13 Jan
2012
13 Jan
'12
2:40 a.m.
On 13 January 2012 02:10, Chuck Guzis <cclist at sydex.com> wrote:
This is because the *I/O routines* were Fortran's. They linked to Fortran
libraries for I/O, in short.
I'm pretty
sure, actually, that SNOBOL4 (the version of SNOBOL most
people are referring to when they say SNOBOL) was written in SIL.
That may be true, but when I used SNOBOL4, it was on a S/360 machine
and catastrophic error messages always came from the FORTRAN
routines.
And there were several versions of SNOBOL4. The one
the book
concerns itself with was 2.0.
And all of them were written in SIL and (on the '360!) used Fortran's I/O
routines. Note that, though: *just* the I/O routines.
If you can acquire the book on the implementation (it's "out there" in
electronic form) it's actually quite an interesting read. Griswold doesn't
disguise his missteps and misfeatures at all. If anything he's
occasionally too harsh on himself.
There was the 1964 SNOBOL and the 1966 SNOBOL3. Was
there ever a
SNOBOL2?
Not that was ever released, no. From what I gathered SNOBOL came out, then
was constantly incrementally modified into a glorious mess and that whole
process is kind of viewed as "SNOBOL2". When they decided on a ground-up
rewrite they called it SNOBOL3. Or something like that. Memory is hazy.
It might be fun to revisit SNOBOL after 40-odd years.
It was
something of a hog on a S360/40 back then. It'd probably run like
the wind on a low-end consumer PC today.
I have a cheap laptop here and can, according to CSNOBOL's timing utility,
pump about 2.5 million operations per second. I'm pretty sure a '360
couldn't do that? :D
--
"Perhaps people don't believe this, but throughout all of the discussions
of entering China our focus has really been what's best for the Chinese
people. It's not been about our revenue or profit or whatnot."
--Sergey Brin, demonstrating the emptiness of the "don't be evil" mantra.