jfoust at threedee.com
Wed Aug 6 13:35:53 CDT 2008
At 12:25 PM 8/6/2008, Paul Koning wrote:
>>>>>> "John" == John Foust <jfoust at threedee.com> writes:
> John> At 11:13 AM 8/6/2008, Paul Koning wrote: I thought it was just
> John> tokenized because there were "de-compilers" that restored
> John> "compiled" code to real source.
> >> You may be thinking of RT-11 BASIC, where the "list" command
> >> actually did exactly that.
> John> No, I'm think of circa '83 VAX-based BASIC-Plus (2?).
>BP2 is a true compiler, not a P-code system.
As has happened many times before, the same topics get rehashed on
this list even by the same people. So Paul, we crossed this topic
about three years ago. :-)
There are two decompilers mentioned in this RSTS history:
1974 February-Clark Baker, George Robbins, Dan Grim and
Ed Baker write the PPCODE Basic Plus Decompiler.
1978: October-Nick de Smith returns to Dulwich College,
where he had been a student between 1967 and 1976, to
write the DECOMP Basic Plus Decompiler.
1990: July-SPL announces special anniversary offers on
various software products including The Link, BAS24K, BP3,
RPM, DECOMP, SORT1 and REPGEN.
(The name George Robbins caught my eye, as I knew him from
Commodore and the Amiga. Other googling makes it apparent
it's the same guy who worked on the 1974 decompiler,
and sadly, it seems he passed away in 2002.)
Perhaps we're just arguing about semantics. If BP2 was a true
compiler, to me that means it was emitting native CPU code in
an executable format - no interpretation needed. So you're
saying this decompiler was reading CPU opcodes and mapping
them correctly back to BASIC code? That it wasn't just
detokenizing something that ran through an interpreter?
It would probably be easier for us to find on the net the
BASIC source code for this decompiler I'm thinking of, than
for me to find my copy in my own archives. Then we could
study the source code.
The Spring '80 DECUS tape from Chicago RSTS SIG matches my
timeframe and locale, as well as the tape being submitted by
someone close to George's origin in Delaware:
"[80,19] BASIC-PLUS-1 DECOMPILER: two programs, one data file
BASIC-PLUS-1 reverse compiler"
Bill Gunshannon was looking for it in 5/07 on comp.sys.dec, too.
More information about the cctalk