The Nineteenth Byte

@Downgoat you have a job?

@Pavel No, it's the name of the runtime, isn't it?

@DJMcMayhem Not V? :P

@ASCII-only That's the Common Language Runtime

.NET is the entire stack

@mınxomaτ hyperdrive, I guess?

@ASCII-only What's that

@Pavel nvm, it's a framework

@Pavel this

12:14 AM

@ASCII-only The framework is .NET Framework, not .NET.

.NET by itself doesn't refer to any one thing.

@Pavel exactly

Well that's what I said

It's not an umbrella term either though

But I'd say the framework is probably what people usually mean when they just say ".NET"

See here

1:04 AM

The feeling when you replace a bug with another bug that's more consistent and you aren't afraid of debugging.

Zacharý

1:26 AM

I'm scared of debugging, just not the code kind.

New Meta Posts

0

Q: Getting Started

I've been lurking for quite a while and have realized (after a technical interview) that I'm really out of practice solving problems like what is here. Is there a language that you recommend I start with? I'm a Python enthusiast (and I know Java and PowerShell) and I plan to try and golf with it,...

discussion

@Downgoat Congrats on successfully performing black magic

Christopher

2:42 AM

What did he do

@Christopher Pls follow reply link ty

3:21 AM

CMP: For my command-line-based minesweeper game, should I make the unflag key different from the addflag key?

A fast way of filling the grid quickly is to trace the outline of the numbers and press F-<space> a lot because that's just autoflag-autostep, which is guaranteed to not explode.

however, sometimes I end up tracing onto a flagged square, and then F-<space> unflags and steps on it which causes me to die

ETHproductions

You could have flag/unflag/toggle

true

how about d-addflag, f-toggle, g-unflag

ETHproductions

sounds good to me

3:46 AM

@Dennis This.

Alternatively, you can have a "print Python representation and return z" atom.

@HyperNeutrino ... What is autostep?

@ASCII-only But need userscript to support keyboard. (although that indeed make reading transcript backwards easier)

@user202729 ... what?

@user202729 here...

3

Interesting. (-> star so people can know of that (star with the mouse because I still didn't read it))

But still need userscript. Unstar.

tfbninja

Opinions on whether I should learn DVORAK keyboard layout? (Hopefully from people who have)

@tfbninja I tried (~2 years ago) and failed.

tfbninja

noice

I kinda feel like that might happen

3:55 AM

Because we can't have description-golf, we need something else to make people look at the sandbox more.

tfbninja

@MagicOctopusUrn I have gone my whole life thinking your name was "MagicOctopusum", only to look closer and see urn

4:16 AM

@user202729 basically if you space-bar a numbered square and it has the exact number of flags around it, it will step on every unflagged square adjacent to it. if the flags are wrong then it explodes

Does anyone have a link to that userscript that makes the top bar be sane

Downgoat

I hvae a userscript to make search bar sane but that's it

script tampermonkey topbar navigation theme

5:16 AM

@user202729 but that is a userscript?

@user202729 huhwhat

41

Q: Top-navigation choices

Stack Exchange Top Nav Choices ( install ) ( GitHub ) Top Nav Choices changes the new Stack Exchange/Stack Overflow top navigation based on the preferences you set. The preferences are accessed on your normal Stack Exchange preferences page1. Default changed top navigation on Stack Overflo...

@ASCII-only Thanks!

@Pavel That reminds me, I should reinstall it too

It doesn't work with the gradscript properly, it makes the settings button appear multiple times >_>

Not that big a deal, just a little annoying

5:31 AM

@Pavel well, gradscript is super outdated

Last update was fixing top bar support I believe

@Pavel well it's entirely possible it's a hacky fix

6:15 AM

If I git commit multiple times but only git push once, would github show it as multiple or one commit?

@user202729 Multiple

Because I just do a 5-in-1 commit and after some more reading know it's bad.

@user202729 why would multiple commits just turn into one...

I'm told that was one the big advantages of git over svn: Committing and pushing are seperate systems.

(what is svn? I don't know)

6:19 AM

@user202729 Subversion, another VCS

It's what people used before git was invented, when people suddenly realized that having a bunch of labelled floppy disks lying around was an inefficient way of keeping track of version.

It's not very good.

I'm pretty sure it just stored every single commit as its own directory.

Useful to read: meta.stackexchange.com/questions/92060/…

(at least when you have to write a comment explaining to a off-topic asker about something)

Can anyone other than mods remove others' comments?

@user202729 Don't believe so, no.

6:54 AM

I just had this thought. I know JS, but I never went out of my way to learn JS. Sometime in the last few months, I wrote a program in JS and I didn't have to look up any syntax or anything. I just kinda figured it out by knowing Java and hanging around people who actually use JS.

@Pavel I never went out of my way to learn any language >_>

@ASCII-only You probably didn't just know Python, right? You probably had to look up how a lot of the stuff worked.

@Pavel as in module functions? yeah, of course, nobody has language documentation pre-installed in their brain :P

oh

but doesn't "hanging around people who actually use JS" count as learning JS...

I mostly mean people complaining about how horrible JS is

@Pavel >_> <_< ok so JS is so horrible that people have complained about all of it?

7:07 AM

Just about, yeah. It's a really simple language.

Like, at some point, you were writing a Python class for the first time and had to look up how Python classes worked

JS didn't have anything like that for me. I just kinda already had an idea about how everything worked.

@Pavel not really. pretty sure i learned everything i needed from PPCG and looking at Jelly's source or something?

TBH I still don't know how to create a private field in Python, or if there even are private fields.

I guess I did have to look up documentation on how to write a decorator though

I feel like interpreted languages of all kinds get OOP way wrong.

@Pavel python.org/dev/peps/pep-0008

@Pavel example pls

7:11 AM

How is interpreted related here?

@user202729 Compiled languages don't have hacky things like reflection

IDK. But I find that the way Java, C#, C++, Swift, all do classes in a way I understand, while Python, Ruby, Perl, Lua, etc. all do it in a way that I just can't seem to wrap my head around.

totallyhuman

java, really?

it's not really a good implementation of OOP... :P

Yeah Swift classes are remarkably intuitive, I have to agree.

Yes. I fundamentally understand how a Java class works and all of the kinds of things that can go into a class. Whether or not it's a good way of programming is something else.

totallyhuman

7:14 AM

ah well i suppose that makes sense

Whenever I define a class in Python I feel like I'm writing an awful hack.

Mostly this is due to the inability to declare variables without assigning them, how self works, and the fact that I don't get to use modifiers like static, private, readonly.

@Pavel well, more importantly than being interpreted, those are scripting languages

Yeah, that's what I meant.

Scripting languages are bad at OOP. I suppose that makes sense.

They're meant to be easy to use, not follow OOP conventions exactly

@ASCII-only Decorators would work much better in a language where functions are curried by default

7:17 AM

@EsolangingFruit Yes. But in that kind of language you can't really do optional function arguments the same way

Actually, even just a language like Scala where you can do def f(a)(b)

Scala might even have decorators, idk

Nope, looks like it doesn't

Decorators can generally be replaced by function composition, I think.

F# has function composition for the kind of decorator that actually modifes the input function, and .NET attributes for the @app.route sort of decorator. And automatically curried functions.

I almost typed tio.run/#fraction. Same issue as Mr.Xcoder.

(was searching for fractions module in Python)

@user202729 wait wat.

Anyway how is Fractran now?

Ørjan Johansen

7:26 AM

It's a very rational language.

7

(on TIO)

Ørjan Johansen

Absent, it seems.

@Pavel How are you finding Dyalog APL's classes so far?

I think Dennis was writing an inperpreter for Fractran for TIO

@Potato44 It's so easy

def fractran(i, fractions):
    while True:
        for num, denom in fractions:
            if i % denom == 0:
                i //= denom
                i *= num
                break
        else:
            return

7:32 AM

@EsolangingFruit and inefficient.

@Adám Better than Python's, but defining constructors and such through :Implements makes me prefer C#.

@user202729 An efficient implementation would prime-factorize and treat each prime factor as a counter, yeah

@Pavel You mean re-using the class name for a method to indicate that the method is the constructor?

@EsolangingFruit That's even worse.

@Adám Yeah. That said, :Implements is a much better solution then __init__.

7:36 AM

@Pavel Of course, the naming thing wouldn't have worked with how multiple constructors are allowed.

I was pleasantly surprised at how simple it was to overload the constructor to take different amounts of args.

I actually really like Dyalog's classes.

Or what I know of them so far, at least.

APL Cultilvation seems to always coincide with a day we have a test at school, so I might have missed this, but I have no idea how to have a class with a field.

@Pavel Lesson 17.

@user202729 The way to program in Fractran is to treat primes like counters. Each number in the program represents a bag of prime factors, so each fraction has the semantics of "For each counter, check if it is higher than a certain value. If this is true of every counter, subtract all of those numbers out and then add to a different set of counters. If this is not true of every counter, move to the next fraction."

So taking the prime factorization of every number beforehand would lead to a more efficient implementation.

But prime factorization is non-polynomial in the worst case.

Taking the GCD is better.

Everything about APL makes me think, this is hella weird, but I have no idea how this could possibly be made better. To someone used to C# and Java and the like, :field looks odd, but after thinking about it for a bit, the syntax is actually quite nice.

7:45 AM

@Pavel Btw, you only need fields if you want to make them Public and/or ReadOnly and/or Instance and/or uninitialised. Otherwise you can just use variables like in simple namespaces. Unqualified variables are just like :Field Private Shared.

I'd rather declare them. I think that makes it clearer that a variable is used for the class as a whole, and not in just one spot, and it'll help prevent mysterious VALUE ERRORs when it hasn't been assigned yet.

@Pavel Yeah, it works fairly well. There is one tiny oddity in how :Field "clashes" with ⎕NC, but it is really minor. I hope we can extend the syntax to have :Method public shared dfn←{blah}

Are there C#-style properties? (Custom get/set for variables)

@Pavel Yes, that's what I'm planning to teach on Tuesday.

Neato

How does Dyalog do memory management? Garbage collection, or do objects keep track of a reference count and delete themselves, or something else?

APL inspired me to have a property that I like to include in my code that goes like this: string io {get=>Console.ReadLine(); set=>Console.WriteLine(value);}

It acts like ⎕

7:53 AM

@Pavel Dyalog does of course do memory management entirely under the covers. I don't know all the details (but I can find out specifics, you you can ask Jay). I think everything has refcounts, but there can be circular refs, which the garbage collector takes care of.

@Pavel Neat. I see the inspiration.

@Pavel Actually, more like a ⎕/⍞ hybrid.

⍞ is for StdErr, no?

Oh, I see what you mean. Yeah.

@Pavel ⍞← is StdErr, but ←⍞ is normal reading in from StdIn. If I understand your code, getting io is reading from StdIn and setting io is writing to StdOut. But StdOut in APL is ⎕← and ←⎕ is almost like ⍎⍞.

C# doesn't have an eval :P

@Pavel Of course not.

That said, some guy wrote a C# eval library called DynamicExpresso. It doesn't invoke the compiler or anything, he just wrote an interpreter for a subset of C#.

You can't define your own classes, but you can call methods and do Linq queries and stuff.

8:03 AM

@Pavel Do you want to see how that would look in Dyalog, or keep the suspense for Tuesday?

@Adám I'd rather see now, because I have a history test that on Tuesday Feb 27 18:30 UTC.

Every. Single. Time.

@Pavel You could also invoke the compiler :P

@Pavel I'm sorry about that. Anyway, sneak peek:

:Property io
    ∇ r←Get
      r←⍞
    ∇
    ∇ Set x
      ⎕←x.NewValue
    ∇
:EndProperty

How nice!

@ASCII-only Probably slower.

@Pavel Definitely slower

8:07 AM

@Pavel Funny how the verbose vs. concise parts are reversed. The OO stuff takes up a lot of room in APL, but is almost implicit in C#. The actual I/O takes no space in APL (rememeber that you can actually omit ⎕←), but is very wordy in C#.

Yeah. C# properties are such a basic part of C# that we got something called autoproperties: int foo {get; set;} will define a property equivalent to int foo_; int foo {get=>_foo; set=>_foo=value;}

Why not just write int foo; in the first place? Remarkably convoluted reasons, but changing a field to an autoproperty actually can be a breaking change.

@Pavel recently = like 3+ years ago?

@ASCII-only I didn't mean to write recently. Autoproperties were C# 3.0

So, over a decade ago.

@Pavel I've never needed properties in APL. Apparently, MiServer does use a properties a little bit.

Having a non-constant public field in C# is considered a great sin.

8:16 AM

@Pavel Oh.

Some people were skeptical of autoproperties as being "Glorified public fields" and continued to use the GetFoo()/SetFoo() pattern that Java still has, but these days all public fields are properties or constants..

8:43 AM

Mm, I suppose public fields live in in structs and small record types.

Is it me or is stackoverflow down (as in error 500)?

up for me

9:01 AM

Uh, cracked... — NieDzejkob 30 mins ago

Would this be considered valid?

In the post I said

> Output as characters bytecode values.

The rule said that

> You may also output with leading and trailing characters

Is my wording sufficient to imply that there is no leading and trailing character?

I realized that I made a mistake (it doesn't output the correct result in the intended language. Should I delete it?

Yes.

Q: What is a clipping path service?

Sorry, NieDzejkob.

New Main Posts

0

I need to pix some of my images and one of my friends suggest me to contact clipping path service provider. Are they any good? Do they professional enough and provide satisfactory service? How much they cost and which online CPS is best for this kinda job?my friends suggest me this site: https://...

image-processing

9:31 AM

Why are people so fast at closing off-topic things...

Looks like spam.

@user202729 So fast? We were terribly slow. And I don't think it's spam.

There is an external link.

also ^

Then yes it's spam

Obviously spam. Flag-and-don't-delete.

9:34 AM

I flagged it already

> If six flags get through, that post is deleted and locked by Community (and gets its content shoved into the revision history), which will also remove 100 reputation from that user (assuming they had any, which they almost always don't) and eventually ban their account and/or that IP address.

"eventually"? How many downvotes are necessary for a 6-month question ban?

I wish we could see the number of flags a post already has.

... My previous faulty submission may be a hint...

9:54 AM

... No CMC in 3 days. Sleepy tnb

10:19 AM

Should an imaginary infinity + a real infinity be a complex infinity, or NaN?

Complex infinity.

@Οurous complex

(Mathematica has a directed infinity thing, this would be IIRC DirectedInfinity[1+I])

@user202729 That's what I've been basing it on but it didn't want to simplify I*Infinity+Infinity to anything

> 1/0 is not a directed infinity, but a complex infinity

wat

10:22 AM

Doesn't work.

@ASCII-only But it is. What's the problem?

nothing

it's just surprising

@Οurous Because I*Infinity+Infinity is not Infinity*(I+1)?

@ASCII-only But that's what I'm asking. Is an imaginary infinity, added to a real infinity, a complex one (with, or without direction)

these kinds of things are why we usually work with only 1 infinity when working with complex numbers.

@Οurous Of course. Maybe Mathematica doesn't simplify because it's more specific like that?

@ASCII-only Well, it doesn't know that the infinities are the same size so (I+1)*Infinity is out. complex Infinty, or saying it is undefined are really the only thing that make sense. Or not simplifiying as it has done here because it is probably not programmed to handle that case.

10:34 AM

Reposted my TPLQ answer.

Let me see...

Yes, if the infinities has different sizes it can be in any direction. So, ComplexInfinity.

\o/ Physica can now call a (monadic) function using infix notation

Without direction. (although the direction is limited to 0 to pi/2.

@Mr.Xcoder ...?

@user202729 It can, for example, do Square$x instead of Square[x]

@Potato44 Well, it is defined...

How is that infix?

10:36 AM

@Potato44 Well, the infinity can only be in the first (?) quadrant so it's more specific than complex infinity

@ASCII-only Are we going to have a RestrictedInfinity together with the DirectedInfinity? No.

@user202729 I chose to call it this way because $ is, in fact, the function, which takes two arguments: another function, Square and its argument, x

(like Haskell)

@Mr.Xcoder so, infix function application operator

@user202729 Exactly, so it doesn't simplify

10:38 AM

@Potato44 Yeah, like Mathematica's @

Assume we have RestrictedInfinity[MinAngle, MaxAngle]. Is there any situation where some calculation results in another types of infinity?

(i.e., is it closed under operators?

@user202729 I can see some situations where it is undefined, like when angles opposite each other are in the ranges of two Restricted Infinites being added

@Potato44 Would that not be just ComplexInfinity again

@user202729 Depends on what you mean by another type

@ASCII-only no, undefined for the same reason as Infinity + -Infinity

Q: Codejam for women: CEO search

10:55 AM

By that reasoning wouldn't Infinity + I*Infinity be undefined too

New Meta Posts

0

Where can I find the code source for the solution of this problem? https://code.google.com/codejam/contest/8384486/dashboard#s=p1&a=1 Here is my code I don't know what's wrong. import collections with open('./B-small-attempt7.in') as f: lines = f.read().splitlines() res = "" case...

discussion code-challenge

Have we ever got off-topic on meta?

(apart from this of course)

11:16 AM

CMP: Do you find this intuitive for printing the primes between 1 and 100 (referring to the syntax)?

Yes, but that's O(n²).

Which language is that?

@user202729 I was talking about the syntax

@user202729 Physica (not on tio yet)

@Mr.Xcoder Why did you choose $?

Because @ is a niladic call (i.e Input@ is equivalent to Input[])

And I like $

@NieDzejkob How about Input$ => Input[] and Function @ x => Function[x]? I am willing to change that

Yeah I am going to make that change

11:40 AM

@betseg D: I tried it with the oldest version of our submission and it still borks

looks like the pbrain interpreter got updated

@AlexA. I know it's kinda weird but, eh, I don't think I can appropriately put that in a comment or five :)

or not... I think eof is handled differently

,. with no input prints nothing

12:00 PM

@EriktheOutgolfer Some resources are not stored correctly.

For example the site golfingsuccess.github.io/ararslan-stepdown/… contains to the image golfingsuccess.github.io/content/img/progress-dots.gif which should be codegolf.stackexchange.com/content/img/progress-dots.gif .

Is that intentional?

yes

mainly because I'm afraid relative paths won't work otherwise

Replace them with absolute path to codegolf?

12:25 PM

@EriktheOutgolfer then add the images to the repo...

but they're already there

@EriktheOutgolfer Then why aren't they loading

I'm afraid I can do absolutely nothing about the ones which aren't loading

@EriktheOutgolfer Why not

@EriktheOutgolfer you could have downloaded the website via Chrome and done a better job...

...but that's what I did

12:37 PM

@EriktheOutgolfer as a full website?

yes, that's why there's the folder

hmm well i don't see why you can't add the images now

ngn

@Pavel very good question, it's a combination of both - there are refcounts (unfortunately implemented the naive way and updated more often than necessary) but due to the mutable OOP madness, circular references are possible, so there's also mark-and-sweep; memory use has a limit ($MAXWS) and when it's reached, objects are moved around in order to pack them tighter and all pointers are updated (this used to make sense before virtual memory and 64-bit addressing came along)

@Pavel I don't recommend using ngn/apl but it may be worth mentioning I experimented with something like properties (without OOP): ngn.github.io/apl/web/…

1:12 PM

for a few days now, there are no reviews left for me, wondering why...

... But HexagonyColorer requires RT.Util. Why can't I compile RT.Util on Linux while I can compile HexagonyColorer?

(reply to nothing)

@user202729 wat.

... I think I can somehow modify the EsotericIDE sln/csproj files to get it to compile under Linux.

3:16 PM

@Downgoat yer langs are decidedly weird (see cheddar), but no one can fault you on your website design

very nice website 9/10 not enough (cat|goat)

flawr

3:33 PM

@Riker this maybe?:)

3:55 PM

I've found this amazing work of art:

Ren'Py

The Ren'Py Visual Novel Engine is a free software engine which facilitates the creation of visual novels, a form of computer-mediated storytelling. Ren'Py is a portmanteau of ren'ai (恋愛), the Japanese word for 'romantic love', a common element of games made using Ren'Py; and Python, the programming language that Ren'Py runs on. Ren'Py has proved attractive to English-language hobbyists; over 1000 games use the Ren'Py engine, nearly all in English. == Features == Ren'Py includes the ability to create branching stories, save file systems, rollback to previous points in the story, a variety of scene...

funny, ren'ai means "romantic love" in japanese and the exact same characters in chinese translate to "in love"

@HyperNeutrino tha'ts intentional

yes ik :P

Sen'Py?

2

@HyperNeutrino lawl

brb using that and making a novel about TNB /s

xD I'd be interested to see what the product of that would be lol

4:00 PM

imagining geobits-san and marky-chan as a fun period of TNB history

who is marky-chan o_O

DJMcMaysan

Question: How do I get unprintables in a JSON string? \033 seems to highlight a syntax error on \0 in Atom

Pretty thing: paintschainer.preferred.tech/index_en.html

tip: don't believe in syntax highlighters

@HyperNeutrino geobit's old chatbot

oh lol

4:07 PM

@Blue tha'ts pretty cool

@EriktheOutgolfer using json.loads on it (python) breaks anyway

Yush

I've got some really messy code and was wondering if someone could try helping me make it semi readable/maintainable

In Python?

Python/Ren'py :P

It's rather long

https://pastebin.com/U54JpggU
https://pastebin.com/qbeFLzsX

Only about 3000 bytes... yes that's long on PPCG.

4:15 PM

And that's with some stuff removed due to lack of relevance :P

5:07 PM

So, I'm thinking about making my own regex engine designed for golfing (short regexes, not something like Retina), but that can still be useable (not Unicode shortcuts etc.) Any suggestions for features?

have you thought about a bitcoded huffman-style representation for normal regexes?

Um, I don't know what "bitcoded huffman-style" means

dzaima

@NieDzejkob I have thought about compressing regexes for SOGL, but it turns out there is a lot of stuff that goes into something like that..

5:27 PM

@cairdcoinheringaahing like, you preserve the semantics of regexes, but encode them differently, possibly at a bit granularity to make common characters take like 0.3 bytes

I've had a great idea

I don't have a chromebook, and running chrome apps on windows (8.1, because it broke and rollbacked, now I can't go back to Win 10) is no more than store apps

But chrome apps are mostly just html

so we can maybe run chrome apps on android

even though google doesn't like it :P

@dzaima The way V compresses regex is stupidly wonderfully simple, but it seems pretty effective to me.

@cairdcoinheringaahing Are you thinking ASCII only?

Zacharý

5:43 PM

@DJMcMayhem He's not present now. :p

Haha

@DJMcMayhem Ideally

@NieDzejkob No, I think you misunderstood. I'm thinking about features in the regexes, such as balancing groups in .NET flavour, or having an escape for [A-Za-z]

Then I have no ideas. V uses some encoding tricks, but I can't think of anything else

@cairdcoinheringaahing Definitely use vim's \zs and \ze

What do they do?

Basically, if you have 'abcdef' and run :s/ab\zscd\zeef//, you'll get 'abef'

So it's kind of like a look ahead/behind, but not quite

The mnemonics are selection start and selection end

Is that clear?

5:48 PM

So kind of like s/(ab)cd(ef)/\1\2/ but more general?

Can there be more than one selection in a single regex?

Almost. It's like saying everything before \zs and everything after \ze needs to be there, but isn't removed or captured

dzaima

@DJMcMayhem I wanted something more complex though - stuff like that [...] would never be able to contain multiple characters, making (...)? and (...|) be equal, ect.

@NieDzejkob Never tried it before. I don't think so

Terminal Minesweeper update: autostep reveals less information now and the display is a lot faster (only noticeable on large boards) and added modifiable settings (keybindings and display) (cc @betseg)

5:51 PM

\zs	Matches at any position, and sets the start of the match there: The
	next char is the first char of the whole match. |/zero-width|
	Example: >
		/^\s*\zsif
<	matches an "if" at the start of a line, ignoring white space.
	Can be used multiple times, the last one encountered in a matching
	branch is used.  Example: >
		/\(.\{-}\zsFab\)\{3}
<	Finds the third occurrence of "Fab".
	This cannot be followed by a multi. *E888*
	{not in Vi} {not available when compiled without the |+syntax| feature}

@HyperNeutrino and im still trying to find the problem ;_;

?

@DJMcMayhem apparently, yes

o in your code? ;-;

I really have to start making an app but I haven't ever done such thing

5:51 PM

@lol what app?

@NieDzejkob Where does it say that? I tried it and it didn't work

Maybe I made a typo

@NieDzejkob The thing I said just before

> Can be used multiple times

abcdefghijk and :s/a\zsbc\zeef\zsgh\zei didn't work

@DJMcMayhem ...like the one right there?

5:52 PM

...

redacts history

Oh doi I see it now

I missed the 'd' in the regex

OK, that works

@HyperNeutrino the problem in bigger grids

you mean in your program or mine o_O

@DJMcMayhem elevening has been done to death already...

@betseg hi

Running chrome apps on Android, effectively making it a Chromebook backward (where Chromebooks run android apps with the base of Chrome OS)

5:54 PM

@cairdcoinheringaahing What type of commands will there be? It would be cool if (like in V) there were several different commands that take a regex as an argument

@Cowsquack hi

@DJMcMayhem What do you mean by commands?

I can run chromium os on it, but that's too complicated and I don't wanna break my tablet

do you know what is up with the pbrain interpreter?

So sub is one command, gsub is another command. There could also be sub until text doesn't change or search (only applicable if you have a cursor)

5:55 PM

or I can port Linux chromium to android which is far harder

@DJMcMayhem I think caird only wants to make a custom regex engine (e.g. string pattern matching)

@DJMcMayhem I'm trying to make a regex engine

@DJMcMayhem sub until text doesn't change that's also like sed

exactly what Xcoder said

Ooh. I thought you were trying to make a regex based language

5:57 PM

Lol, Jelly minds think alike (citation-needed)

@DJMcMayhem No, I got fed up with Python's regex engine, and got bored

@cairdcoinheringaahing Something that annoys me about vim regex: You can't use a character class in a range. That would be really nice. So in order to make the range [0-9a-h], conceptually you should be able to do [\da-h] but that doesn't actually work. I don't know if other flavors support that though

Ooh, that could be a neat idea, thanks!

You can do [:digit:a-h] but that's disgustingly verbose

@cairdcoinheringaahing Something I like about vim regex: magic

What does that do?

5:59 PM

@Cowsquack no idea