« first day (1700 days earlier)      last day (947 days later) » 

5:12 AM
@rak1507 It doesn't quite convert to 'false', but rather ⊂'false'. This allows for the lossless conversion of [false,"false",0]
 
 
3 hours later…
8:28 AM
@KamilaSzewczyk ?
 
8:47 AM
@Razetime it says Dyalog APL Help 18.0
But the link leads to the 12.1 version
 
 
2 hours later…
11:12 AM
@PaulMansour right, but why not 0? chances are if you're using 'false' in JSON you'll want 0 in APL not ⊂'false'
ok, then it isn't lossless, but that's just a consequence of APL not having a separate boolean type
 
 
1 hour later…
12:24 PM
@rak1507 True, but it's convenient to be able to write out a JSON false value and a 0 and not to suffer the consequence of APL not having a separate boolean type. No?
 
@PaulMansour Not if you then have to convert the boolean value into 0/1 yourself.
interestingly, if you do ⎕JSON⍠'M' the type you get is 6, or 'No APL equivalent'
 
@rak1507 I thought there was a variant for importing true/false as 1/0, but I think I'm mistaken.
 
Can't find anything about that in the docs, but there could be. It does sound plausible
 
1:26 PM
@Adám @KamilaSzewczyk they know about this issue...
 
1:54 PM
1
Q: APL skill without lambda functions

Vishal SolankiI am creating Alexa echo show skill, I have created one sample Alexa skill without lambda function that hits myserver api and I return response in Alexa accepted format like below: { "version": "1.0", "sessionAttributes": {}, "response": { "outputSpeech": { "type": "PlainText", ...

 
 
1 hour later…
2:55 PM
@StackExchangeFeeds cool apl is alexa programming language
why does it exist?
 
 
2 hours later…
4:45 PM
@rak1507 Common Lisp has a similar problem. Not between false and 0, but between false and the empty list. The JSON library I'm using defaults to the "natural" way (making them equivalent), but a flag allows you to read it in a reversible way.
Clearly the same is needed in the APL case. No one solution fits all use cases.
Of course, one could make the default the reversible form, and then provide some function that converts a structure from that format to the natural one.
 
4:57 PM
@EliasMårtenson yeah, I think that would be preferred, it's definitely too late to change it now because of backwards compatibility but maybe a variant option to do false true->0 1 would be good
 

« first day (1700 days earlier)      last day (947 days later) »