The Nineteenth Byte

MilkyWay90

02:10

Anybody know string slicing in Sisi

I got that so far

Riker

I haven't seen DLosc recently (chat says 27 days ago, main says 2 days, but hasn't posted since early april), and I'm not certain anybody else knows that one

MilkyWay90

Oh

Riker

yeah he hasn't done anything visible since april 5th, you'll either have to figure it out yourself or give up I believe

MilkyWay90

02:43

How to write a program

Jo King

03:27

Any last minute thoughts on my digitsum question before I post to main?

xnor

@JoKing Is there some upper bound on the input? You might get the usual questions on integer types, etc

especially because interpreting in a high base b and then taking the result modulo b-1 works for digit sums at most b-1

Riker

@flawr nice comment from mechanics.se

@BlueRaja-DannyPflughoeft Thanks, you've convinced me that flipping my car upside down for repairs was impractical just moments before I attempted it — wedstrom Feb 8 '18 at 18:58

Jo King

@xnor I was going to let the standard "number won't be above your representable number data type" handle it (though looking for a meta ruling, we don't actually reference that outside of the standard loophole)

so i'll state it explicitly

xnor

@JoKing What do you say to the inevitable Python 2 question of whether we can assume it's an int (<2^64) rather than a long, so that its string rep doesn't include an L at the end?

Jo King

@xnor is that still a thing? I thought that was removed in later versions of Python 2

TIO

xnor

03:40

Only in the later version known as Python 3 :) It happens with backticks

TIO

Jo King

oh. I'm going to say use the actual str function, not the repr one then

xnor

ok, that's reasonable

flawr

06:42

@Riker flat earth confirmed :D

@Riker but if it works on my bike why not on my car?

New Main Posts

08:33

0

Q: Words Everywhere

Given a list of words, output a square grid of characters, so that all the words can be read from the grid, by moving horizontally, vertically, or diagonally between characters, without reusing characters in the same word. The grid must be the smallest possible which allows all words to be read. ...

Anush

08:57

hello!

Anush

09:35

at https://codegolf.stackexchange.com/ it says "
Maximizing by swapping[...]modified 6 hours ago Sriotchilism O'Zaic 35.8k"

but if you click on that link it wasn't modified 6 hours ago

what does this mean?

New Meta Posts

0

Q: Is there support to approach 3rd parties to build a performance tool?

For code-golf questions, TIO has been a superb resource. However, for fastest-code challenges there is no equivalent tool available. I believe we could explore working with one or more 3rd parties to build such a tool; it would be in the clear interests of performance analysers, cloud providers...

discussion

user202729

@Anush Probably some answerer modified their answer.

Anush

@user202729 wouldn't that also show as being top in the answer list at codegolf.stackexchange.com/questions/185001/… ?

maybe they added a comment and then deleted it?

user202729

That user was the last one to reopen the post.

Anush

interesting

but it wasn't reopened 6 hours ago was it?

the user has a great name too!

user202729

09:47

That user changes their name like every month...

Anush

aha ;)

New Main Posts

10:11

0

Q: (Soft-Question) How to implement my board-game idea?

I am a big admirer of logical board games like chess, sudoku etc. In my free time, I think of ways of tweaking rules to make games even more interesting (and sometimes think of inventing a new game altogether) but every time I feel restricted as I have no way to demonstrate my idea. I think of ma...

programming-puzzle

New Main Posts

11:10

0

Q: Intercommunication between threads

So glad I found this cool site. Wanted to try sharing a coding challenge a friend gave me, an interview question. Usually it is written in java, but you can use whatever language you want, just have fun! Challenge: write efficiently and securely a program that creates an array of n threads the c...

array-manipulation java multi-threading array

New Sandboxed Posts

11:31

0

A: Sandbox for Proposed Challenges

A factorization game Player one picks a positive integer \$10^3 \leq x \leq 10^{15}\$. Player two outputs \$k\$ distinct positive integers \$y_1,\dots,y_k\$ such that \$(y_1+1)(y_2+2)\dots(y_k+1) = x\$. She gets \$k\$ points. The challenge is to write code for player two. Input A single int...

Veskah

12:34

@NewMainPosts Hmm, pretty sure this should shed the java tag as its open to all languages

Anush

12:52

if anyone has any comments on my challenge, gratefully received codegolf.meta.stackexchange.com/questions/2140/…

Adám

@Anush I don't understand the sequence (y1+1)(y2+2)…(yk+1). What is the third term?

Anush

(y_3+1)

@Adám does it make sense?

Adám

@Anush So you alternate between +1 and +2?

Anush

@Adám no! Hmm.. let me see why this is confusing

Adám

@Anush Oh, so now I really don't get it. What is the fourth term?

Anush

13:02

@Adám typo!

thanks

:)

they should all be +1

Adám

@Anush Why have the +1 there at all?

Anush

@Adám it makes the challenge a little more interesting

that is a little step away from factorization

Adám

@Anush No, k is the same no matter what you add or subtract.

Anush

let's take 127381 and see (you may be right)

Adám

If all the yₙ are distinct, then so are all the zₙ if zₙ=yₙ+c

Anush

13:08

actually 1099511627776 is more interesting

Adám

@Anush If I understand your text right, your challenge is actually just: given a positive integer x, determine the maximum number of distinct integers that have the product x.

New Main Posts

-1

Q: Send first Email of each unique domain using C/java

Challenge Given n email addresses of different domains, please send an email to the first address(in alphabetical order) of each domain. Assume function sendmail() to send emails. You need to return an array with the list of all first emails corresponding to each domain. Take any sample array li...

string code-challenge array-manipulation fastest-code array

Adám

@Anush If there exists a set y such that (y₁+1)(y₂+1)…(yₖ+1)=x then there must also exist a set z such that z₁∙z₂∙…∙zₖ=x, no?

Anush

hmm.. I am thinking :)

Adám

@Anush Ah, there is actually a difference. If the elements of y must be strictly positive (not just non-negative) then y₁≠0 and so you just cannot include the factor 1. If the number is even, y₂ will be 1 to represent the first factor being 2. So with your complication, one just has to subtract 1 from k.

Anush

13:14

thanks! I knew there was a reason :)

Adám

@Anush Sure, but it isn't really interesting to subtract 1 from the result, is it?

Anush

not that interesting, no

the output is just k of course

if you have a preferred wording I would be happy to use it

Adám

@Anush Given a positive integer x, answer the maximum number of distinct integers that have the product x.

That's it. Then some test cases and you're done.

Anush

so 1099511627776 gives 9 in this case

right?

Adám

@Anush I suppose, if 8 was correct before.

Anush

13:22

question changed

thanks

Adám

@Anush Why the whole story about players and naming the output and talking about "optimal output"? Just state what's to be returned.

Anush

@Adám I thought it was more fun :)

but I can make it more minimalist if people prefer

Adám

@Anush Then put a tl;dr at the top or a background story section marked "safe to skip". At least some people prefer less fluff.

Anush

it is quite minimalist now

(I changed it)

Emigna

The player-bit at the start makes it feel a bit like we're making koth bots

Anush

13:27

take another look

there is very little left in the challenge :)

Adám

@Anush Only the interesting part is left.

Anush

:)

I should say, it's non trivial

if you want to be able to solve it for the full range I gave

Emigna

I would add the actual numbers to the output example, like (1, 17, 59, 127) for the 127381 example, to make it a bit clearer

Adám

@Anush Why is the upper limit specifically 10¹⁵?

Emigna

I didn't immediately realize that we should include 1 (although it should be obvious since we're maximizing factors)

Adám

13:31

@Anush I'd limit it to 10⁹ so languages that represent integers with 32-bit can compete.

@Anush Also, why include a lower limit at all? For 1, the answer would be 1, no?

Anush

I can't read 10⁹

what is that?

10^9?

Emigna

yes

Adám

@Anush Yes.

Anush

are there really languages restricted to 32 bits these days?

moonheart08

@Anush BYOND.
Of course, BYOND is not a language, nor is it a game engine. It is purified awfulness in a tin can

Anush

13:36

:)

moonheart08

But yes, BYOND is 32-bit only. Probably because it depends on Internet Explorer for it's functionality

that and because it's probably so chock full of undefined behavior (It is, i've disassembled parts of it) it simply cannot be upgraded to even a new compiler version

Anush

CMC: Find 9 distinct positive integers whose product is 1099511627776.

just to keep people busy :)

and hopefully make them interested in my challenge

moonheart08

Anyways, i need to go wash off my hands now. Typing out that accursed name has tainted me

Anush

which name?

moonheart08

The B name

Seriously, it's awful, the fact it still exists is an evil itself

Emigna

13:38

@Anush [1, 2, 4, 8, 16, 32, 64, 128, 4096]

Anush

@Emigna too fast!

:)

Emigna

Well, it was a 2**x number

moonheart08

Anyone else here do x86-64 assembly?

Anush

@Emigna true.

Emigna

So not that hard just to take the first 9 2**x numbers and increase one

Anush

13:39

@Emigna did you do it by code or hand?

Emigna

code

Anush

oh cool

1*2*4*8*16*32*64*256*2048

also works

of course :)

Emigna

Yeah, there are a few different solutions

512, 1024 also works for last 2 ofc

Anush

true

oh a comment by me has been starred!

Emigna

You should probably change the wording on the sandbox post to show that it is "a possible solution". Sounds like it's the only one right now

Anush

13:44

it's so minimalist I am tempted to post it now

moonheart08

Question: Is there a pathfinding challange on PPCG?

Adám

@moonheart08 path-finding

flawr

I'm sure I've already participated in multiple:)

Anush

thanks for the improvements all

Skidsdev

Is there a guide or reference sheet I can use to read up on fancy math notation syntax in SE?

flawr

13:46

(some of them are not that obvious)

moonheart08

Sidenote: Should i learn C#? I currently know Rust and Javascript, and am using C# a bit for unity, but don't really know it

Skidsdev

I mean if you're using Unity you should probably learn C#, but it's really up to you

Adám

@Skidsdev The sidebar help when asking a question links here. You don't need the parts about the document and choosing a flavour. The basics are here.

moonheart08

@Skidsdev what i mean is learn it for use beyond just Unity :p

Adám

@moonheart08 How about learning something radically different? — like APL…

moonheart08

13:51

:p
APL is on my TODO list

J. Sallé

@Adám I love the topic named "Getting Stuff On Top of Other Stuff". That's exactly what I googled when I needed to use TeX for the first time.

moonheart08

LISP is on my TODO list as well

Also plan on learning RISC-V Assembly, but for that i want real hardware, and real hardware isn't very cheap at the moment

Skidsdev

@Adám Thanks!

Anush

@Emigna your code gives the output 3 instead of 9, right?

Emigna

@Anush: yes, but for a different number (1 higher than your example)

Anush

14:02

@Emigna oh ok.. it looked similar

ideally it would run for the test cases. But who am I to be fussy

Emigna

I changed it to a a number beginning with 2

Anush

thanks

Emigna

@Anush Inefficient code is short code ;)

Anush

yep :)

I still haven't found a way to ask for short code that isn't inefficient without risking the wrath of the ppcg community :)

Emigna

The standard-ish way of "must complete test-cases within x seconds" seem to work decently. Although there is always the question of on what specs

It definitely is harder making questions asking for optimized programs in time/memory

New Main Posts

14:07

2

Q: A factorization game

Input A single integer \$1 \leq x \leq 10^{15}\$. Output The maximum number of distinct positive integers that have the product \$x\$. Examples Input: 1099511627776. Output: 9. One possible list of factors is: (1, 2, 4, 8, 16, 32, 64, 128, 4096). Input: 127381. Output 4. One possible list ...

code-golf math

dzaima

@Anush it just doesn't make sense to ask for that - you're asking for two completely different tasks in the same answer, which kind of fits for this. For this I don't think it's that bad, but in the very big powers question golfing is definitely just an afterthought, despite being the main focus

New Sandboxed Posts

0

A: Sandbox for Proposed Challenges

A Note On Frequencies In western music, we denote "musical notes" as a letter A-G, optionally followed by a ♯ or ♭. An "octave" is a collection of 12 notes, such that the frequency of any given note is double that of the same note in the octave below, and half that of the same note in the octave...

Skidsdev

@NewSandboxedPosts I've finally learned my lesson abuot posting seemingly simple challenges without sandboxing first :P

ais523

@Skidsdev your challenge is surprisingly close to a kolmogorov-complexity; there are only 12 possible outputs, and it's not immediately clear to me whether calculating them or hardcoding them is shorter

Skidsdev

@ais523 there are a lot more than 12 possible outputs, my wording was just unclear that the number in input refers to the octave

Skidsdev

14:47

@ais523 Is that better?

Anush

@dzaima my perspective is just completely different. I see it as "write the shortest code for any non trivial algorithm for this problem"

we have tags for restricted-time and restricted-complexity for exactly this reason

Skidsdev

@Anush people want to write fast code or short code, not both. If you want short code, but only want non-trivial answers, you need to find a challenge that has no trivial answers, but be warned you'll get far fewer answers

ais523

@Skidsdev yes, that's clearer

Anush

@Skidsdev I do see this.. it would be much easier if we had a TIO for fastest-code

it's not satisfactory to have to time code on your home pc

ais523

@Skidsdev I'm normally happy to golf restricted-complexity but it can often take a while to find a good algorithm

Skidsdev

14:50

@Anush Somebody mentioned that on meta recently

dzaima

@Anush but when writing the algorithm is like 99% of the challenge, golfing, being the 1%, feels way less required, despite being the scoring system

Anush

@dzaima well you have to design an algorithm which can be golfed

@dzaima there are often faster algorithms which are huge in code

Skidsdev

See I have no formal CS or Math education, everything I've learned is self-taught, so when it comes to PPCG I tend to prefer more trivial challenges, like yesterday's Shantae Dance parser

ais523

the reason is that with pure unrestricted code-golf the shortest program is basically the program that describes the problem as succinctly as possible, with restricted-complexity you need to force a particular algorithm, so now you need a program that describes the algorithm as succinctly as possible

Anush

it all seems like perfectly legitimate coding challenges to me

ais523

14:51

and algorithms are way longer than problem descriptions, normally

Anush

@ais523 that makes sense

@Skidsdev I see that

ais523

I am not normally happy to golf fastest-algorithm because there's normally numerous optimisations you can do to the algorithm, and having to golf them too takes the fun out of it

Anush

but the task of designing a fast and succinct algorithm seems worthwhile to me. I realise not everyone agrees :)

you can choose if the winning critierian is code size or speed

restricted-time/restricted-complexity sets the winning criterion as code size

ais523

well, golfing restricted-complexity is really worthwhile from the language design point of view; a language that is good at that would be a language which is good at finding optimised algorithms given a problem description, and a language like that would be really useful

Anush

I agree

as an example, codegolf.stackexchange.com/questions/185062/… is a much less interesting challenge because I was bullied (joke :)) into not adding restricted-time

if I had put a max of 10 seconds, then the problem is very cool and quite hard

ais523

14:54

that's a really good pure code-golf problem too, IMO

so people didn't want the alternative wincon because it wouldn't reward golfed programs

it might be good as a restricted-complexity problem too, but possibly not (it wouldn't surprise me if it's NP-complete)

Skidsdev

@Anush See this is where the issue lies, it's not that there's any flaws in your logic, golfing algorithms under restricted-complexity or restricted-time is still golf, and there's no inherrent problem with challenges that do that, it's just that opinion-wise, most people prefer not to.

Anush

@Skidsdev that makes perfect sense to me

dzaima

@Anush for that, I'm actually not against restricted-complexity (less so with restricted-time but still acceptable)

Anush

I just don't like it when I get 5 close votes and 10 down votes in one day for usign restricted-time

ais523

restricted-time is really hard to make objective

the same program won't run in a consistent time even on the same computer, usually

Skidsdev

14:56

From what I see in PPCG, especially when golfing in non golflangs like Python and JS, a lot of the fun of golfing is stretching and twisting the language in ways so time/performance inefficient that one would never get to explore in practical code

Anush

restricted-complexity is a lot less useful than you might think if you want actual code

Skidsdev

Although thinking about it I could definitely get behind some restricted-complexity golfing

Anush

the problem is that real code needs a max size for the problem which means all algorithms are constant time

ais523

why? plenty of languages have bignums

Anush

so we would have to agree what we mean by the complexity of real code

@ais523 that's quite a restriction

dzaima

14:58

@Anush PPCG usually should let you slide by not defining the max input size & requiring that the code theoretically works on all inputs (or so i remember from a while ago and don't currently have any meta links to back me up)

ais523

@Anush well for restricted-complexity we normally have something like "even if your language doesn't have bignums, you can assume that it does", together with specifying the complexity to be assumed for, say, adding or multiplying bignums

Anush

@dzaima that seems reasonable as long as everyone agrees what it means for a piece of C code to theoretically scale to large types

@ais523 that sounds good

Skidsdev

@Anush Except we're not dealing with "real code" in any definition. Code on PPCG is and should never be production-worthy, and algorithm complexity is a theoretical construct, meaning with restricted-complexity, the theoretical complexity of the code is what matters, not the "real" runtime

ais523

that said, even a trivial problem like multiply two numbers becomes really hard as fastest-algorithm

I think the fastest algorithm was only discovered recently

Skidsdev

and the latter point is only an issue if you overlook the former point

ais523

14:59

and is really complicated

hmm, apparently it hasn't even been proven optimal, so there might be still faster ways to multiply numbers out there

Anush

@ais523 well only if you care about it being exact. I just want it to terminate in a reasonable amount of time. So 1 minute is fine, 1000 years is not

ais523

20

Q: What is the fastest algorithm for multiplication of two n-digit numbers?

I want to know which algorithm is fastest for multiplication of two n-digit numbers? Space complexity can be relaxed here!

algorithms mathematical-programming

Anush

I guess I could add restricted-complexity and say it should be poly time

would that be acceptable to people?

ais523

@Anush yes but you have to be objective otherwise it's impossible to compare answers with each other

Anush

@ais523 isn't that objective?

the winning criterion is still code size

but the code has to run in poly time in theory

hal.archives-ouvertes.fr/hal-02070778/document

dzaima

15:05

@Anush but where do you draw the line? is 59.99s fine? is 60? is 59±10s? is 60±60?

ais523

oh, polynomial time

Anush

@dzaima sorry I moved to poly time. How about that?

ais523

your factorisation problem possibly can't be solved in polynomial time, though

Anush

@ais523 poly time after the factorisation step :)

ais523

there is a polynomial-time prime-test, so we can determine if the answer is 1 or not in polynomial time

Anush

15:06

hmm.. maybe I should just supply the factors as the input

ok so.. say I provide the factors as the input

and then the algo has to run in poly time

dzaima

@Anush that's very very easy, no?

ais523

see it as a set-theoretic problem: "you have a multiset, what's the most submultisets you can divide it into so that all the submultisets are different?"

New Sandboxed Posts

0

A: Sandbox for Proposed Challenges

Generate A038666 code-golf (Needs a better description, of course.)

ais523

… and now that's helping me to come up with a better code-golf algo for the original problem

dzaima

oh wait, grouping different numbers together is a thing

Anush

15:10

@dzaima no! I mean, what do you have in mind?

@dzaima right

if I change it now people will kill me of course... sigh

ais523

@Anush: it's best as two different problems anyway

dzaima

@Anush this is why you should keep a question in the sandbox for more than 3 hours :|

ais523

the code-golf as you have it atm, and the fastest-algorithm based on a multiset

Anush

sure but if I pose another related problem now they will also kill me :)

ais523

if you disguise it correctly people won't be able to tell it's the same problem :-D

Anush

15:12

@dzaima I agree in principle but the difference in attention you get in the main site is just massive. There should be a "1st hour" on the main site where you answer it at the risk that it will be changed :)

@ais523 good thinking!

Erik the Outgolfer

actually, I agree, reposting a challenge under a completely different winning criterion is allowed

Anush

@ais523 oh you think fastest-algorithm! I would prefer restricted-complexity

@EriktheOutgolfer I know but they killed me when I did this last time!

dzaima

@Anush posting similar questions a bit of time apart, not immediately after is a start

ais523

trying to work out the fastest solution to this is the interesting part IMO (although you have to be careful because there are two size parameters not one, the number of distinct factors and the maximum multiplicity of a factor)

Anush

@dzaima I'll wait until tomorrow.. still scared from last time :)

Erik the Outgolfer

15:14

multiplicity shouldn't count here, since we're only concerned about multiplicative partitions with distinct factors

ais523

@EriktheOutgolfer multiplicity matters because p² is different from p

Erik the Outgolfer

maybe you mean something else

ais523

say your input is 8, i.e. 2³; the solution is [2¹, 2²]

here, the only distinct factor is 2, with multiplicity 3

Erik the Outgolfer

are you talking about prime factors?

ais523

yes

Erik the Outgolfer

15:16

that's a different story altogether though, because the challenge is about all factors, not just prime factors

maybe that would matter as an implementation detail

ais523

no, the challenge is effectively about grouping the prime factors; that's what a multiplicative partition is

you're dividing the prime factors into groups so that no two groups are identical, and trying to form as many groups as possible

dzaima

@Anush I'd suggest a bit more, and definitely keep in the sandbox for a bit longer. I'd like to even have a little bit of chance of seeing the question before it's posted :p

Erik the Outgolfer

@ais523 yep, that would be an implementation detail :P

moonheart08

The ultimate form of madness: Be insane enough to attempt to write a self-hosting JIT VM

ais523

atm I suspect that the greedy algorithm is optimal here, but I'm not sure, and it's not clear how efficiently it could be implemented; you probably don't have to form the actual groups

Erik the Outgolfer

15:19

and then Dennis sweeps in and proves you right... ;) :P

ais523

@moonheart08 that's useful due to the eigenratio of 1 if you do it correctly (i.e. each extra layer of VMs gives only a slowdown of a constant amount of time, rather than the constant factor as usual)

moonheart08

@ais523 why layer VMs?

It's JIT. Have the VM just manage itself

ais523

not because you have to, but because you can

moonheart08

Of course there will still be 1-2 layers to bootstrap it

Gotta have an interpreter first before i can prototype anything :p

but yea it is useful, hence why i'm doing it

ais523

although some self-modifying esolang ideas benefit from having very efficient self-interpreters; the inner ones can be running a slightly different language from the outer ones

moonheart08

15:20

it's just that pulling it off is, uh, crazy

as the VM'd language has to be complex enough to generate native code and interface with the OS

and of course, do it efficiently so the programmer actually writing the VM doesn't go insane

Anush

@dzaima :) sounds good

moonheart08

@ais523 you wanna discuss this in a new room? I'd love some advice

Skidsdev

Is there a room specifically for discussion of design and development of languages and language infrastructure?

moonheart08

idk

Skidsdev

ie Compilers, Interpreters, VMs etc

moonheart08

15:25

Probably the big thing that will cause me difficulty is the fact i'm going to need a full type system, and can't get away with a small one that omits some, say, integer types, early on

also the fact i'm going to have to design a usable language atop the underlying bytecode, and chances are high it'll end up usable to me, and opaque to everyone else

moonheart08

16:26

Question: Is it possible to unmap the executable data from RAM on linux?

i want to try and reclaim the memory used up by ELF data

ngn

16:40

@moonheart08 wouldn't that cause a segfault on the next instruction?

moonheart08

@ngn All my code will be elsewhere eventually, hence why i want to do it

I just want to free up the ELF data once i stop using it.

Also, I found a decent solution to the type system issue. Just omit what i don't need, and make the language into a fully selfhosted system, with the bootstrap interpreter pretty much only being used to generate the first version into a ELF

because it'll be simple enough early on i can get away with just doing something C like for the initial works

of course being able to generate a ELF file doesn't mean it'll stop being a VM

hence why i want to be able to clean up the unused ELF data once setup completes

Skidsdev

18:48

Finally getting some more work done on Clam :D

HyperNeutrino

19:18

ok so i'm trying to make my site to support mobile and I have <meta name="viewport" content="width:device-width, initial-scale:1" /> and in my style i have body, html { width: 100vw; }

but for whatever reason, the body is not as wide as the screen (testing both on my phone and chrome dev console)

it probably has something to do with CSS pixels not being real pixels or whatever because the body says it's the right number of pixels wide but it's too narrow still

any idea why? i can send the link (not posting to SO because it'll take a really long time to copy-paste every relevant thing and I can't just give away a repo link and I don't even know what's relevant to this issue)

ais523

22:23

@moonheart08 I'm not an expert on JIT/VMs, most of my language experimentation is in compilers and traditional interpreters

Jo King

22:45

@HyperNeutrino Should you be using html as the tag? Try body instead

ais523

22:59

@HyperNeutrino try setting margins and padding on html and body to 0 to see if that's the issue (note that you'll probably want some margin or padding in the final version of the page, writing right up to the edge is ugly)

ais523

23:11

@moonheart08 I believe it's possible to unmap the segments of a running executable, just so long as the code is currently running from some other segment

Anonymous

@HyperNeutrino You want =, not :, in the content key-value pairs

ais523

@Mego not in CSS, that uses : between key and value, = is for HTML/XML

Anonymous

@ais523 Viewport tags are not CSS

ais523

ah, I see, there's both CSS and non-CSS in the comment

Anonymous

developer.mozilla.org/en-US/docs/Mozilla/Mobile/… has the best docs for viewport tags that I've found

ais523

23:22

MDN is my go-to reference for JavaScript (and occasionally HTML/CSS when needed, although I'm not really a frontend dev)

Transcript for