Article: 113095 of alt.folklore.computers Path: iad-read.news.verio.net!dfw-artgen!iad-peer.news.verio.net!news.verio.net!news.maxwell.syr.edu!newsfeed-east.nntpserver.com!nntpserver.com!chi1.webusenet.com!news.webusenet.com!cyclone1.gnilink.net!spamkiller.gnilink.net!nwrddc02.gnilink.net.POSTED!53ab2750!not-for-mail Sender: cstacy@BONK Newsgroups: alt.folklore.computers Subject: Re: LISP was: Eliza References: <3e4bc4d4$1@news.ucsc.edu> <3E4EC64B.6736A2DA@gazonk.del> <8E05D08E573CFC64.E9921937DDEB8011.CD841B2EEFB5B0D2@lp.airnews.net> <0071C11D420D5A18.46554D7FD33BD420.7C23C3D57F0D9773@lp.airnews.net> From: cstacy@dtpq.com (Christopher C. Stacy) Message-ID: Lines: 70 User-Agent: Gnus/5.09 (Gnus v5.9.0) Emacs/21.2 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Fri, 21 Feb 2003 22:02:38 GMT NNTP-Posting-Host: 68.160.161.205 X-Complaints-To: abuse@verizon.net X-Trace: nwrddc02.gnilink.net 1045864958 68.160.161.205 (Fri, 21 Feb 2003 17:02:38 EST) NNTP-Posting-Date: Fri, 21 Feb 2003 17:02:38 EST Xref: dfw-artgen alt.folklore.computers:113095 >>>>> On Fri, 21 Feb 2003 08:06:00 -0500, Bernie Cosell ("Bernie") writes: Chris>> Would it surprise you to know that the PDP-10 was designed for Lisp? Bernie>I thought it was the PDP-6, and the PDP-10 was just the next-generation You are correct, but I was addressing (so to speak) folks who I thought were not familiar with the PDP-6. When I was at the AI Lab decades ago, I had learned from Greenblatt that the convenience of the PDP-10 word/halfword architecture for Lisp was no mere fortunate accident. More recently, Alan Kotok, who worked with Bell designing the PDP-6 and PDP-10 has stated, "Facilitating a good LISP implementation was an important consideration." Bernie> And in a similar vein, wasn't the terminology for LISP derived from the Bernie> 7094 -- 'car' was "contents of the address register" while "cdr" was Bernie> contents of the decrement register", or something like that? It was the IBM 704. Some of the looping instructions on that machine worked by accessing two 15 bit register fields (in instruction words) called "Address" and "Decrement". You could load and store these fields into the index registers. A CONS cell is just a pair of pointers, and it's convenient if you can load and store them together at once. So they took advantage of the hardware and implemented CONS cells as a pair of 15 bit pointers. When he was implementing the first Lisp in 1959 at MIT, Steve Russell wasn't feeling inspired to come up with a nice abstract name, and so the operators got named CAR for "Contents of the Address part of Register") and CDR for "Contents of Decrement part of Register"). The other 6 bits in the words were used as flags for the garbage collector, by the way. Here's a fun link I ran across that mentions Steve Russell, Alan Kotok, and Spacewar: Ironically, some programmers consider CAR and CDR to be very abstract names, as they refer to the parts of a data structure called a CONS cell and mean nothing else; certainly nothing to bit fields on some ancient IBM computer! Some other programmers prefer the alternate function names that Lisp provides that mean the same thing: FIRST and REST. (McCarthy and company actually wanted to change the names to FIRST and REST right away after they tried teaching Lisp, but CAR and CDR somehow stuck.) Personally, I write FIRST and REST when I am operating on a chain of CONS cells linked together to form a list, and write CAR and CDR when talking about a CONS cell in isolation (usually some kind of leaf node, as in an "association list"). CONS cells are the oldest data type in Lisp, along with symbols, and 44 years ago Lisp was used primarily for writing programs that dealt with lists of symbols in an Artificial Intelliegence domain. Programmers of today's impovershed languages might be surprised at how useful lists and symbols are in all kinds of programs. Especially when writing macros. But it should be noted that modern Lisp (for more than 20 years) includes many additional fundamental data types, such as all kinds of numbers (double-floats, complex, infinite precision, ratios), byte fields like on the PDP-10, strings, characters, arrays, vectors, hash tables, streams, pathnames, and of course user-defined structures. It also has first-class functions, closures, and all the control structure you might want. It also has typing and object-oriented features that are a lot more sophisticated than what you find in languages like C++ or Java. And lots more. Lisp has too many goodies to even list in a short message like this, combines them together wonderfully, it's had all this stuff for longer than those other languages have existed. (Not trying to start a language war, here, just advertising that Lisp is not at all like what most people think it is. It _was_ sort of like they think, up until about 1970.) Earlier you said something about your mother? Article: 19509 of alt.sys.pdp10 Path: iad-read.news.verio.net!dfw-artgen!iad-peer.news.verio.net!news.verio.net!news.maxwell.syr.edu!newsfeed1.cidera.com!Cidera!cyclone2.gnilink.net!cyclone1.gnilink.net!spamkiller.gnilink.net!nwrddc02.gnilink.net.POSTED!53ab2750!not-for-mail Sender: cstacy@BONK Newsgroups: alt.sys.pdp10 Subject: Re: PDP-10 Macsyma References: From: cstacy@dtpq.com (Christopher C. Stacy) Message-ID: Lines: 80 User-Agent: Gnus/5.09 (Gnus v5.9.0) Emacs/21.2 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Mon, 14 Apr 2003 04:33:11 GMT NNTP-Posting-Host: 151.203.61.59 X-Complaints-To: abuse@verizon.net X-Trace: nwrddc02.gnilink.net 1050294791 151.203.61.59 (Mon, 14 Apr 2003 00:33:11 EDT) NNTP-Posting-Date: Mon, 14 Apr 2003 00:33:11 EDT Xref: dfw-artgen alt.sys.pdp10:19509 >>>>> On Sun, 13 Apr 2003 13:36:24 -0700, Mark Crispin ("Mark") writes: Mark> On Sat, 12 Apr 2003, John Saeger wrote: >> Does anybody know if runnable PDP-10 Macsyma is available anywhere? And >> maybe the source code? Mark> The only known runnable PDP-10 Macsyma is the TOPS-20 port of Macsyma 304. Mark> Unfortunately, it has a 1983 Symbolics copyright and trade secret notice Mark> for "enhancements", meaning that it can't be distributed without the Mark> permission of the current owner of the assets of Macsyma, Inc. Mark> The ITS archive had a copy of MIT (not Symbolics) Macsyma Mark> source, but I don't know if that is available or what needs to Mark> be done to whip it into shape. If anyone has a copy of those archives, I would suggest being very careful about the legal implications of releasing them. MIT sold Symbolics the exclusive rights to that very code. Afterwards, the U.S. Department of Energy demanded rights to a derivative work of that very same code. Various legal battles ensued to resolve that predicament, and the Macsyma that was eventually made available under license from DOE was _not_ the MACLISP version. It was a subset that had been ported to NIL, a long dead dialect of Lisp for the VAX. The "Maxima" program is yet a third derivative of that, whose rights to GPL are from an agreement with DOE, deriving from DOE's license with MIT. The version of Macsyma from Symbolics was eventually rewritten by them into Common Lisp. This is not, however, the basis of the "Maxima" program, which was an independant effort operating on a different fork of the code. Also, the ITS version doubtless includes various works that are owned by private individual contributors (rather than MIT or Symbolics, and certainly not DOE). One should be careful to identify the proper owner of each file (which could possibly conflict with what the copyright notices say). Just because MIT or somebody perhaps put an incorrect copyright notice on something doesn't relieve any later third parties who want to copy those files. Also, be aware that the name "Macsyma" is trademarked by Symbolics (whose assets have been bought by various successor companies, whose company names include the "Macsyma" or "Symbolics"). I did a little work on it about 9 years ago during the period it was owned by "Macsyma, Inc.". To use the MACLISP version of Macsyma, one should seek to secure a legal agreement license from the current incarnation of Symbolics (which probably still owns the exclusive license to that code), and possibly also a second legal agreement from MIT, which has entered into licensing contracts with other third parties (ie. DOE). All that being said, I think it would be really cool if the original MACSYMA that ran in MACLISP on ITS and TOPS-20 was available for the PDP-10 emulators. Perhaps someone will sort out all the legalities. However, the GPL'd version of MACSYMA, aka MAXIMA, is better than the original program on the PDP-10, if only for its modern graphics. MAXIMA is of course written in Common Lisp, not MACLISP. If you just want a really good version of Macsyma, not having to do with PDP-10s, you can currently purchase a copy of "Macsyma" from Symbolics. It started as the original MACLISP version, then was ported to Zetalisp (the Lisp Machine), then to ANSI Common Lisp. Macsyma runs on Windows, and has lots more features than "Maxima", and represents many years of additional commercial development. I think the most recent incarnation of Symbolics is called "New Symbolics Technologies" or something. They are based in Northern Virginia. I could dig up the contact information. ObHistory: The final product from Symbolics was a Lisp Machine called "Open Genera" that ran in software: a complete emulation of the Ivory processor chip that powered the workstations. The emulator only runs on DEC Alphas (that being the fastest thing around at the time, and the only 64-bit platform, anyway.) This product is also still available, but you don't need it to run Macsyma. Their PC version of Macsyma is much better, having had extensive development post-LispM era. Article: 20199 of alt.sys.pdp10 Path: iad-read.news.verio.net!dfw-artgen!dfw-peer!news.verio.net!newsfeeds.sol.net!hammer.uoregon.edu!arclight.uoregon.edu!logbridge.uoregon.edu!news.u.washington.edu!140.142.17.34.MISMATCH!news.u.washington.edu!Ikkoku-Kan.Panda.COM!mrc From: Mark Crispin Newsgroups: alt.sys.pdp10 Subject: Re: Any online copies of INTERLISP and LISPUSERS for TOPS-20? Date: Fri, 30 May 2003 09:03:46 -0700 Organization: University of Washington Lines: 13 Sender: mrc@Ikkoku-Kan.Panda.COM Message-ID: References: <6b26c83b.0305300147.1bc02550@posting.google.com> NNTP-Posting-Host: nntp4.u.washington.edu Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Trace: nntp1.u.washington.edu 1054310642 23110 (None) 140.142.17.38 X-Complaints-To: help@cac.washington.edu In-Reply-To: <6b26c83b.0305300147.1bc02550@posting.google.com> Xref: dfw-artgen alt.sys.pdp10:20199 On Fri, 30 May 2003, Derek Peschel wrote: > Although twenex.org has the version from 31-Dec-84, some of the files > are missing. I recall trying to run one of the fancy full-screen editors > and I recall one of the missing files being SARITH.COM but I'll have > to check again. What twenex.org runs is what I saved from Stanford. I don't know of any other online copies. -- Mark -- http://staff.washington.edu/mrc Science does not emerge from voting, party politics, or public debate.