>One of my first, and most powerful, J epiphanies was that f@:g~ is >actually (f@:g)~ and not f@:(g~) ... but that _it didn't matter_. Can >you think of a case where f@:g~ -: f@:(g~) doesn't hold?
That is a good observation (It seems to be equivalent even for the dyadic case). Yet, this makes Martin's point even more relevant because a beginner might imply mistakenly a parsing rule and get confused afterwards, for example, by the behavior of f&g~ ________________________________ From: Dan Bron <j...@bron.us> To: Programming forum <programming@jsoftware.com> Sent: Sun, November 22, 2009 12:13:51 PM Subject: Re: [Jprogramming] On a curious property of 3435 Oleg wrote: > isMunchausen=: = +/ @: ^~ @: digits Martin replied > This spacing suggests that ^~ is a sub-phrase > of the expression -- which is not true. One of my first, and most powerful, J epiphanies was that f@:g~ is actually (f@:g)~ and not f@:(g~) ... but that _it didn't matter_. Can you think of a case where f@:g~ -: f@:(g~) doesn't hold? When I had this realization, years ago, I remember thinking "I'm beginning to get this stuff..." Oleg wrote: > Note also, no space is expected before =: > unless aligned vertically in a group of lines. I disagree here (and believe this expectation is very prevalent). I prefer space between the name and the copula, so that the name stands out. Assingment is a side effect, and I want my side effects obvious. Often, I'll align all the global (and sometimes even the local) copulae in an entire script, just so it's obvious at a glance what names are assigned. But that's a (long) extension of "aligned vertically in a group". -Dan ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
