amit@umn-cs.UUCP (Neta Amit) (10/14/87)
I have a question for the active members of this newsgroup (Lee Dickey?), that has probably been discussed a thousand times. Post your response on this newsgroup, or mail directly to me. Here goes: I'm looking for a PD (unlikely) or commercial (several hundred $) APL product for the AT, under either MS-DOS or any form of Unix, that will have the following similarities to IBM's APL2: 1. Will be able to apply system-defined operators to user-defined functions 2. Will allow user-defined operators (and obviously to apply them to any fun) 3. Has the Bracket operator, which I find indispensible! 4. Supports nested arrays -- but this is really of less importance, except in the context of (3). - I checked with STSC; their PC product doesn't support anything of the above. Their Unix product supports just nested arrays, which are not terribly important to me. - Being also a Lisp (Scheme) programmer, Q'Nial may be appealing. If its cost (for Univ P.O.) is not steep, I'd consider it as a second package to get -- but not as the first: I prefer the standard APL notation and keyboard. I have their address in Kingston; do you have their phone #, by any chance? - As far as I can remember, MIPS Software Development, of Michigan, marketed DYALOG APL from Dyadic Systems of UK. I have no idea if they have a PC product (DOS/Xenix), and if it satisfies my requirements. - Same question about I.P. Sharp; here I have neither their address nor their phone#... - IBM. Do they have their own APL2 product for the AT? Is it complete, i.e. the same as their mainframe APL2? Who should I contact about this? - Any other implementation?? Any info about features and limitations, performance, support, stability of the product, hardware and software requirements, (discounted?) prices, addresses and telephone numbers will be greatly appreciated. Better yet, I'll be most interested in your personal experience. Thanx in advance, -- Neta Amit U of Minnesota CSci Arpanet: amit@umn-cs.cs.umn.edu
ljdickey@water.UUCP (10/17/87)
In article <2356@umn-cs.UUCP> amit@umn-cs.UUCP (Neta Amit) writes: > ... > I'm looking for a PD (unlikely) or commercial (several hundred $) APL > product for the AT, under either MS-DOS or any form of Unix, that will > have the following similarities to IBM's APL2: > > 1. Will be able to apply system-defined operators to user-defined functions > 2. Will allow user-defined operators (and obviously to apply them to any fun) > 3. Has the Bracket operator, which I find indispensible! > 4. Supports nested arrays -- but this is really of less importance, except > in the context of (3). > ... > ... I have their [Nial's] address in Kingston; do you have their phone #, by > any chance? > Try Nial Systems, Kingston, Ontario: 1-613-549-1432 > - As far as I can remember, MIPS Software Development, of Michigan, > marketed DYALOG APL from Dyadic Systems of UK. I have no idea if they have > a PC product (DOS/Xenix), and if it satisfies my requirements. I don't know either. I do not have their address with me. The company is in a suburb of Detroit, I think. > - IBM. Do they have their own APL2 product for the AT? Is it complete, > i.e. the same as their mainframe APL2? Who should I contact about this? So many questions, I can not answer them all. From your shopping list, it sounds like you are already using an advanced product. You will find no PD product like this. Sharp APL has, by far, the most complete APL for the PC. From them, you get the *full* mainframe implementation, complete with an emulator for the 370 instruction set, all for a micro pricetag. They have even emulated the "large-number" arithmetic anomalies [bugs?] that were present in the 360 and perpetrated in later products of that line. The idea is that results on the micro should be exactly the same as the results on the production system. For a while, they gave it away for $100, but I think they charge something more for it now. People did not believe that it was the complete system for that price. I think it is a wonderful bargain. I.P.Sharp Associates keeps offices in a lot of major cities throughout the world, so it is possible that there is an office of near you, but if not, call them in Toronto at their main number, 1-(416)-364-5361. -- L. J. Dickey, Faculty of Mathematics, University of Waterloo. ljdickey@watmath.UUCP UUCP: ...!uunet!watmath!ljdickey ljdickey%water@waterloo.edu ljdickey@watdcs.BITNET ljdickey%water%waterloo.csnet@csnet-relay.ARPA
ljdickey@water.UUCP (10/19/87)
>In article <2356@umn-cs.UUCP> amit@umn-cs.UUCP (Neta Amit) writes: > ... > - As far as I can remember, MIPS Software Development, of Michigan, > marketed DYALOG APL from Dyadic Systems of UK. I have no idea if they have > a PC product (DOS/Xenix), and if it satisfies my requirements. > ... MIPS Software Development, Inc should be reachable at 1-313-855-3552 -- L. J. Dickey, Faculty of Mathematics, University of Waterloo. ljdickey@watmath.UUCP UUCP: ...!uunet!watmath!ljdickey ljdickey%water@waterloo.edu ljdickey@watdcs.BITNET ljdickey%water%waterloo.csnet@csnet-relay.ARPA
moran@yale-zoo-suned..arpa (William Moran) (10/19/87)
I'm almost 100% sure that IBM has no APL2 for a PC/AT. The reason I say this is
that when I was working there, I got hooked on APL2, and would have killed for
a PC version of it. I did a fair amount of investigation, and the responses I
got were of the sort: "APL2 for the AT? Ha Ha Ha Ha Ha Ha Ha Ha Ha Ha...etc.
you must be kidding". My general experience is that people writing these sorts
of things don't think that anything less than a 3081 is a real machine. It's
just luck that these things run on a 4341.
Bill Moran
******************************** William Moran Jr. *************************
* Arpa: moran-william@yale.arpa or cs.yale.edu * Arioch, Arioch etc. *
* UUCP: {harvard,decvax,cmcl2}!yale!moran-william * 1. Nf3 ++/-- *
* Bitnet: moran-william@yalecs.bitnet * *
****************************************************************************Richard_s_Stoneston@cup.portal.com (10/19/87)
Okay, for PC-AT and other MS PC/dos people and computers there are several APL's available. I know of no PD APL's available for anything- someone may enlighten me on this. I do know of a project going on in England to come out with a PD Apl for the XT/AT machines but do not know to what extent the language will be implemented. My guess it will be a stripdown but if the source is released people may want to extend the parser to include more things. Commercial APL's for PC/XT/AT types include, from STSC, Pocket APL ($100), and STSC APL/PC ($$$ not sure, but expensive), IBM-APL, and perhaps ibm ver 2 of apl. IBM-APL REQUIRES 8087 + CGA. As far as I know, pocket APL will run on an unmodified XT/AT with either CGA or EGA hardware (no character generator required) with or without an 8087. I also believe the same is true for the full-fledged professional STSC APL PLUS/PC, which is expensive... and probably can use either 8087 emulations or 8087 coprocessor, depending on what is in the computer.
amit@umn-cs.UUCP (Neta Amit) (10/20/87)
I know of no PD APL for a PC-AT. There is at least one free APL, in the contributed software in the Unix 4.[23]bsd distribution tape. The version I have access to is a rather old one, and not particularly impressive. Nevertheless, it is free, and all the sources are avilable. I don't know to what extent it is system dependent. I also don't know if there are newer versions. It was developed in Purdue-EE in the early 80's (at best...). Will someone from Purdue-EE shed some light?... As for AT implementations, I'm also in the market. I called STSC and IP Sharp. IP Sharp has a supported product, which they (and Lee Dickey -- thanx Lee) claim is very compatible with their mainframe implementation. They also say it is rather slow in execution, and they are not active in marketing it. In fact, they recommend the STSC product... You may be interested in the product after all, because it supports nested arrays. The list price is $395. They have quantity and educational discounts (about 50% off). STSC appears to have a much more active operation in this market. Their MS-DOS product (the full fledged, not the pocket version) is a solid implementation (so they say), with full support for graphics, statistical package and what not. It is reasonably fast. The current version (6.x) will be replaced soon by version 7, as will the price tag ($595 going to $650 -- or was it $695?). They also have educational discounts ($295 going to $350). Nested arrays are not supported. They ARE supported on the AT-Xenix implementation (bigger bucks: $500 ed-discount). I have no details about the IBM implementation(s), and would love to hear about it. -- Neta Amit U of Minnesota CSci Arpanet: amit@umn-cs.cs.umn.edu
ken@braegen.UUCP (10/26/87)
Sharp's pc product is indeed compatible with their mainframe version. However, they achieved this compatibility by writing an emulator of the mainframe environment. This makes the APL about the slowest available language for the PC environment. Also note that while it has nested arrays it does not have the other extensions that make APL2 so interesting. In the PC world with the higher end processors expect STSC to produce a DOS nested array version within the next year or so. -- Ken Marchant The Braegen Group, Toronto, Ontario (allegra,linus,ihnp4,decvax)!mnetor!yetti!geac!braegen!ken
dietz@zhmti.UUCP (10/28/87)
In article <2410@umn-cs.UUCP> amit writes: >There is at least one free APL, in the contributed software in the Unix >4.[23]bsd distribution tape. The version I have access to is a rather old >one, and not particularly impressive. Nevertheless, it is free, and all >the sources are avilable. I don't know to what extent it is system >dependent. I also don't know if there are newer versions. It was >developed in Purdue-EE in the early 80's (at best...). Will someone from >Purdue-EE shed some light?... You're right on most counts. The original author was Ken Thompson at Bell, Purdue inherited it. See the enclosed README from the release. BTW, There are mods to the BSD release tape version that fix some null pointer derefs and that get the floating point to do the Right Things. (w.r.t. IEEE model, 68k's) It is free. It is not that impressive. It does work. It's probably not really worth hacking. Ken Thompson wrote it before stdio existed, so the comment about "rather old" in the quote is an understatement (yes, Virginia, there are sources that are predate stdio), It plays all sorts of games and tricks with IO, although Purdue seems to have hacked that for some of their own terminal support. Don't touch any of the stdio routines either (so that ld brings these in) or we conflict. (i.e. they don't work well together) Most of the work after Bruner was done by Ken Yap (ken@cs.rochester.edu) (Hi Ken!) with some further hacking by us here at zhmti (not much)... Ken and I never have gotten around to giving back the patches to Berzerkeley.... When I contacted John at LLL, he refered me to Ken, since he had not touched the stuff in quite some time... -----Included File: ~src/apl/README---------------------------------------- Title: APL Authors: John D. Bruner Lawrence Livermore Laboratory P.O. Box 808, L-276 Livermore, CA 94550 (415) 422-0758 Prof. Anthony P. Reeves Cornell University, Phillips Hall Ithaca, NY 14853 (607) 256-4296 Description: This is Purdue/EE's APL, which runs on both PDP-11's and VAX-11/780's. This APL originally was written by Ken Thompson at Bell. It went to Yale for a while, and came to Purdue via a Chicago distribution in (I think) 1976. Jim Besemer (now with Tektronix in Oregon) made many of the extensions to the original V6 PDP-11 version, including quad I/O functions, the state indicator, internal label processing, and a number of primitive functions. I began support of APL when Jim left in 1978 and have been handling it since then. The driving force behind all of the development and maintenance of APL at Purdue has been my major professor, Dr. Anthony P. Reeves. Please forward bugs/comments/suggestions to Dr. Reeves or to me (UUCP site "pur-ee", login names "reeves" and "bruner"). [...install notes deleted -- dietz] -----End of Included File-------------------------------------------------- Dieter H. Zebbedies ('dee-ter ayech 'zeb-ed-eez) Zebb-Hoff Mach. Tool's Automated Manufacturing Project Cleveland, OH (USnail): 9535 Clinton Rd, Cleveland, OH 44144 (+216 631 6100) (+216 741-5994) (UUCP): ...{decvax,sun,cbosgd}!mandrill!zhmti!dieter (CSNET/ARPA/BITNET): dieter@mandrill.CWRU.EDU
tower@bu-cs.BU.EDU (Leonard H. Tower Jr.) (11/14/87)
X-Home: 36 Porter Street, Somerville, MA 02143, USA +1 (617) 623-7739
Just notice this seminar announcement from California. At least,
helpful to the Unix-PC folks. Maybe they'll do a PC/AT port soon.
This is all I know about it, but I expect both companies are is
Silicon Valley, so try telephone information out there
(408-555-1212,415-555-1212,etc.).
enjoy -len
IBM Almaden Research Center
650 Harry Road
San Jose, CA 95120-6099
RESEARCH CALENDAR
November 9-13, 1987
...
SAX: SHARP APL UNDER UNIX
J. K. Tuttle, I. P. Sharp Associates, and P. C. Berry, Digital Insight
APL Int. Tech. Liaison Mtg. Tues., Nov. 10 1:00 P.M. Room: Front Aud.
Sharp APL/UX (informally called SAX) is a new implementation of APL
based on Iverson's "A Dictionary of the APL Language," recently
published as Volume 18, No. 1 of the ACM Publication "APL Quote Quad."
The thrust of Iverson's recent work has been simplification and
generalization. Only a few new symbols have been added to the
language; mostly, they denote what he now calls adverbs or
conjunctions: that is, symbols that do not themselves denote actions,
but modify the way in which other actions work. The concept of the
automatic extension of scalar functions has been generalized to the
complementary notions of "call" and "frame," permitting a function
(whether primitive or user-defined) to be applied in parallel fashion
throughout the cells contained in a framework of data. The
partitioning of data into frame and cells is inherently neither in the
data array nor in the verbs applied to it, but is governed by the
conjunction "rank" which specifies how a verb treats its arguments.
To permit the parallel cell-wise application of several functions, the
language provides three conjunctions that form compositions. Where
parallel execution produces results of inconsistent shape, the
system's tolerant frame-builder provides for automatic adjustment to a
rectangular frame. APL/UX supports three types of simple data:
character, numeric and "boxed" (otherwise called "nested" or
"enclosed"), as well as mixed arrays containing any of these. The
approach to boxed arrays differs somewhat from that taken in APL2, and
does not require (or provide) operators to govern the application of
functions to nested arrays. Iverson's Dictionary includes the
definitions of a selection verb and a merge conjunction which, taken
together, replace the old notations for indexing and indexed
assignment. However, these are not yet supported in SAX. SAX is
written in C, and runs under Unix(TM). Versions are now available for
the Sun workstation and the AT&T Unix PC. They include the usual
support for shared variables, and through them to auxiliary processors
that run APL tasks or communicate with the operating system to execute
Unix commands or handle native files. The talk will discuss the
principles behind SAX APL. Assuming that we succeed in connecting the
Sun or AT&T equipment to the projector at IBM, it will be accompanied
by a live demonstration.
Host: R. Dunbar
...