This probably should have gone on chat rather than programming -- we are not doing any programming here.
Also, I think that the interesting thing about J's arrays is that they deprecate storage issues from the type system, and instead focus on some [in my opinion useful] set of mathematical properties. That said, the issues which are significant in dealing with storage issues are still treatable. And, in fact, when viewed historical, the original purpose of the language (of Iverson's notation) was to document those storage issues. So... When I look at computing from outside of J, J is "too complex". The starting assumptions for other computing frameworks that I have examined are inextricably wedded to an ornate treatment of storage issues. J can be represented in those frameworks, but only by largely ignoring the benefits from these ornate enshrinements of bit-level formats. And yet, I see little in the literature which motivates the sanctity of bit-level formats. Most everything I read wants to neglect and ignore them -- and, instead, they become unstated assumptions which pervade and mold the discussions. All of which make the subject of their usefulness a difficult one for me to grapple with. Of course, on the other side of the fence, J has does its own "implicit treatment" of issues which are treated explicitly in storage-format-based-type-system-implementations (which is how I will attempt to distinguish most programming languages). -- Raul ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
