kim@amdahl.UUCP (05/07/87)
[ For all you do ... this line's for you ... ] I recently received a brochure from Spencer Organization, Inc. for a product called "APL.68000" (in response to my having circled a number on the bingo card in Computer Language magazine). Though I am specifically interested in the Amiga version, I'm cross posting this to all the comp.sys.groups that cover machines for which the product is available (Amiga, Mac, AtariST, and PClone). BTW, the PClone version comes with a 68K board, and is called MultiAPL. Has anybody had any experience with APL.68000 [or MultiAPL], or with Spencer Organization, Inc? The brochure says APL.68000 is a trademark of MicroAPL Ltd. Are these folks in Great Britain, or ...? I've always wanted to play with APL since college [no "language wars", please], and am wondering if this package would be worthwhile. It seems to be a pretty complete implementation from the (meager) information in their literature, and does provide interfaces to the graphics facilities, serial ports, windows/menus/etc. on all supported machines. Two concerns on the Amiga implementation ... dunno if multitasking is supported (no mention), and don't know if access to the native file system is provided (mentioned for all *other* implementations). They are asking $295 ($995 for PClones, w/68K board), and have a 30-day money back guarantee, which seems in-line with (Amiga) compiler prices. Any info would be appreciated; I'll summarize any email responses that I receive to the net. Thanks! /kim -- UUCP: kim@amdahl.amdahl.com or: {sun,decwrl,hplabs,pyramid,ihnp4,seismo,oliveb,cbosgd}!amdahl!kim DDD: 408-746-8462 USPS: Amdahl Corp. M/S 249, 1250 E. Arques Av, Sunnyvale, CA 94086 CIS: 76535,25 [ Any thoughts or opinions which may or may not have been expressed ] [ herein are my own. They are not necessarily those of my employer. ]
ward@chinet.UUCP (05/14/87)
If you are only interested in "hacking", and have access to a PC or would consider a clone (you WERE considering a board for a PC, no?), might I recommend STSC APL? It costs $595 (list), but they have a version minus the full-screen editor and a few other things, called "Pocket APL". STSC is an EXCELLENT company, good products, good support. 800-592-0050 (Rockville MD)
graifer@net1.ucsd.edu (Dan Graifer) (05/15/87)
In article <1021@chinet.UUCP> ward@chinet.UUCP (Ward Christensen-) writes: >If you are only interested in "hacking", and have access to a PC or would >consider a clone (you WERE considering a board for a PC, no?), might >I recommend STSC APL? It costs $595 (list), but they have a version >minus the full-screen editor and a few other things, called "Pocket APL". > STSC is an EXCELLENT company, good products, good support. > 800-592-0050 (Rockville MD) STSC APL*PLUS/PC is an excellent product, except for one drawback. Because of the brain damaged segment register architecture of the intel processors, there is a 64Kbyte size limit on all objects(vars or fns). For floating point numbers, that is 8K numbers max per object. I believe that, to keep the interpreter size down, the also make no storage distinction between boolean and integer data. Since I believe it uses 4 byte integers, booleans can have at most 16K elements. This is most limiting when doing things like string searches. Fortunately, they provide a quad-ss system function to do this. I have the first release of the APL*PLUS/Mac system. Still a few minor problems in the first release, but very fast. Uses as much memory as you've got. (I've got 1MB, I have friends with 4MB). I've created variables with 100,000 integer elements. Used Domino to do OLS on 2200 cases of 20 variables (ie y <domino> x, where x has 44000 floating point elements). You can't do that on any PC/clone. Dan Graifer graifer@net1.UCSD.EDU Disclaimer: Nobody ever listens to me anyways; Why should they start now?
ljdickey@water.UUCP (05/16/87)
I have just returned from APL87 held this past week in Dallas. Both companies, Spencer Organization and STSC, were had booths in the exhibit area. I agree with comments made previously that STSC has an excellent product and would add that APL.68000 is also an excellent product. Timing tests may sometines be deceiving, but I mention that on some tests, APL.68000 on the Mac is about 5 times as fast as STSC's APL*PLUS on the Mac. The reason for the speed of the APL.68000 is that it is written in hand crafted assembly code. A very nice advantage, indeed. This product has been around for several years, and has most of the bugs shaken out of it by now. (You might have seen it on Wycat for instance.) The major marketing thrust now is for the Atari ST, the Mac, and the Commodore Amiga and the ports wore done in that order. Hence, claim the developers, the Amiga version has the best interface of the three. One representative from STSC claims that they have uniformity across different versions. Among others, they offer versions for the Sun, the IBM PC, and for the Apple Macintosh. The Mac version seems to be descended from a product called PortaAPL, previously of Cambridge, Massachussets. Since STSC claims now to have uniformity, it looks to me as though they bought the Mac interface and the customer base and plans to expand on that. In recent years, STSC has brought many developments to APL, including nested arrays. Maybe these things will be coming to to the micro versions. The company provides excellent user support. -- L. J. Dickey, Faculty of Mathematics, University of Waterloo. ljdickey@water.UUCP ljdickey%water@waterloo.CSNET ljdickey%water%waterloo.csnet@csnet-relay.ARPA ljdickey@watdcs.BITNET UUCP: ...!watmath!water!ljdickey
bchso@uhnix2.UUCP (Dan Davison) (05/18/87)
In article <3172@sdcsvax.UCSD.EDU> graifer@net1.UUCP (Dan Graifer) writes: >STSC APL*PLUS/PC is an excellent product, except for one drawback. Because >of the brain damaged segment register architecture of the intel processors, >there is a 64Kbyte size limit on all objects(vars or fns). > Dan Graifer Not any more. Version 6.3 ('m not sure about 5.X versions) now support objects greater than 64K. Amazingly enough, the speed does not seem to be changed much...although I added a V20 and an 8087 at the same time. BTW, updates and patches are FREE from the STSC bulletin bboard. I upgraded from 6.0 to 6.3 for about $5.00 in phone calls.
miner@ulowell.cs.ulowell.edu (Rich A Miner ) (05/18/87)
In article <6487@amdahl.UUCP> kim@amdahl.UUCP (Kim DeVaughn) writes: >I recently received a brochure from Spencer Organization, Inc. for >"APL.68000". I am specifically interested in the Amiga version I just received APL.68000 for the Amiga, here is a quick review. I cannot comment much on the language, I am new to APL. Included in the cloth bound box was: a single Amiga disk, with the APL interpreter, example programs, and libraries; an APL.68000 quick reference booklet; a spiral bound manual with sections on, getting started with APL, APL concepts, and an APL.68000 reference; key stickers to map out the APL keyboard; and a separate booklet on the Amiga specific features of this implementation. This version of APL does allow you to access most of the Amiga advanced features. They have demo programs that shows the opening of windows and the generation of graphical objects, all with little code. You can also use music and sound features using built in functions like OPENSOUND and MAKEWAVE. I have not done any benchmarks though I have started playing around with some heavy math, circular convolutions and single value decomposition (time to get that 68020,68881 combo). We will be using the APL to prototype Image Processing algorithms and maybe as a front end language for a Parallel Image Co-Processor board that we have built for the Amiga-A2000. >Two concerns on the Amiga implementation ... dunno if multitasking is >supported (no mention), and don't know if access to the native file system >is provided (mentioned for all *other* implementations). It does support multitasking and it allows communication using several methods. You can spawn off different APL tasks during your session. These different tasks can all access the Amiga ClipBoard. One method for inter- task communication is using the APL.68K commands PUTCLIP and GETCLIP to transfer small bits of information between applications. An other more interesting method is the use of their component files. The APL.68K componant file is shared between all APL sessions. The component file system can thus be used to pass any APL data (including overlays of APL functions) between different APL sessions. There was no mention of using message port and I am not sure weather these features can be accessed. There are also commands to access and create files using the native Amiga file system. If you have any specific questions send EMAIL. -- Rich Miner ULowell-Cntr for Productivity !ulowell!miner 617-452-5000x2693
ken@braegen.UUCP (Ken Marchant) (05/19/87)
> STSC APL*PLUS/PC is an excellent product, except for one drawback. Because > of the brain damaged segment register architecture of the intel processors, > there is a 64Kbyte size limit on all objects(vars or fns). For floating > point numbers, that is 8K numbers max per object. > Dan Graifer > graifer@net1.UCSD.EDU The one drawback mentioned no longer fully exists. With version 6 of APL*PLUS/PC there is no longer a limit of 64K on data objects. They can now expand to the workspace size. There remain a couple of limitations; Using "large" (>64k) objects slows things down quite a bit and some primitives are not supported (domino, gradeup, gradedown most notably) and give a "nonce error" when used. The restriction is usually there because use of the primitive would likely cause a "ws full" to occur. Also, for pre version 6 users, some of the problems can be avoided by placing a large data object into a RAM disk and manipulating it using Native file commands; This can be surprisingly fast. Upgrades from earlier versions are available for a nominal fee. The Braegen Group is the major Canadian distributor of STSC products (micro, super-micro, mini, and mainframe). For Canadian users our phone number is 416-366-6363. -- Ken Marchant The Braegen Group, Toronto, Ontario (allegra,linus,ihnp4,decvax)!utzoo!yetti!geac!braegen!ken