APL90 Reflections

Transcribed from the printed issue; not yet reviewed. Read it in the PDF of the issue, page 95

The conference was very well organized and APLers are definitely improving in their standard of presentation and use of high-tech audio visual aids. Nearly all of the authors spoke to or about their papers rather than just summarised them, and thus made actual attendance at the conference worthwhile. The organisers are to be congratulated in getting the proceedings out in advance.

Conference Highlights¶

First ever delegates from Moscow and Leningrad. First time the proceedings were distributed to early registrants in advance. Premiere of Iverson's new "J" notation, with free software. Public showing of Dyalog APL/X for HP and RS/6000, and new DOS version 6.1. Public showing of IBM's APL2 for RS/6000 and MicroAPL's APL.68000 level II. Stanley Jordan's APL music and impromptu guitar performance.

Conference Lowlights¶

STSC's reluctance to display any experimental APL version under X windows. Rainstorm timed to coincide with conclusion of fireworks in Tivoli gardens. The price of beer in Copenhagen (although the Swedes come over to drink it).

Exhibition Wars¶

There was healthy competition in the exhibition, and Dyadic systems were clear winners with their readiness to show the pre-release Dyalog APL/X and upcoming new version 6.1 of the DOS product. STSC gave a paper about APL under X but did not show any software. They appear to be hamstrung by the size of their installed user base and are very careful about new features. They have no immediate plans to call APL from C (they say no one has asked for it), but they hope to release a port of APL*PLUS II to Unix by the end of 1990.

Future of APL¶

As always, there was much discussion about what sort of future, if any, APL has in conventional data processing. I usually leave a conference feeling much more optimistic than at the start, having seen what can be done, and this conference was no exception. Although conference numbers were down to about 300, the depth and variety of APL applications were as broad as ever. APL continues to be a competitive option for programming language choice.

Optimistic points:

• in Europe, APL continues to be one of the few environments which support different national languages yet provide international compatibility.
• Jim ("APL2") Brown of IBM has almost convinced the US Dept of Defense that APL2 is the best basis for a common prototyping language for DARPA (the defense advanced research projects agency). If this comes off it will be big.
• APL is used in the forefront of technical developments in medicine, music and space science, as well as computer processor research and language theory.

Pessimistic points:

• Reuters have "stabilised" Sharp APL so that there will be no new language developments, and support for the mainframe product could be reduced. There were a lot of ex-Sharp people at the conference looking for work.
• IBM have not included APL2 in the SAA because there is no AS/400 version. This means that a lot of "true-blue" IBM sites will not consider APL.
• APL has been around for so long that it has no appeal for managers or users who insist that their systems be leading edge.

Whither APL Conferences?¶

The family of APL languages has reached quite a mature stage. The conferences now include developments in ISO APL, APL2, A, J, NIAL and even Mathematica. Through the conferences, attendees learn of the latest developments in many related fields, such as OOPS, neural networks, parallel processing, and other buzzwords, mainly because someone can explain them or prototype them very clearly using an APL model or notation. Despite too much navel-gazing, the conferences serve a good purpose. Long may they continue.

The next conference is planned for Stanford University, Palo Alto CA, in August 1991. The one after that may possibly be in Leningrad, June 1992 or 1993, so start saving your roubles for some white nights.

See you in Palo Alto!