Without getting into the Risc/Cisc issues, let me add that it was a sheer pleasure to proram the PDP-10. It rewarded programmer skill and made for succinct, expressive code. I always appreciated the architecure of this system. Bob Mayans ________________________________________________________________________ Check out AOL.com today. Breaking news, video search, pictures, email and IM. All on demand. Always Free.
At III for several years we had a DEC-10 based compiler called TRIP, which exceeded in capability and usability all the C compilers then in existence. It was originally designed by Lowell Hawkinson, a LISP expert. I still miss it. Steve Gray ----- Original Message ----- From: <bobmayans@aol.com> To: <math-fun@mailman.xmission.com> Sent: Tuesday, June 06, 2006 3:21 PM Subject: [math-fun] order codes
Without getting into the Risc/Cisc issues, let me add that it was a sheer pleasure to proram the PDP-10. It rewarded programmer skill and made for succinct, expressive code. I always appreciated the architecure of this system.
Bob Mayans ________________________________________________________________________ Check out AOL.com today. Breaking news, video search, pictures, email and IM. All on demand. Always Free. _______________________________________________ math-fun mailing list math-fun@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/math-fun
Quoting Steve Gray <stevebg@adelphia.net>:
At III for several years we had a DEC-10 based compiler called TRIP, which exceeded in capability and usability all the C compilers then in existence. It was originally designed by Lowell Hawkinson, a LISP expert. I still miss it.
You don't say what it compiled; presumably "C" although your comment implies that it might have been C-like but not following it exactly. Did Lowell write it in LISP, or directly in machine language? He also had one of the many versions of LISP and one of the few that I know of which was capable of handling arrays directly as arrays. But I am not sure that it ever got widely used. The last I knew he was at SDC in California, although I have heard that he later formed a company of his own. - hvm ------------------------------------------------- www.correo.unam.mx UNAMonos Comunicándonos
The language was vaguely C-like but actually predated C. K&R and the famous Bell system technical journal issue both date from 1978. Trip was developed around 1971. Trip was an incremental compiler. You typed a line of code, it compiled and executed it. It was obviously influence by lisp, by people who wanted a lisp-like experience by also wanted something effecient. Among other advanced features it had structures, a pretty printer, and a permanantly resident symbol table. Ah the memories!
TRIP was first written in PDP-10 assembler although there was a later effort by Dave Dyer to rewrite it in itself. I don't know how far that got. TRIP had nothing to do with C. TRIP had its symbols in memory at all times. There was no distinction between compile and run time. It needed no separate debugger. You could redefine any or all routines at any time. Compilation was very fast, even with the old hardware. The 10's operating system had a feature whereby you could stop any run "on a dime" and save the complete program state onto disk and later resume operation exactly where it left off. TRIP had a few 10-specific features and was weak on data types, but these problems could have been fixed. It was used only at III and never exported to other machines. When I was later forced (due to no more PDP-10's being available) to work in C, it was a great comedown and a big step backwards. TRIP perhaps most resembles the Mathematica environment, without all the math and without lots of the other features of Mma. TRIP dates back to about 1967 when III was in Tech Square in the same building as Minsky's Project MAC. Quite a few people on this list were associated with us, including our host, Rich. Steve Gray ----- Original Message ----- From: <mcintosh@servidor.unam.mx> To: "math-fun" <math-fun@mailman.xmission.com> Cc: <mcintosh@servidor.unam.mx> Sent: Tuesday, June 06, 2006 5:29 PM Subject: Re: [math-fun] order codes
Quoting Steve Gray <stevebg@adelphia.net>:
At III for several years we had a DEC-10 based compiler called TRIP, which exceeded in capability and usability all the C compilers then in existence. It was originally designed by Lowell Hawkinson, a LISP expert. I still miss it.
You don't say what it compiled; presumably "C" although your comment implies that it might have been C-like but not following it exactly. Did Lowell write it in LISP, or directly in machine language? He also had one of the many versions of LISP and one of the few that I know of which was capable of handling arrays directly as arrays. But I am not sure that it ever got widely used. The last I knew he was at SDC in California, although I have heard that he later formed a company of his own.
- hvm
------------------------------------------------- www.correo.unam.mx UNAMonos Comunicándonos
_______________________________________________ math-fun mailing list math-fun@mailman.xmission.com http://mailman.xmission.com/cgi-bin/mailman/listinfo/math-fun
Quoting Steve Gray <stevebg@adelphia.net>:
..... When I was later forced (due to no more> PDP-10's being available) to work in C, it was a great comedown and a big step backwards. TRIP perhaps most resembles the Mathematica environment, without all the math and without lots of the other features of Mma. .....
I know that feeling. Some six years creating a LISP processor down the drain when IBM switched over from 7090 to 360. Nowadays I could have salvaged the code, even with something like Mathematica(TM), but back then I was barely beginning to write something resembling a symbol manipulation program, and nobody else really had anything either. C and FORTRAN weren't anywhere near perfect, but thay had (and still pretty much have) the great advantage of universality (not Turing's, just that everybody and everywhere, you could find one), especially if one avoided any and all "advanced" features. Is there anyone who doesn't rember (or have participated in) the flame wars, whether FORTRAN was better than Algol? Or whatsoever other pair? One of the great shortcomings of all these languages was the fact that you couldn't stop in the middle of something to compile and execute some source code that you had just created. Sounds like this TRIP might have done just that. - hvm ------------------------------------------------- www.correo.unam.mx UNAMonos Comunicándonos
Quoting bobmayans@aol.com:
..... let me add that it was a sheer pleasure to proram the PDP-10. It rewarded programmer skill and made for succinct, expressive code. I always appreciated the architecure of this system.
Although I never wrote machine language code for the PDP-10, my recollection is that this (or the PDP-6) was the first use of the hardware push-pop; probably the most significant change with respect to the IBM 709 or 7090, which I did program extensively in machine language. But my main recollection of the PDP-10 is twofold: first, their FORTRAN had COMPLEX data type, which is probably the only FORTRAN or anything else that was able to handle complex numbers directly via single symbols. And I {\em do} include C++ and Objective C in this comment. Secondly, and what was likely the most important feature from the user's point of view, was the timesharing operating system. Long before IBM had one. I don't know that either of these features had much to do with hardware, but the timesharing was certainly exemplary. And those DECtapes! It is interesting to speculate as to why the PDP-10 (later metamorphosed to PDP-20)never caught on better than it did, but that has been the fate or numerous products. It is painful to think that the PDP-10 gave Billy Gates his head start on microcomputers via the Intel 8080 cross assembler. - hvm ------------------------------------------------- www.correo.unam.mx UNAMonos Comunicándonos
--- mcintosh@servidor.unam.mx wrote:
... But my main recollection of the PDP-10 is twofold: first, their FORTRAN had COMPLEX data type, which is probably the only FORTRAN or anything else that was able to handle complex numbers directly via single symbols. And I {\em do} include C++ and Objective C in this comment. ...
Modern versions of FORTRAN do have COMPLEX data type. Gene __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com
participants (5)
-
bobmayans@aol.com -
Dave Dyer -
Eugene Salamin -
mcintosh@servidor.unam.mx -
Steve Gray