« first day (641 days earlier)      last day (894 days later) » 

01:27
@ngn do you have a suite of tests for apl?
is it private?
 
4 hours later…
ngn
ngn
05:49
@JeromeIbanes in ngn/apl i used only these tests. they are public, of course (mit-licensed).
@JeromeIbanes i think you should ask in the apl orchard too. there are many implementations now.
 
2 hours later…
08:17
@ngn did you ever read this? bitsavers.informatik.uni-stuttgart.de/pdf/ibm/apl/… i think it's quite interesting.
ngn
ngn
@ktye i don't remember. i'll read it now.
when you did, answer the question: was there array rank in apl360? or is it more like ktye (length only). codestrings are also interesting.
ngn
ngn
08:35
@ktye i didn't see anything about the language itself. they focus on internals, like the codestrings you mentioned (almost bytecode), how they manage memory, and how they protect the stability of the system.
i would assume, if it was an apl, it must have supported ranks
in Fig 2, there is only number of elements, and 1 for a scalar.
ngn
ngn
08:49
@ktye this representation is close to the source code and doesn't seem fully parsed. maybe "length" here is just the length of the strand literal, and later they turn literals of length 1 into scalars.. who knows
ngn
ngn
09:01
a sequence of ints could be only a fragment of a strand (at least in later dialects like dyalog), so the scalar-vs-vector decision must be deferred until runtime: a←2 ⋄ 0 1 a 2 3
 
3 hours later…
11:34
i guess you are right. jsoftware.com/papers/APL360TerminalSystem.htm Table 3 lists array ranks.
 
5 hours later…
16:05
Is there an easy way to filter the code golf stackexchange for k answers? :S
16:35
@Traws thanks!!
@ngn thanks

« first day (641 days earlier)      last day (894 days later) »