc.

But check out the hardware of the B5000 and Barton's paper on it ca. 
1961 to see where byte-coded interpretive VMs came from. Also look at 
the KDF9, etc. Several of these systems influenced Richards (I'm glad 
to hear that he is well).

Cheers,

Alan

At 6:25 AM -0500 12/16/03, David T. Lewis wrote:
>On Tue, Dec 16, 2003 at 05:28:55PM +1300, Richard A. O'Keefe wrote:
>>  There are many books and papers about VMs.
>>  For example, the book about BCPL by Richards and Whitby-Strevens
>>  describes the OCODE vm for BCPL.  Ultimately it all goes back to
>>  Dijkstra's Algol 60 compiler, one of the first Algol 60 compilers.
>
>Funny that you should mention this. I expect to be speaking with
>Martin Richards (by phone) for the first time a couple of hours
>from now, and the BCPL book happens to be sitting on the desk next
>to me right now. Richards' BCPL system and Cintpos runtime are
>quite interesting by the way. Transparent all the way down, just
>like Squeak.
>
>Dave


--