> I like the wiki presentation much more. In addition to content > changes, the formatting on the wiki is really richer than the J pages.
Thank you. It's all been copy/pasted from the Lab, but I admit it does look prettier. The Lab lacks text formatting, but it does do the work for you, so avoiding you misreading anything. All a matter of taste, maybe. > I see now that my description of "open" applies more to "matrix" and > is thus all wrong. No, I thought you had a point. I may well have been stretching the meaning of "open" as it is used in J-argon. You could argue from the naming of the primitive Open (>), plus system verbs: boxopen, boxxopen, that "open" refers exclusively to the unboxing of a boxed atom; or something which could have resulted therefrom. (Oh... plus displaying a script in an IJS window.) I don't want to get too deep into theology. But *par abus de notation* -- as my old friend Nicholas Bourbaki used to say -- perhaps I should have called my "open" stringlist: "SP-separated", by analogy with "LF-separated"? Then in time ZULU could support "CR-separated", "TAB-separated", "CRLF-separated (they're all important to a developer)... though things could get a little out of hand! :-) Perhaps it boils down to this: does my use of the term: "open (stringlist)" cause confusion and distress to a novice user? IMO, things like this actually, can -- even though the novice can't articulate what the matter is. The subconscious is more logical than you'd credit. I don't relish the prospect of retrofitting a more orthodox terminology into ZULU, much less my usage in my own scripts. It would entail my replacing 'o' everywhere, as in 'b2o', with 's' (for SP-separated) -- or should it be 'p', by analogy with my choice of 'f' for LF-separated? Formerly I had 'l' in place of 'f' -- but it looked too much like '1' and made me go boss-eyed -- but the stress of eliminating 'l' everywhere was just too awful to repeat. I do notice however that stdlib.ijs employs the term "namelist" to describe the y-argument of verb: erase. As in: erase 'foo bar'. The verb: erase itself uses (;:) to attack 'foo bar' --but at least I can draw comfort from J "case law": 'foo bar' is called a "list" here, implying a list of two names, not 7 letters. So the term: "open stringlist" isn't entirely the imagination of one IanClark. I humbly await being sat-upon. Ian On Tue, Sep 18, 2012 at 9:47 PM, Brian Schott <schott.br...@gmail.com> wrote: > I see now that my description of "open" applies more to "matrix" and > is thus all wrong. > > I like the wiki presentation much more. In addition to content > changes, the formatting on the wiki is really richer than the J pages. > > On Tue, Sep 18, 2012 at 9:20 AM, Brian Schott <schott.br...@gmail.com> wrote: >> One thing that has occurred to me is that in the early J Phrases >> document, they referred to certain nouns as "specimens" and maybe your >> nouns could build on that framework. Now that I am reminded that zulu >> is a phonetic name, it does not seem so bad, but until I perused your >> work, my only association with zulu was the international date line. >> >> I have attempted below to suggest a brief introduction to this addon. >> Perhaps you could consider some of its ideas? >> >> Because strings vary, sometimes dramatically, in their length, it can >> be challenging to arrange and delimit them in the array structures >> native to J. Some common strategies for managing strings include >> putting each string in a J box, or separating strings with common >> delimiters such as whitespace and linefeeds. Another format, called >> "open" here, is to pad the end of each string with space characters to >> force all strings to be the same length and place each padded strings >> on its separate row of an "open" J array. This addon utility >> provides verbs for understanding these various formats and for better >> converting between such formats. The addon defines several categories >> of specimen nouns, especially the phonetically defined alphabets >> called "zulu" here. By the way, the name "open" is derived from the J >> result of opening an array of boxes. > > > > -- > (B=) <-----my sig > Brian Schott > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
