The Nineteenth Byte

New Posts

00:33

0

Introduction Super Mario 64 has a heavily overcomplicated RNG, ably explained here by Pannenkoek. I thought it might be interesting to implement it. Challenge Implement the RNG function, except for the two special cases. Input and output are both 16-bit integers. The standard C implementation is ...

thejonymyster

@Neil most things in unicode should be combining characters

do we have a specific meta for "challenges should be self contained"

Zion mycelia adamancy

@thejonymyster No, because technically speaking, a challenge does not need to be self-contained. However, it's implied by our many metas/help center pages/etc. that say that a challenge must have a clear specification of what constitutes a correct submission

thejonymyster

fair enough :P i think it should be so im still saying it in my feedbacks

Bubbler

The C code is the spec in this case I think

Zion mycelia adamancy

Yeah, feedback can always include that a challenge should be self-contained, you don't necessarily need a source for that :P

thejonymyster

00:44

right ofc i jsut wanted to have authority about it xd

@Bubbler i dont know C code :P

plus like.. cmon just write it out in english why make me jump through hoops

Bubbler

It's just a ton of bitwise manipulation, I don't think it can be explained better in plain English

Zion mycelia adamancy

The problem with writing out an algorithm in english is that it can be incredibly long. Pseudcode might be worth it

thejonymyster

pseudocode sounds ideal honestly yea

Bubbler

though I guess things like (uint8_t)xx can be changed to make the bitwise operation aspect clearer

Zion mycelia adamancy

For example, look at how long it takes me to explain 25 lines of Python code

Bubbler

00:48

and a bit of parenthesizing

thejonymyster

a long challenge description is fine by me

Zion mycelia adamancy

A long challenge description isn't necessarily a bad thing, but when it can be short and concise, it should be

thejonymyster

sure, but what appears to be sparsely commented C code isn't a description

maybe im oppinionated and ppl wont mind except for me LOL

Bubbler

Set s0 to 0
Copy lower half of input to upper half of s0
Take bitwise XOR of s0 and input, and store it in s0

Do you really think it is more readable than the C code...?

thejonymyster

yes?

Zion mycelia adamancy

00:54

I stand by that this style of C bitwise coding is poor style

thejonymyster

your comment about making it more "agnostic" is exactly what i was feeling but didnt know how to say, ty

like with the description sure its still complicated but i immediately know what each word is saying and telling me to do

Zion mycelia adamancy

It doesn't take much to figure out what & 0xFF00 does, but ^ 0x8180 or ^ 0x1FF4? Those seem like arbitrary magic constants to me that just make the code unnecessarily dense and confusing

thejonymyster

overall: should have sandboxed :')

Bubbler

Set s0 to 0
Copy lower half of input to upper half of s0
Take bitwise XOR of s0 and input, and store it in s0
Copy lower half of s0 to upper half of input
Copy upper half of s0 to lower half of input
Take lower half of s0, shift left once, XOR with input and store back to s0
Shift right s0 once, flip upper 9 bits, and store it to s1
if s0 is odd, calculate s1 XOR 0x8180; otherwise calculate s1 XOR 0x1FF4
and return the result

thejonymyster

i do enjoy mario 64 mechanics though

Zion mycelia adamancy

00:57

A few minor edits being required doesn't mean it should've been Sandboxed

Bubbler

Well I'd argue that the two magic constants are what makes the RNG a RNG

Zion mycelia adamancy

Of course, I always recommend Sandboxing unless you've got a lot of experience, but almost every challenge will need a bit of editing on main

thejonymyster

you know what, reasonable

Zion mycelia adamancy

@Bubbler Fair enough, but when the rest of the code is a mess of bitwise commands, it's not especially easy to tell what's a "magic constant" and what is a "bit constant" (e.g. 0xFF00)

Bubbler

Usually the operands to & are regular bitmasks and those to ^ are magic

OldSandboxPosts

01:00

Sandbox posts last active a week ago: Make an ASCII art arrow [code-golf] [ascii-art]

lyxal

01:37

If y'all want to learn about some propositions and logic stuff, you should come join a lyxture

lyxal

02:53

As to the above, the lyxture is over for now

Sandbox Posts

08:25

0

A: Sandbox for Proposed Challenges

Swap every two elements in the list every possible way Inspired by this question. Challenge Let L be a list of n distinct elements. Let P be the set of all pairs of positions in P. Let R be a result of applying a pair-swap operation on L by every pair in P in any order. Example: L = [1, 7, 8] P =...

thejonymyster

14:09

Do we have a question abt resolving a bunch of competing cardinal/ordinal directions?

like the 2d version of bf optimizing >>><< to >

something like NE NW SW => NW

idk if N N N S should output different from N N S or something though

Seggan

15:29

LDQ: is a mondaic or dyadic sort function better?

i.e. the absolute ordering or a comparison between 2 values?

Adám

What?

I'd expect a dyadic sort function to be "sort-by".

And monadic sorts by itself.

Seggan

i meant a monadic or dyadic sort function as input to the sort by

Adám

Oh, you mean like sort by < where < is a dyadic function.

Seggan

yeah

Adám

What would sort do with a monadic function?

Seggan

15:36

ik its used in vyxal, a monadic would be any -> int and the sort is determined by the ints

Adám

So, in APL lingo, this would be a monadic sort operator taking a single monadic/dyadic operand.

Ah, so a mapping.

Seggan

not exactly

Adám

I'd say dyadic then.

Seggan

i.e. sort by length is mondaic

Adám

Yes, sorts by the arguments mapped to their lengths.

Seggan

15:38

yeah

Adám

And if you want the reverse sort (longest first) you'd have to sort by negated length.

But you can always give a dyadic function which preprocesses the arguments before doing some sort of comparison.

Seggan

true

Adám

In APL, that'd be using the Over operator, so if $ is the sort operat, you could write $(>⍥≢) to sort descending by length.

NoHaxJustRadvylf

I literally cannot remove my heat sink screws

Is there some sort of trick I don't know about

Adám

You shouldn't; you'll overheat.

NoHaxJustRadvylf

15:41

I'm gonna end up stripping the screws

Adám

That's better than screwing the strippers.

NoHaxJustRadvylf

Yesss I got it

I just need a properly fitting screwdriver and a lot of force

Seggan

@Seggan ya know what, since i can detect the arity of a function, ill have both in the same operator

Adám

15:56

Aha!

NoHaxJustRadvylf

16:11

I can't get the other one and the screws are nearly stripped

fuck

Yesss, my dad managed to unscrew it

Seggan

dads can do anything

NoHaxJustRadvylf

If this thing boots I'm gonna scream

Alright, everybody cross your fingers

It works!!!

Seggan

16:31

AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA

Neil

17:25

@NoHaxJustRadvylf but Adam said no screwing the strippers!

NoHaxJustRadvylf

17:36

Desperate times call for desperate measures

NoHaxJustRadvylf

19:13

I am...so...incredibly...smart
I spent forty minutes trying to figure out why upgrading to cat7 cables broke both of my servers' connections
Turns out no, it was yanking a whole array of drives from the first and sticking it in the second that did that

Sandbox Posts

0

A: Sandbox for Proposed Challenges

Cube calendar numbers Credit: Marco Verch CC BY 2.0 A two-cube calendar, as shown in the picture, uses two cubes with digits painted on the faces to display the date. For dates in the range 1-9, a leading zero is used ("01", "02", ..., "09"). Now, if you do the math, you might come to the conclu...

Unrelated String

19:34

@Seggan i feel like sort by key is generally more useful in golf than sort by comparison, to the point that if you have a sort-by operator then its overload on dyads should probably pass its second argument to the dyad instead of doing comparisons

definitely have a sort by comparison digraph or some such though

Seggan

eh, i think ill go with having both depending on whether its monadic or dyadic

Unrelated String

19:48

you probably don't have many useful comparison functions that you can express tersely, so if you do need one it's not a big loss to spend an extra byte or so

but sorting by a key that depends on another input is pretty big

granted you can use grade up for that either way but from the perspective of trying to make things as golfy as possible

emanresu A

20:07

@Seggan As a JS user, I can say that dyadic sort is close to useless

I’d force it to be a monad if I were you

Seggan

In java its the only way lol

emanresu A

And it sucks

Unrelated String

21:03

lmao

Wheat Wizard

I just had a random bolt of inspiration on how to golf an answer I hadn't even thought of in probably 3 years.

Unrelated String

nice

Wheat Wizard

Have to put that in the back of my mind for slowest guns in the west at the end of the year.

Seggan

21:23

When the stack snippet has difficulties parsing 1.646 :P

I'm going on a Fig answer rampage rn

NoHaxJustRadvylf

CMQ: Do you prefer the -rwxr-xr-x style of permissions, or 755?

Wheat Wizard

I lightly prefer the former, I can't read either without a guide though.

user

The former. Much easier to read for someone new to Linux

NoHaxJustRadvylf

I'm more of a fan of the second one

Just easier to parse the second at a glance, don't have to group into into triplets

I'm sure if I looked at the first more often I'd get used to it tho

user

I mean, if you're used to the second then it'd certainly feel more natural and it's nice that it's concise

But it takes me around a minute to decode the numbers

A compromise: 111-101-101

NoHaxJustRadvylf

21:33

Ooh I like that

Jul 14 at 19:45, by NoHaxJustRadvylf

Okay fine I'll say what it was. 60 MiB/s for read and 20 MiB/s for write

After testing on my new machine, 160 MiB/s and 60 MiB/s

Seggan

21:45

Whew 6 Fig answers in 40 minutes

Zion mycelia adamancy

22:54

@Zionmyceliaadamancy This challenge should be given another chance. Although trivial in mainstream languages, it's interesting in certain esoteric languages, or languages that are mainstream but not designed to do arithmetic. — Deadcode 22 mins ago

@Deadcode Sure, it's interesting in esoteric and unfit languages. As are basically all trivial challenges, there's nothing about this that stands out, even by that metric. My biggest problem with the challenge is that fundamentally, it doesn't offer anything new to the site. We have countless challenges which are trivial in most langs, but fun in esoteric ones. This is just combining exponentiation and addition, nothing new.

Not every challenge has to be a new and novel idea, but I'm definitely not going to upvote a challenge that doesn't, and I'm for sure going to downvote those that offer so little as to almost guarantee no interesting answers outside the most extreme languages.

Deadcode

@Zionmyceliaadamancy "doesn't offer anything new to the site" - but it was the ONLY question on the entire site I was able to find where I could apply the really cool thing I just discovered about number-of-possible-matches unary regex output.

GingerBot

hi, this is ginger

for various reasons I currently can't get into my main account, so I'll be using GB for a while

just a heads up

@NoHaxJustRadvylf 100% the former, do I look like I can translate things into binary?

Zion mycelia adamancy

@GingerBot Just a gentle reminder about the ban on bots in TNB. It's fine for you to use it, just not as a bot :)

GingerBot

I know, didn't intend to.

just as a temp account until I can get back into my main

speaking of my bots, both of them are offline because we're doing a server upgrade right now!

switching from a pi3 to a pi4

Zion mycelia adamancy

23:10

@Deadcode To be completely honest, while that's a cool regex trick, I don't want (or see the need) for a question for every neat trick in various languages. Besides, the question is open and not deleted, there's no real problem. I just expressed my opinion as to why I believe it merits a downvote

@GingerBot Cool, just checking :P

GingerBot

no problem0

Deadcode

@Zionmyceliaadamancy And your opinion got 13 upvotes. So I was replying to it not just to you personally, but as a proxy to everyone else who upvoted it and presumably downvoted the question.

Zion mycelia adamancy

As a tangent: I dislike the stigma behind downvoting. More people should downvote questions and answers they believe are low quality, or that don't provide value to the site, without feeling like they need to justify doing so

@Deadcode Fair enough

GingerBot

but downvoting answers takes away your rep, which makes it feel like a big serious action of doom

Steffan

it takes away literally 1 rep each

GingerBot

23:15

which is in fact more than zero

this is why I downvote questions but not (usually) answers

Steffan

I likely have more answer downvotes than questions lol

Seggan

@GingerBot same here

Zion mycelia adamancy

Tbf, I say people should downvote more, but I only have 54 downvotes (on non-deleted posts)

GingerBot

fear not, we're all hypocrites here

Steffan

I thought I had downvotes but those were spam posts lol

So they're yeeted now

NoHaxJustRadvylf

23:25

→ 9 messages moved to Off-Topic TNB

GingerBot

tqabgs

Deadcode

Well, since we're talking about votes, I feel I have to ask this. Why would somebody compliment a post but leave it at 0 votes? I've seen this happen multiple times now. I just want to understand it, and currently it totally befuddles me.

GingerBot

[ginger's patented shrugging disapproval face goes here]

NoHaxJustRadvylf

Most likely forgetfulness, I guess? or maybe they saw it in a review queue and just posted a stock "wow this is good, welcome to the site" comment without upvoting

Deadcode

To me, voting something up from 0 to 1 is worth doing just for any post which I know solves the challenge, am sure that it works properly, and seems to have had effort taken towards golfing it efficiently. That's just the minimum. A post that's actually interesting deserves more than just 1, by my system.

NoHaxJustRadvylf

23:36

Same here, so I'd imagine it's usually a mistake when that happens

Seggan

23:49

@Deadcode ^

Deadcode

@NoHaxJustRadvylf I have a hard time believing that.

@Zionmyceliaadamancy Are you aware that in a question with negative votes, answers don't even show up in the site's list of active questions? So I had to bounty my answer for it to even be seen.

And it still doesn't show up on the main list, only the separate bounty list.

Transcript for