> PS: I love reading your (Ian's) posts because they always teach me > interesting new English idioms.
Thanks, Dan, for your bouquet -- if it was a bouquet ;-) But be warned... my daughter says my attitudes and turn-of phrase are seriously out-of-date. I'm beginning to realise my knowledge of APL is out-of-date as well. And as for my J...! Ian On Mon, Oct 25, 2010 at 6:07 PM, Dan Bron <[email protected]> wrote: > Ian wrote: >> If I input: >> >> foo=: +/ >> fie=: foo&99"_ >> fie f. >> foo=: i. 9 >> fie f. >>> ...is [fie or foo] now a pronoun, whereas it was hitherto a proverb? > > Certainly foo is now a pronoun. But that its definition has changed > shouldn't be surprising, as you've intentionally and explicitly redefined it > (if you think of your IJS like a page in a dictionary, where you used to > have "foo /v/, ...." and then you edited it to read "foo /n/, ...." you > wouldn't be surprised if people reading it came away thinking "foo" was a > noun, right?). > > The nature of fie is a bit murkier. But first we must be clear that in > redefining foo, upon which word fie's definition depends, you've also > redefined fie (I expect this statement is uncontroversial but I wanted to be > sure). So the question isn't why fie's definition changed, but what it > changed to. > > The answer is, I don't know. The reason I don't know is because m&n is not > defined in the DoJ. So fie's definition is similarly undefined. If m&n > were defined, then I'd say fie is whatever (i.9)&99"_ is. > > That said, I'm not sure the interpreter would agree. For example, > > foo =: 1: > fie =: foo + 99"_ > > fie f. > 1: + 99"_ > > fie 0 > 100 > > foo =: 1 NB. Changed verb to noun > fie f. > |domain error > | fie f. > fie 0 > |domain error: foo > | fie 0 > > > Here, changing the class of a name upon which a definition depends just > confuses the interpreter; this is in contrast to simply changing its value, > which doesn't: > > foo =: 2: NB. Change to a different verb > fie f. > 2: + 99"_ > fie 0 > 101 > > > All that said, I see your motives for asking this question (teaching APLers > J, without going back to square one) are different from my normal MO > (understanding J from first principles), so I'm not this is the style of > answer that you're looking for. > > I will say that when I was learning APL, knowing J was an immense jumpstart > (and a bit of a bias :). > > -Dan > > PS: I love reading your (Ian's) posts because they always teach me > interesting new English idioms. > > > > > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
