In terms of helping a newbie grok J, "An array is a list of items" was very useful to me. I'm not sure many of the more technically accurate statements would have had the same impact.
For me Raul's "Arrays arrange data." & "Empty arrays arrange nothing." are also interesting but not replacements for the statement above. When introducing complex subjects, it is often useful to "ignore" some of the nuances/exceptions to start with and then "come clean" later on (Lots of good examples of this in Henry's book). Sure the ideal would be not to make inaccurate statements, but if that is getting in the way of getting across the key concepts then I think it is a worthwhile compromise. Obviously what is "acceptable" in an book introducing J, may not be acceptable in the dictionary, where accuracy should be paramount. > -----Original Message----- > From: [EMAIL PROTECTED] > [mailto:[EMAIL PROTECTED] On Behalf Of Roger Hui > Sent: Thursday, 5 April 2007 12:08 > To: General forum > Subject: Re: [Jgeneral] Need a word for a non-atomic array > > You can finesse it by focussing on items instead of saying > list of items. Talking about list of items has other > potential problems when the number of items is 0. > > The dictionary says this in section II a > http://www.jsoftware.com/help/dictionary/dicta.htm > > A cell of rank one less than the rank of b > is called an item of b; an atom has one item, > itself. > > If I need to, I would say things like, > > Items are important; many verbs in J such as > the dyads i. and { and the monad /: are > defined on items. > > without ever saying "list of items". > > > > ----- Original Message ----- > From: Henry Rich <[EMAIL PROTECTED]> > Date: Wednesday, April 4, 2007 4:33 pm > Subject: RE: [Jgeneral] Need a word for a non-atomic array > > > > So, the question is, how often do you need to say "all > arrays other > > > than atoms"? My guess is, not very often. Less often in > J than in > > > APL with the concept of prefix agreement in J. > > > > Not often, I agree. But I have a few phrases that I think every J > > programmer would do well to commit to memory, and one of those is > > > > An array is a list of its items. > > > > which unfortunately isn't true of an atom. So now I have to water > > down the phrase. Ecch. > > > > The phrases suggested so far are accurate but lack zip. > > > > Since atom = 'a-' not + 'tom' divide , i. e. that which cannot be > > divided, I could just use 'tom' to refer to a > > non-atomic/composite/higher-order array. That has zip. > > But no one would understand, and like Roger says, this is not a > > frequent-enough need to introduce a new term for. > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
