...resending: (Google rejected because of my tinyurl link...) ---- km says:
> Who are your "beginners"? That is indeed the Question To Ask. The whole area was comprehensively researched back in the late 70s by Morton, Hammond, Barnard, Long (et al), at the Medical Research Council Applied Psychology Unit, Cambridge, England. They came up with something known as the BIM (Morton Block Interaction Model), expressed as a simple but immensely powerful diagram of how users approach a new computer system. Let me stress this was not some sophomore's starry-eyed woolly thinking, but the distillation of years of hard-nosed research by world-class experimental psychologists. IBM gave it over 9 years of funding from 1975 onwards. The results seeped out into various areas, including the design of new "task oriented" computer documentation (though much credit must also go to Alphonse Chapanis & his students at Johns Hopkins, who liaised closely with this research). Quickly Googling: Morton block interaction model ...the only good refs I can find on the web are for pay-for papers. A free source showing the BIM diagram is here: [tinyurl] bp36fhg (see: slides 22 and 23 of this PowerPoint presentation) Applying it to J and its documentation, you can explain (and even predict) why students brought up on the Texas Instruments calculator have difficulty with J, and where. Ditto why the very word "trains" will invite misunderstanding over the behaviour of: (u v w) noun as opposed to: u v w noun You can also predict why a fresh young mind will grasp J with a lot less trouble than an experienced programmer in whatever language, even APL. For a junior, the boxes in the BIM are mostly empty, so they can't interfere with each other. Brief exposure to BASIC at school will rapidly fill the boxes and ruin the said junior for a quick uptake. IMHO... Hitherto we've all just been dabbling here. A proper analysis of J and its documentation using the BIM would pay serious dividends. If I were a silicon valley venture capitalist, I'd demand such a study as part of a funding program for J. It would hilite the problem areas of "beginners" using J (especially "experienced" beginners) and inform the writing of documentation to remedy the problems uncovered. The pay-off? You could propel J centre-stage for statistics and data mining worldwide, and knock Mathematica and Python (beloved of Google) into a cocked hat. The effort? It's the sort of thing you could give a computer science or human-factors MSc student for his/her dissertation. Volunteers? IanClark On Fri, Nov 30, 2012 at 6:25 PM, km <[email protected]> wrote: > Who are your "beginners"? Beginning calculus students are thrown by right to > left evaluation because they are used to Texas Instruments calculators' > Algebraic Logic System. For them J is not a nice calculator. I gave them > rules for how to do arithmetic calculations with J, and showed them only > explicit programming. These were university students. Roger has reported > pleasant experiences with young students, and I notice your son was 10! > > Kip Murray > > Sent from my iPad > > > On Nov 30, 2012, at 10:17 AM, Bo Jacoby <[email protected]> wrote: > >> J is a rich language, and it might be a good idea to define an elementary >> subset for beginners. J is a nice calculator for elementary computations. >> You can do a lot of computing without knowing anything about binomial >> coefficients and taylor expansions and capped forks. If you need to >> understand everything in order to be happy, then you may be unhappy. I >> taught my son elementary APL when he was 10 years old, and he loved it! >> >> - Bo >> >> >> >> >>> ________________________________ >>> Fra: Ian Clark <[email protected]> >>> Til: [email protected] >>> Sendt: 16:06 fredag den 30. november 2012 >>> Emne: Re: [Jprogramming] @: and capped fork >>> >>> I too have felt the need for Really Prominent Pages (RPPs?) and I've >>> regretfully come to conclude that the only workable solution for the >>> beginner is the forum, as Dan hints. Hence certain issues will turn up >>> over and over again in the threads. >>> >>> Experienced programmers, trying to pick-up J from the reference manual >>> as they go along, are especially vulnerable. What "experience" can >>> prepare you for J? >>> >>> Trains are indeed a pratfall, and I think it boils down to this. When you >>> see >>> f g h noun NB. (1) >>> even if you've read and re-read all the stuff about J's bracketing >>> rules, you're still tempted to think you can write: >>> v=: f g h NB. (2) >>> v noun >>> ...and it gives a different result. So you begin to suspect J is only >>> for people addicted to brain-teasers. It doesn't help to be told that >>> (2) is a "train" and (1) isn't. (sic) >>> >>> But I didn't start this thread just to gripe (yet again) about the >>> opaqueness of J and its documentation. Researching computer >>> (un)usability and how to fix it has been my career, and I was drawn >>> into J in the first place by these very considerations. It has proved >>> a rich hunting-ground, even more so than APL, which hitherto had been >>> the best I'd encountered. Industrial experience has taught me you need >>> to spend millions to write good documentation -- unless you want to >>> stick to telling people things they know perfectly well already, which >>> J emphatically doesn't. >>> >>> Oh, don't get the idea I only value J the way a doctor values a sick >>> patient. Having climbed the learning curve, it's my language of >>> choice, and I don't regret it. Why, I'd even pay good money for it. >>> (And if you're an APL user you say that with a gulp!) >>> >>> On Fri, Nov 30, 2012 at 2:50 AM, Alex Giannakopoulos >>> <[email protected]> wrote: >>>> I agree that all this needs more and *clear* and *visible* explanation! >>>> It was the first question I asked, and in the two+ years (on-and-off) that >>>> I have been mucking around with J I have seen asked on this forum umpteen >>>> times. >>>> I have also seen it debated at length (and breadth and depth). By >>>> experienced users, experts and newbies alike. Unfortunately these >>>> discussions get buried in the (not easily searchable) list archives and the >>>> question keeps popping up, kind of like Nosferatu. >>>> >>>> Maybe we need a REALLY prominent page explaining it once and for all. >>>> I must add that for someone coming from another language, the idea of >>>> trains is massively alien (and a pain until you see their utility). >>>> ---------------------------------------------------------------------- >>>> For information about J forums see http://www.jsoftware.com/forums.htm >>> ---------------------------------------------------------------------- >>> For information about J forums see http://www.jsoftware.com/forums.htm >> ---------------------------------------------------------------------- >> For information about J forums see http://www.jsoftware.com/forums.htm > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
