« first day (819 days earlier)      last day (954 days later) » 

00:08
Steffan153 requested changes on PR #1278 (Vyxal/Vyxal): "we should also add some for the digraphs that use typable chars."
Apl poggers
in The APL Orchard, 3 hours ago, by Richard
ok done. Just 14 votes now? We must be able to do it better. And almost half of these languages I do not take serious (including Basic and Pascal)
We must rally ourselves
Those unoriginal nerds
Adam saw our commotion about week.golf and immediately announced it in APL orchard
I've got one more source of votes I can rally
But I'll leave that source until it's truly needed
00:23
The third stage is the top two with most votes, so I think we're good
but if we do make it to stage 3, we still have to vote against apl
And that's a problem
Because the discord server is full of old golfers
And they stick with what they know
we might be the only ones that actually use it
If we get far enough ahead we could get everyone to vote for say, J and possibly kick APL off the runnings
@emanresuA that just leads to the same problem
00:35
You'd want something like prolog as the opposition in phase 3
i feel like the real solution is to just convince the guy that holding a vote to add one language every week is a terrible idea :P
One language every month
And yes, I fully agree
Only one a month is inefficient as all hell
00:36
^^^^^
so it's WAY too slow to add all the languages that are worth adding, but once that is mostly dealt with it'll just keep adding languages whether multiple people care or not
I know
Don't worry I've thought this through at length
I can give a whole bunch of reasons why it's beneficial to just add the major golfing languages straight away
keep people happy and keep things fresh
But I can't exactly do the convincing because it could seem like I just want to cut in line language wise
@lyxal Just add every lang straight away
00:39
it's hard to think of reasons not to have already added apl and vyxal
@emanresuA That's true too
considering how ancient apl golf is and how large the vyxal gold community is
Vyxal gold™, the premium teir subscription of vyxal
get exclusive access to the latest updates seconds before they're committed to the public repository
You get patches in real-time as we write them :p
00:42
Most of the time you'd get unmatched parenthisis errors
@UnrelatedString just on that - it'd be a good idea to add a golfing language that is friendly to those not exactly the best at golf
oh yeah that too
The current language selection is already dominated by those who've had years of experience in praclang golfing
So for someone who is less experienced, there's not much availability in terms of language choice
Because all the major praclangs have already been golfed to optimisation
and having a golf-optimized builtin set also lowers the barrier to entry in terms of knowing idioms for those kinds of things
Exactly
00:46
Because there aren't idioms, just builtins
APL, while good for new users solving practical problems, isn't the best for new users to golf with
Neither is jelly
And we have some extremely overengineered tooling
You'd need something like vyxal, 05ab1e, golfscript or pyth (and maybe husk) for new users to have a chance to solve problems
how do i even navigate this website without an account
You don't
00:49
i tried registering with an address with a plus tag because i don't entirely trust this guy and the verification email just never got sent lmao
Just link SE lol :p
i have no idea what "access your private actions and information" entails but it sounds slightly out of scope :P
There's a discord for week.golf if y'all want to suggest the add more than one a month option there: discord.gg/zrEa6MghUm
As I said, I can't exactly do it because of conflicts of interest
01:45
Honestly I dislike the challenges.
They're all chameleon challenges and they'd recieve a bunch of downvotes here
^^ ^
@emanresuA still did
huh okay
@emanresuA tbh same
That's why they need a golfing language with a whole bunch of built-ins to solve challenges with
APL doesn't do that very well
@emanresuA exactly this
You go to do one aspect of the challenge only to find that "hey you actually need to do another unrelated task first"
And it's annoying
Because it means that the majority of the bytes are boilerplate dealing with the inconvenience of the annoying task
Also I bet SBCS languages will be scored as utf-8 instead of SBCS
So rip to any fractional byte languages wanting to be added to week.golf
@lyxal frick.
Perhaps we should create a core version of Vyxal that's ASCII-only
Well by the time you've done that, you'll probably end up with the same score you would using utf-8
Also,that's a planned feature for v3
01:56
huh okay
@lyxal i couldnt and had to do github
Jun 19 at 7:41, by lyxal
okay so for version 3, I was thinking of making it so that every character can be typed using # digraphs made of ascii chars
@Seggan odd, I couldn't get github working
@lyxal nooooo
lets just hope the creator doesnt read our criticism
@Seggan the way the chameleon challenges are going and the fact that hard coding solutions were deleted by the owner in the first week, utf-8 everything seems to be a natural conclusion
And obviously languages like Fig can still be added
It's just that they won't have their full potential realised
@lyxal wdym by "hard coding solutions were deleted by the owner in the first week"?
02:04
Perhaps we should just leave this alone. I'm not sure this is a place we want to be.
@Seggan some solutions must have been hard coding the answers and the owner just removed those submissions
Straight up yeeted them
@emanresuA it's worth the advertising and the investment in the long term success of the site
It may be a little bad at first, but it has potential
Perhaps it will.
Wait...
yeah honestly i would try not to do anything that prolongs the life of that site
02:08
^
The Unrelated String has spoken
i still got a ping from the hyperping
it's not called a hyper ping for nothing
except it just took me a few more seconds to find where the ping was cuz it got discarded
I also still got a ping
02:10
@lyxal wait did the owner announce that or how did you find out
@UnrelatedString the way I see it, being the first golfing language on there would cement that language as being the OG OP language that everyone uses. In the short term, it's a bad idea to invest, but long term, it could be beneficial
@hyper-neutrino discord conversation
hm
I would not manipulate the submission list without a very good reason and being very transparent about it
also if it's utf-8 it's possible to just use bytes 0 to 255 for vyxal right?
wait i'm dumb that's not how utf-8 works
@hyper-neutrino Weird
I think someone was directly accessing the test cases and outputting the corresponding result
but yeah maybe part of this is just subconscious bias against non-fluent english and the tokyo ghoul pfp but i just don't have the feeling that this can turn out well
02:13
@hyper-neutrino Oh wait I think I got mixed up lol
Moving stuff to rooms people can't access stops that from notifying them I think, but it appears it doesn't stop pings
@UnrelatedString I think it's worth a try for at least another month
Besides it's not the first of its kind lol
There's other one challenge a week sites that have vyxal
What?
(unless you count codidact)
oh yeah reminds me the guy's github has a link to that "codingame" thing
i remember we've had some weird experiences with people who came here from there
02:17
It has vyxal 2.4.1 available
Huh okay
Is it sbcs or utf?
Utf-8 baby
For all sbcs languages on there too
I kinda want to create my own site
and write it in vyxal
02:19
@UnrelatedString isn't that the clash of code site used in the code.golf (not cgcc) discord?
i have no idea
oh yeah i think there was also an incident with someone lifting a challenge from something called clash of code
Even with attribution that's probably a bad idea
03:11
codingame has a golf thing in it
two other modes too
i forget
Huh, checked the updated corous and the most used element is now ; lol
that's not an element tho
I know lol
(it uses the lexer, not the parser)
Thanks @lyxal for updating that btw
What on earth, I added ÞT like a few weeks ago and it has 15 uses???!?!?!??!!!?!!1111!?
How many of those are from when it was transpose though?
03:23
oh it used to be transpose. nvm
Not one of my best ideas tbh
In hindsight it should have been a monograph from the start
Idea: We add a "current Vyxal version" tag to online interpreter permalinks
It's just metadata, and doesn't affect anything else
03:40
There's currently a /version thing but that's the most recent commit hash
Which isn't particularly helpful
it would also be nice if you could run older versions. we can disable versions with ACEs :P
but idk
As much as that's a good idea, the limitations of Pythonanywhere prevent that
Which is why I'm really looking forward to just hosting on github pages with scala to js
is heroku any better?
we should really just find better hosting, spin up a docker container, and enable sympy lol
We can enable Sympy by just filtering quotation marks out of strings
...then we might as well do that right now
what takes so long?
03:50
@Steffan but also you gotta remember the fact that when it comes to hosting a backend, I really don't know what I'm doing
@Steffan just considering whether or not it's worth working on symvy
It's been something I've been thinking about for a while
we could just filter out quotation marks while we wait months for u to decide
on the other hand, we can't just filter out quotation marks. that won't do it.
you can use ord and stuff
Dang
Didn't think of that
Filter out ord too then
And chr
lol, then use bytes
Filter out bytes
there's zillions of other ways lol
03:53
Or uppercase the string
man that is a terrible idea and I love it
but what about expressions with x and y
Lowercase after evaluating
maybe just cook up a super hueg regex about what we can do and what we can't
That or transliterate with non-normalising Unicode characters
look over this: codegolf.stackexchange.com/questions/209734/… so many ways to get out.
03:59
Sounds like symvy is needed after all
CMC: Find a character with all five digraph extensions
There are many 4s, but (afaik) only one 5
found it :p
very clever way of saying
:O copilot made a guess that was correct
Next thing we know it'll be in charge of Vyxal
found with a script using parsed_yaml.js
@UnrelatedString very clever way of saying you found it :p
04:10
ninja;d
04:35
Could y'all come have another look at this?
chunkybanana opened PR #1310 (Vyxal/Vyxal) (Vyxal:yeet-oeis → Vyxal:main): Yeet OEIS
chunkybanana deleted branch Vyxal/fix-a-test-case-for-ḭ
05:16
Wait what dammit
1 message moved to ­Trash
I saved a draft.
I'm using gists from now on
Lyxal opened PR #1311 (Vyxal/Vyxal) (Vyxal:fix-a-little-issue-with-split-and-keep-delimiter → Vyxal:main): Make `vy_eval` account for fractions and make split and keep delimiter
only 7 more failing tests to go
I'll be very happy if we can fix them all before the next release
And thank you so much for putting so much time into this
Remember to look after yourself, too :)
@emanresuA dw I am
this is me getting my mojo back after just not being able to even touch code for 12 days
Lyxal opened PR #1312 (Vyxal/Vyxal) (Vyxal:change-some-tests-for-replace-first-occurance → Vyxal:main): Change some tests for `øḞ`
05:27
Oops
Trying to think of an algorithm for set intersection with two infinite lists
Lyxal opened PR #1313 (Vyxal/Vyxal) (Vyxal:change-some-tests-for-replace-nth-occurence → Vyxal:main): Change some tests for `øṄ`
4 more to go
5 actually because I'm considering re-working one of the decisions about types I made
chunkybanana approved on PR #1313 (Vyxal/Vyxal): "Loookfs goodddi to oo ememmmeeee e!!1!!"
no, 4
what should "12" ∪ 12 be?
last night I though ["1", "2", 1, 2]
but now because of øḞ and øṄ I'm thinking ["1", "2"]
That ... sorta makes sense
But honestly I'd prefer to define it as undefined
Which prompted this:
in The Nineteenth Byte, Jul 1 at 2:10, by emanresu A
Do golflangs need heterogenous lists?
05:39
I think ["1", "2", 1, 2]
because you can S on 12 if you want ["1", "2"]
meaning the last 4 items of #1164 are all easy
because they're all the same type of issue
at least I think so
IMO, this just doesn't particularly matter
maybe, but there's broken tests
and they need some sort of resolution
The tests are running a lot faster now that most of them aren't failing :)
Whyyy does python handle "local variable referenced before assignment" at runtime???
Lyxal opened PR #1314 (Vyxal/Vyxal) (Vyxal:make-multiset-difference-respect-original-type → Vyxal:main): Improve the behaviour of `Þ∨`
06:02
LMAO, dup and co don't preserve the infinite state of a list
For that matter, neither does deep_cop
I've come up with something quite elegant:
def gen():
    while True:
    a = next(lhs)
    b = next(rhs)
    if a in rhs.generated:
        yield a
    if b in lhs.generated:
        yield b
06:21
chunkybanana opened PR #1315 (Vyxal/Vyxal) (Vyxal:set-intersection → Vyxal:main): Make ↔ work on infinite lists differently
Lyxal approved on PR #1315 (Vyxal/Vyxal): "One day we'll have to change deep copy and more to account for infinite status "
@VyxalBot @lyxal didn't lol
I've fixed it
And what "more"? Steffan made it work for everything I know of
Cursed idea for when we run out of digraphs: Bring back a single-byte infinite list of positive integers and pass it into stuff like L for extra "digraph" overloads
30 mins ago, by emanresu A
LMAO, dup and co don't preserve the infinite state of a list
Those all rely on deep_copy, which I fixed
Oh you did fix it?
I didn't see that
06:44
Line 126 of helpers.py
I see why I didn't see it
Github mobile hid the def deep_copy() part in the "show more lines" area
Meaning it wasn't syntax highlighted :p
 
4 hours later…
10:35
@emanresuA *depressed music*
On the other hand, I'm learning selenium for a reason ;)
10:50
Guys we did it
The owner of week.golf has expressed interest in adding more than one language per month
yay!
> I see all the same that there is a real enthusiasm around several languages. So I'm going to make sure that as many people as possible are happy this month.
Basically this means that the addition of vyxal is likely
Because getting in at least the top two is easy for us due to our sheer size
As we've seen from voting in phase 2 this month
 
2 hours later…
13:14
And that's it
once those are merged, the tests will all pass once more
 
1 hour later…
14:36
We had the oeis url so that someone here could try and access the oeis because firewall
It's from like June or July last year
15:04
Oh
 
2 hours later…
17:10
14 messages moved to ­Trash
10 messages moved to ­Trash
27 messages moved to ­Trash
!!/prod
Vyxal-Bot opened PR #1320 (Vyxal/Vyxal) (Vyxal:main → Vyxal:production): Update Production (2022-07-15)
 
3 hours later…
20:26
@lyxal Niko/okie IIRC
But we don't need that now, and the oeis.html page doesn't exist anyeay
chunkybanana approved on PR #1319 (Vyxal/Vyxal): "WOOOHOOO!!!!!!!1!11!1!!!!!"
21:11
cgccuser merged PR #1319 (Vyxal/Vyxal) (Vyxal:fix-last-test → Vyxal:main): Fix last failing test
cgccuser deleted branch Vyxal/fix-last-test
🎉
I fricking deleted it.
1 message moved to ­Trash
@user Would you mind having a look at this?
22:12
I tried, but I don't understand it, sorry
The aliases themselves look good to me tho
22:50
@user Thanks :P
@user It's pretty much "try and match as many alphabetic chars as possible, if that fails, try digraphs"
23:35
Steffan153 merged PR #1320 (Vyxal/Vyxal) (Vyxal:main → Vyxal:production): Update Production (2022-07-15)

« first day (819 days earlier)      last day (954 days later) »