--- Dan Bron <[EMAIL PROTECTED]> wrote:
> Regarding tree data structures in J, I have drafted an article on the wiki:
>
> http://www.jsoftware.com/jwiki/DanBron/Temp/Tree
>
> I would like to publish this as an Essay or a FAQ, but want to make sure it
> answers the question effectively, and presents the answer appropriately
> (i.e. I want J in the right light).
>
> Please read the article and make any changes you see fit. That includes
> changing the answer, if you like. When it settles down, I'll cut a final
> draft and move to Essays (with a pointer from the FAQ).
A few observations:
- in general, it should sound more upbeat and leave out unneccesary
technical detail: "J punishes you with 5 bytes per item"...
How much does Lisp punish you per item?...
- no need to speak about "horrific" obstacles in long gone
previous versions that had been addressed
- should rather focus on strengths of J, e.g. integer
list can very compactly represent self-referencing structure,
easily traversed with transitive closures
- should emphasize APL legacy, e.g. A Programming Langugage book
has many in-depth and sophisticated tree processing
algorithms discussed, which is, of course, all based on arrays.
- Remind J advantages over generically building those structures
and loops in C etc. Of course somebody wrote STL trees and XML
components already, but similar libs can be done in pure J or
called into (e.g. xml/sax addon)
- instead of complaining what you don't have (e.g. {::),
appreciate what you do have ("map" script, unfetch)
http://www.jsoftware.com/help/user/script_map.htm
http://www.jsoftware.com/pipermail/programming/2007-March/005531.html
____________________________________________________________________________________
Get easy, one-click access to your favorites.
Make Yahoo! your homepage.
http://www.yahoo.com/r/hs
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
