Jelly Hypertraining

@PyGamer0 Not much more to it, but the current way is very good, and I can't think of an obvious, recurring pattern which would be shorter. That said, it wouldn't surprise me if there is a more optimal set of patterns out there

PyGamer0

12:22 PM

in The Nineteenth Byte, yesterday, by ais523

my main focus has been on the chaining rules, because Jelly's aren't very good in practice, frequently requiring extra bytes to disambiguate

Zion mycelia adamancy

12:43 PM

@PyGamer0 Not sure I entirely agree with ais, but ¯\_(ツ)_/¯

1 hour later…

Unrelated String

1:47 PM

i feel like dyadic chaining rules have a highish rate of needing extra bytes but they're still very good
all i can think of to improve that without the sort of overhaul he's thinking of would maybe be some way of using monadic chaining rules dyadically; i.e. treating the right argument as the original left argument

but then you'd need to add stuff that actually lets you do that

having some grouping quicks that compose stuff in a way that's independent of normal chaining rules is a good idea but you don't really want to have too many of those

2 hours later…

PyGamer0

3:51 PM

how useful is the 0,2,2 rules?

Unrelated String

as in the rtl-ified 2,2,0? it's conceptually just giving 2,0 higher "precedence" over 2,2 so you don't end up with a 2,2 then an unparseable 0

you can also solve that by actually grouping 2,0 before applying other chaining rules

though the consequence of making it possible for a monadic chain to 2,(2,0) may be undesirable

i guess you could introduce a fourth pseudo-adicity for "monadic except it always splits the chain"

actually wait that's brilliant

that completely solves the
wait no never mind

PyGamer0

@UnrelatedString i got excited until i read the last sentence lol

@UnrelatedString in general how often do you use that in dyadic chains?

@UnrelatedString and what do you mean by acidity?

and what is your sleep schedule! do you only sleep for like 4 hours a day?

lol

Unrelated String

@PyGamer0 reasonably often; it's also the case that the alternative is almost never desirable

@PyGamer0 highly variable™

PyGamer0

so a 2,2,0 rule for dyadic chain is useful... i will add that then..

Unrelated String

@PyGamer0 i thought it could introduce a meaningful distinction between dyad-nilad-quick and dyad-quick-nilad for quicks that handle monads and dyads uniformly but then i realized those would both produce the always-split adicity

of course if the design is super jelly-like already you'd also have to worry about quicks that take an optional nilad argument

or a non-optional nilad argument come to think of it lmao

PyGamer0

4:06 PM

yeah the design is super jelly like

so the only thing i changed is the handling of unparseable nilads

so 1 2 3+3 5 5 gives [1,2,6,8,8]

1 hour later…

PyGamer0

5:15 PM

actually i can just preparse to find the unparsable nilads and those nilads become a list

that sounds smart