Editorial: Very Well, Alone Then
Our Ken Iverson commemorative issue is still not ready to print and we’ve fallen behind our publication schedule. So here in 21.2 and shortly after in 21.3 are the issues we’d been planning to follow.
This last year has seen the near demise of Quote-Quad and the official APL conferences. ACM-SIGAPL, if not actually dead, has been on life support for a while now. A rescue team has been recruited to refloat it, and we wish them well.
Arthur Whitney was in London in February for what has become Kx Systems’ annual meeting with its users here. (See my note of a talk by one of them.) The day after his conference, he met some of the BAA committee. This was the first time we’d met him since Iverson’s death. “So,” he said, counting faces, “it’s just us now.”
And here it’s spring. The future of Dyalog Ltd has been secured for a while under the management of Morten Kromberg and Gitte Christensen. (See Sustaining Members’ news.) The London community of Q programmers continues to grow. This year’s conference suggests they might be ready to come out of hiding, and we plan to be carrying Q articles in most issues of Volume 22.
In my own work with Optima Systems we’ve been developing tactical systems for a large UK pension provider. Over the last year a handful of us have produced recurring savings in our client’s operations of over £750,000 a year. Our client’s IT division is said to be the second largest in the country, and has for years been deprecating our technology as ‘non-strategic’. But as business conditions get tougher, our results are winning us an audience of managers who wonder if we might not have something to teach them about developing software.
Of course it’s an old story: a handful of APL programmers produce results the ‘official’ developers can’t match. In four decades we’ve heard this story so many times. And we know how it ends: the productive mavericks get sidelined, isolated and finally squeezed out. The tortoise beats Achilles by cutting his tendon.
Perhaps not this time. The results are in on Software Engineering by now. No one seriously believes great advances in productivity remain to be squeezed out of formal methods. Agile Development and Extreme Programming have done serious damage to the doctrine that there is only one good way to produce software. The listening among business managers is changing.
This will be too late to inspire many of our readers, who have watched the decline in APL’s usage and reputation in the last decades of their careers. But there’s money to be made; it’s only a question of who will do it.
Doing their bit in this issue are Stephen Mansour, checking where points fall between boundaries; Sylvia Camacho following up from VECTOR 18.2 on Bell’s Inequality; Paul Cockshott on compiling array expressions; and Devon McCormick follows up on the Monty Hall problem from VECTOR 21.1. Richard Smith writes from Cambridge University, where he has constructed Rowan, an APL-like interpreter, out of .NET parts. Those undergraduates and their spare time!
We have the fourth and last instalment of Dan Baronet’s series on tools. This is on Regular Expressions, a powerful tool fundamental to programming in Perl, and widely used by (e.g.) Ruby and PHP programmers. Even if you don’t care to master this tool, study it long enough to point rudely to it the next time you hear the crack about APL code looking like noise.
Mark Wickens uses Java to translate source code between APLs from different vendors — keep grinning, J and Q programmers. Michael Baas likes distributing his Dyalog APL workspaces as EXE files and shares tricks for polishing the result. More — very soon — in 21.3.