« first day (840 days earlier)      last day (695 days later) » 

00:00
What are some examples of things that would go wrong if dyadic % was replaced by / and " //" was used to introduce comments?
Anything other than people being confused because / isn't the traditional APL symbol for division and/or loss of backward compatibility?
Well, to my knowledge \ is intentionally chosen to be symmetric to /, because most (all?) meanings are closely related. Split-join, decode-encode, reduce-cumulative reduce, etc.
Changing the symbol loses the symmetry.
I'm not suggesting changing / when used with other operators. I'm only suggesting changing / alone when used as a dyadic operator like x / y.
where x & y are lists or numbers.
Again, x/y and x\y are symmetric in that meaning too.
so :/ etc wouldn't change, the way I'm picturing it. But I'm not an expert and I haven't thought it through...
(need to look up K9 manual thing to see what changed to what, but)
00:04
I see what you mean. / and \ as join/split is consistent.
@jordancurve the tricky part there is handling the x f/y scenario (versus f x/y). there's this
(basically, is x in x f/y part of the reduction (i.e. the seed or number of iterations) or is it just a UDF/noun)
@coltim Interesting, thanks.
Is there a way to get &1 0 1 to behave similar to &1 0 2? that is, I want a way to force the "generate x0 copies 0, x1 copies of 1, ..., x_N copies of N" behavior, regardless of whether every element should get at most 1 copy.
What's the difference?
@jordancurve I dunno if this addresses what you are looking for, but...
Doesn't &1 0 1 return 0 2, which is 1 copies of 0, 0 copies 1, and 1 copies of 2?
00:16
Hmm, maybe it does! I was thinking it was just giving me the indices where the array is non-zero, but maybe that's the same thing as the other behavior...
Never mind, then! Thanks, Bubbler.
 
9 hours later…
08:47
@JohnE the Readme for oK says it's an impl of K5 while the manual says K6
which one does it aim to implement
ngn
ngn
@Razetime they are very similar. see 5to6.
09:20
good link
 
4 hours later…
13:19
@ngn you cannot parse lists with empty elements? e.g. (1;;2)
ngn
ngn
@ktye true, i can't :(
13:38
@ngn same for x[;1]:0 to set a column or x[1;]:0 to set a row.
ngn
ngn
@ktye that one may be worth fixing

« first day (840 days earlier)      last day (695 days later) »