« first day (2063 days earlier)      last day (2783 days later) » 

4:02 AM
@RohanJhunjhunwala Pulled.
 
FYI everyone, I updated the TNBDE Useful Queries page to include some of those line graphs I posted on Monday.
I'll try constructing an average wordiness per day graph query tomorrow.
 
Are there any active TIO users here right now? I'd like to pick your brains about something.
 
What?
 
I use TIO (very rarely though)
 
TIO v2 will have servers-side permalinks that store the code on the server and are accessible via a short URL (just like Ideone, repl.it, etc.). Given that, is there still a use case for client-side permalinks (storing the code in the URL, like TIO currently does)?
 
4:05 AM
probably not
 
Do you mean a use case for generating new permalinks that way, or a use case for making sure the old ones still work? Obviously for the latter it would break a lot of links if not done, while I don't see an immediate use for the former.
 
The old links will still work, no worries. They provide a really limited feature set though (code, input, arguments) while the newer ones will be rather complex (multiple test cases, multiple languages, etc.), so if client-side permalinks are still required, I'd have to revisit encoding them.
@Quill That's what I'm thinking, but I wanted to know if I'm overlooking something. I thought privacy may be a concern (in some challenges, you don't want to make your code public), but even with a pseudo-randomly generated, alphanumeric 5-character token, there are 900 million possible links, so finding the right code by accident is rather unlikely.
 
if privacy is a concern, you could always incorporate expiring links
 
There's always the option of simply not saving the code. Unlike Ideone, you'll be able to run something without creating or forking a submission.
 
If privacy is a concern, just add a checkmark labeled "private" for logged in users >_>
 
4:12 AM
Or that.
Authentication still has to be implemented though.
 
Right, but assuming it is, privacy issues should be a bit easier.
You can always leave the option of not logging in, but without some features (like many/most places do).
Plus, how else to show off your impressive array of TIO links if not on a profile page :P
 
@Dennis pls do not forget SSL for TIO if you do make login system >_>
 
That's one of the few things that are already done.
 
I assume you have not deployed https TIO yet?
 
Well, not SSL, but TLS (the good stuff).
 
4:19 AM
;_; SSL was good too
 
@Downgoat It's up, but I'm keeping it private for now. While frontend and backend are partially working, they're still disconnected.
@Downgoat was is the keyword here, since POODLE completely and utterly broke it.
Darn cats...
 
> On December 8, 2014 a variation of the POODLE vulnerability that affected TLS was announced
 
the goat is getting old
 
Not too many servers were affected though
 
Only those who implemented TLS improperly.
 
4:27 AM
@Quill ;_;
 
It's so hard to refrain from posting nonsensical comments on stuff you find on the HNQ list. For me, anyway.
4
 
Why not join the trend? As if there aren't already enough of them
 
¯\_(ツ)_/¯
 
TNQ = HNQ?
 
T=H, N=L, Q=3
 
4:35 AM
Confirmed
 
@1000000000 Hey I've been meaning to talk to you
 
@EamonOlive Yes?
 
@1000000000 What do you think of Brain-flak classic mode?
 
@EamonOlive I haven't looked at it yet
 
@1000000000 You should I would like an opinion
 
4:42 AM
@ConorO'Brien @ASCII-only @ConorO'Brien @Quill @anyoneInterested feedback on cheddar's pattern matching before I forget about it for 3 months and then work on it?
 
my bus stop smells like barbecue sauce
 
@Downgoat why n where n < x not n < x
 
because how do I know whether to use n` or x as the argument variable
@Quill ._.
 
It looks like Haskell
 
@Quill It's probably definitely stolen from Haskell
 
4:45 AM
@Quill that is where i got idea from
 
@Downgoat Argument would just be the one on the left
 
I could assume but that could be difficult to work out
 
Yeah, plus it wouldn't work for modulo
But where is a bit long IMO, some people may want a shorter alternative as well as where
 
CheddarExpressionToken {
    CheddarExpressionTokenAlpha {
        CheddarPropertyToken {
            CheddarAnyLiteral {
                CheddarVariableToken {
@ASCII-only suggestion for syntax maybe?
 
@Downgoat ?
 
4:47 AM
im saying thats how much depth is requires to check if variable in the token list
 
Wow, that's a lot
 
@Downgoat maybe don't prefix everything with Cheddar...
 
;_;
 
@Downgoat I'd go for something like Elegance's n : Number<@ < 2>?
 
for type signatures?
 
4:49 AM
Yeah, super-strictly typed arguments
But that's not easy to do, plus really different to the where syntax
 
can you write up an alternate proposal to detail what you're saying? :3
 
@Downgoat super-strict typing/implicit contracts (contracts being .NET terminology)?
 
Shit... by now I was supposed ot have cheddar v2
@ASCII-only uhhh
no hablo .NET
 
basically the assert in type declaration proposal?
 
Pls explain I have no idea what this syntax is ;_;
 
4:54 AM
@Downgoat what am i saying (which message)
Also we should probably port to bison/jison first
also what is un-proposed generics
 
There is no proposed generic syntax yet
 
5:14 AM
@Dennis Pull Pip, pretty please?
 
Hi
 
@DLosc Pulled.
 
@Dennis Perfect!
 
hello
@Dennis I've been meaning to ask you sth regarding TIO. Is it possible to add sed to it as well?
 
Hello
 
5:36 AM
@seshoumara GNU sed's e is problematic, since it allows executing arbitrary code. I could check if sed works with mbox (the current method of sandboxing), but there's a decent change it won't. As soon as I release TIO v2 (which has proper sandboxing build into it), I'll definitely be able to add it though.
 
I thought about that, plus the r and w commands that can be potentially dangerous, so what if I modify the sed source to remove these 3 commands? The rest is safe.
@Dennis Only very few of my answers require e and other sed people here posting answers in it don't use it either. Anyway, I'm glad you thought of sed for v2, so I can easily wait, no problem.
 
v2 will hopefully be ready soon. There's a lot of languages I want to add (sed, bash, C, Python, etc.). In the meantime, I could try mbox.
 
Oh yeah the input in TIO is bugged with LGC
 
Not sure what to do with that information.
 
@Dennis don't forget to disable :(){ :|:&}, for(;;)fork(); etc
 
5:51 AM
The number of processes is capped.
 
@Dennis Thank you. Check that when you can. I want to let you know that GNU sed also removed the limitation on pattern and hold space max size, so this code :l:s/a/aa/g;tl will consume A LOT of system memory until and if it stops.
 
Between the OOM killer and the time limit for processes, that shouldn't be a problem.
 
great
 
At least in v2. The current incarnation has had a few problems with subprocesses that refused to die.
Two languages in particular create the occasional zombie for reasons I have yet to figure out. v2 will identify processes by context though, so it's much easier to identify and kill them.
 
Do you know my prime checker?
 
5:58 AM
The one you posted on my question?
 
Ya
It doesn't return anything, even with an out at the end
Doesn't even return an error
 
And it works fine locally?
I tried running it and (with Debug checked, so you see STDERR) it shows an error.
  File "/opt/Logicode/logicode.py", line 221, in __getitem__
    return self.parent[key]
KeyError: 'd'
 
I fail to see how that could be TIO's fault.
 
Hang on...
Hmm, it works perfectly well on my side.
 
6:04 AM
Do you use Python 2 or 3? It seems to be designed to work with both.
 
No wait, don't worry
I'm being stupid
 
I tried and it doesn't error with Python 2, but it always prints 0, unless I'm not understanding how to provide input.
 
Hmmm, this is weird
Yeah, this is really weird
It works on my laptop
But it doesn't work on TIO
And it returns 0 and 1
 
Python 2 or 3?
 
Python 2
I have the same code in a .lgc file
and I'm launching it using python logicode.py -f {name of file}
 
6:16 AM
How do you provide input?
 
Uhhh
I just type it in
$ python logicode.py -f adder
Input: 100000
1
 
I tried interactively and I still get the same results: an error with Python 3 and output 0 with Python 2.
$ python /opt/Logicode/logicode.py --file x.lgc
Input: 3
0
$ python /opt/Logicode/logicode.py --file x.lgc
Input: 4
0
For comparison, I tried the primality_tester.lgc from the repo, and it always prints 1.
 
The input is only binary
And Logicode strips down all ASCII chars that aren't 0 or 1
If the input is binp
 
Ah, OK. Is 0 supposed to be true and 1 false?
 
Yeah
 
6:24 AM
Yeah
ninja'd
 
OK, because primality_tester.lgc does it backwards.
Seems to be working with Python 2 then. The output for 1 is incorrect though.
 
Wot u mean
But 1 is both prime and composite...
 
1 is neither prime nor composite.
A prime has exactly two positive divisors, a composite has three or more.
Also mentioned in the challenge btw.
> For the purpose of this challenge, an integer is prime if it has exactly two strictly positive divisors. Note that this excludes 1, who is its only strictly positive divisor.
 
6:41 AM
@seshoumara Seems to be working. sed.tryitonline.net/…
7
 
Hooray!
@Dennis cool example btw
 
:)
 
so, could I use the e command? And is the back-end OS Linux?
 
v1 uses Ubuntu, v2 will run on Fedora. e should work, within the few permissions granted to the sandbox user.
 
Awesome, thank you very much Dennis. Now I can showcase sed as well :)
I was using e for basic commands anyway, like wc, bc, dc and sort.
 
6:46 AM
Yeah, e seems to be working fine. sed.tryitonline.net/…
 
@TimmyD Pretty strict, because it's already a pretty generous bound
 
@Dennis I posted the news on bash, sed and dc chat room as well. For the moment, except me, only Riley is using the room, but we post sed answers regularly. Thanks again.
 
No problem.
 
7:17 AM
Halp I forgot who was making an ASCII-art language apart from DW
 
@Dennis I think a sed process is still running. I tried to kill it, but it just continued. I close the browser tab. It was an inf loop to print sth, testing the limit.
 
@seshoumara It's already killed and should be after 60 seconds or 100 KB of output are generated, whichever happens sooner. TIO currently has incremental output, updating the result every time new output is generated. Unfortunately, that can freeze your browser tab (and will go away in v2).
 
Are the subprocesses monitored as well?
If someone calls some other command from within the main script.
 
@Dennis could you put turtlèd on TIO? github.com/Destructible-Watermelon/turtl-d
 
@DestructibleWatermelon Remind me tomorrow if I forget. It's 3:30 here and I'm going to bed.
@seshoumara Sort of. I usually works.
 
7:33 AM
@Dennis Could you please pull Brachylog?
 
@DestructibleWatermelon oγ
 
7:53 AM
0
Q: "Cannot run program "Rscript": error=2"

user59969after use SparkR,in Linux operations,shows that "Cannot run program "Rscript": error=2",why?wait online

 
if you guys do any image processing, you should read this: blog.johnnovak.net/2016/09/21/…
 
Hello
Is anyone on?
 
sed is on TIO o_O
 
@TùxCräftîñg yes, he was very kind to add it
 
8:07 AM
gtg
 
@Dennis I made this as a token of appreciation sed.tryitonline.net/…
 
8:33 AM
@seshoumara is that a challenge on the site?
I was thinking I might try answering this question with turtlèd:
3
Q: Alexa and the Tea Plantation

twoTimesAgnewHey guys, first time poster here. I went on Coderoulette recently and some guy posted this question. I looked around online but it doesn't seem to be anywhere. I figured I would post it here for anyone who is a fan of these types of problems. Enjoy! Alexa is a druid who loves Mathematics! She ...

hi @TùxCräftîñg
I should probably make a program that parses a string integer into a line of that many chars
 
9:00 AM
@DestructibleWatermelon no, that was just a script I made.
 
make it a challenge pls
 
never thought of that
I'll search the site and if no question has sth like that, then I will make a challenge
 
that tabula recta is similar...
but different challenge still
 
1
Q: List all combinations of possible sums of n from 1 to n

EternityThe challenge is to list all combinations of possible sums of n from 1 to n. For example, assuming n = 4, the result should be: 4 1, 3 3, 1 2, 2 2, 1, 1 1, 2, 1 1, 1, 2 1, 1, 1, 1 Or the same in reverse.

 
9:31 AM
hmmm, I need to figure out how to do better maths with turtlèd
I should really make a less difficult esolang
 
10:00 AM
that moment when the minimum volume on computer is too loud -_-
 
How to insert/overwrite/delete a line at a specific address in sed?
 
@zyabin101 use the line number like so: 3c\new_text or 3i\ or 3d
c is for change, i for insert and d for delete. If you use c and d note that the code after these won't be executed since a new cycle starts. If you don't want that use s/.*/new_text/ and z respectively
 
10:16 AM
@DestructibleWatermelon same happens with phone screen brightness at nights
@seshoumara does 3s/old/new/ work?
 
@zyabin101 yes, it will do the search and replace only on the 3rd line
if you work with a file and not stdin and you want to actually delete or modify sth in that file than use the -i flag, but try making a backup first just in case until you're used to sed
 
@seshoumara im not zyabin101
 
oh, sorry there
I have tunnel vision
 
10:40 AM
--- Code ---
forever
    while 1 do
        until 1 do
            call f(x).
        end
    end
end
--- Tokens ---
[{:id=>:KForever, :value=>"forever"},
 {:id=>:KWhile, :value=>"while"},
 {:id=>:Identifier, :value=>"1"},
 {:id=>:KDo, :value=>"do"},
 {:id=>:KUntil, :value=>"until"},
 {:id=>:Identifier, :value=>"1"},
 {:id=>:KDo, :value=>"do"},
 {:id=>:KCall, :value=>"call"},
 {:id=>:Identifier, :value=>"f"},
 {:id=>:LPar, :value=>"("},
 {:id=>:Identifier, :value=>"x"},
 {:id=>:RPar, :value=>")"},
\o/
 
@TùxCräftîñg is this ur language?
 
it will use prefix notation becuz precedence parsers are hard
 
10:52 AM
--- Code ---
n.
set n 0.
if == n 0 then
    call do0().
or == n 1 then
    call do1().
or == n 2 then
    call do2().
else
    call do_else().
end
--- Tokens ---
[{:id=>:Identifier, :value=>"n"},
 {:id=>:Dot, :value=>"."},
 {:id=>:KSet, :value=>"set"},
 {:id=>:Identifier, :value=>"n"},
 {:id=>:Identifier, :value=>"0"},
 {:id=>:Dot, :value=>"."},
 {:id=>:KIf, :value=>"if"},
 {:id=>:Operator, :value=>"=="},
 {:id=>:Identifier, :value=>"n"},
 {:id=>:Identifier, :value=>"0"},
 {:id=>:KThen, :value=>"then"},
now just to work on the compiler
 
@TùxCräftîñg Why or instead of the clear elif, else if or sorts?
 
becuz it's more natural i think
--- Code ---
a.b.
set a [b 1].
--- Tokens ---
[{:id=>:Identifier, :value=>"a"},
 {:id=>:Dot, :value=>"."},
 {:id=>:Identifier, :value=>"b"},
 {:id=>:Dot, :value=>"."},
 {:id=>:KSet, :value=>"set"},
 {:id=>:Identifier, :value=>"a"},
 {:id=>:LBra, :value=>"["},
 {:id=>:Identifier, :value=>"b"},
 {:id=>:Identifier, :value=>"1"},
 {:id=>:RBra, :value=>"]"},
 {:id=>:Dot, :value=>"."}]
--- AST ---
[{:id=>:Declare, :type=>{:id=>:Type, :type=>"s32"}, :var=>"a"},
 {:id=>:Declare, :type=>{:id=>:Type, :type=>"s32"}, :var=>"b"},
assignements are ok
 
11:07 AM
Should I make a webpage (or IDE) for Logicode?
Is anyone on?
At all?
 
Anyone is on.
At all.
 
11:23 AM
Guys, you know how you can do any boolean operation with just NAND?
We need to make a CPU that only does NAND
And then make an OS for it called NANDOS
</joke>
 
There are single instruction compilers.
mov/add/sub/xchg etc. are all turing complete
 
Hello
 
Hello
I wrote a FRACTRAN answer: codegolf.stackexchange.com/a/94169/47581
 
0
A: Sandbox for Proposed Challenges

bwoebiAdd binary numbers without math operation codegolf restricted-source math Inspired by my own Stackoverflow question: Match correct addition of two binary numbers with PCRE regex Given an input in form of (?<a>[01]+)\s*\+\s*(?<b>[01]+)\s*=\s*(?<c>[01]+) (i.e. given three binary numbers a, b and...

 
11:46 AM
^ trivial
 
Not trivial in Logicode...
Because I have to parse everything
Argh how would I parse things
 
Can I have a little advice on Machine Learning algorithms?
Does anyone here actually know about Machine Learning algorithms?
 
Wot machine learning?
@ASCII-only You're on!
 
@Qwerp-Derp Machine Learning is where a machine learns.
 
Ah yeah
 
11:58 AM
A Machine Learning algorithm was used to create this bot several chatbots with whom I may or may not be personally acquainted.
 
well... i know some ANN thingy
 
Wait what
 
artificial neural network
 
@Qwerp-Derp ide? no, webpage? what do you mean, there's not much point doing anything other than TIO
 
I have a lot of input parameters and one output parameter. I want the algorithm to optimise the input parameters for the highest output value. How can I do it?
 
11:59 AM
Ah fine.
@ASCII-only I've got an idea for one-line circs.
 
huh idk
 
@TùxCräftîñg I have had a look at FANN, but I don't know how to use it for that.
 
Crap it seems like all ASCII-only and I talk about is Logicode now
 
@Qwerp-Derp Haha, yeah
 
It says it's simple, but I suppose that is subjective.
@Qwerp-Derp Are you a bot?
 
12:01 PM
@wizzwizz4 That's up to you.
Shroud of mysteriousness activates
 
@TùxCräftîñg 0/10 shunting yard is really easy
 
@Downgoat Pls implement inputting for TIO in your plugin
 
duckduckgoing shunting yard
 
In computer science, the shunting-yard algorithm is a method for parsing mathematical expressions specified in infix notation. It can be used to produce either a postfix notation string, also known as Reverse Polish notation (RPN), or an abstract syntax tree (AST). The algorithm was invented by Edsger Dijkstra and named the "shunting yard" algorithm because its operation resembles that of a railroad shunting yard. Dijkstra first described the Shunting Yard Algorithm in the Mathematisch Centrum report MR 34/61. Like the evaluation of RPN, the shunting yard algorithm is stack-based. Infix expressions...
^ Done.
 
@zyabin101 0/10 not DDG
@TùxCräftîñg I have implementations in JS and C++ if that's helpful in any way
 
12:03 PM
I won't do a service to google things for others, especially with lmgtfy.com.
 
@Qwerp-Derp Yes, what? I am not sure. Yes.
 
@ASCII-only Should I put the suggestion on GH? Or should I just blurt it out right now?
 
wow it look easy o_O
 
@Qwerp-Derp Both? I'd put it on GH in case it's not implemented at the moment
 
Ah
So
 
12:04 PM
but i dont want to rewrite the parser :P
 
circ asdf(n)->cond n->1/0
asdf(1) should return 1
and asdf(0) should return 0
 
@TùxCräftîñg You just need a shunting yard function and pass the expression into the shunting yard
 
It's a beautiful one-liner
And would save many, many bytes
 
@Qwerp-Derp If you want to save more bytes you could shorten all the keywords wo 1 char
 
But that would be boring
But it's still a beautiful (and intuitive) one-liner
So yea or nay?
 
12:06 PM
Yeah, okay
 
How hard would it be to implement?
 
idk, starting up c9
 
--- Code ---
func main(s32 argc, ptr ptr u8 argv) to s32
end
--- Tokens ---
...
--- AST ---
...
--- C ---
typedef s8 signed char;
typedef u8 unsigned char;
typedef s16 signed short;
typedef u16 unsigned short;
typedef s32 signed long;
typedef u32 unsigned long;
s32 main(){}
it half works
 
signed long main() wat
 
TODO: add typedefs to define hell :3
 
12:09 PM
@betseg int
@zyabin101 liek int other
 
Naww, only 107 views and 5 UAVs on define hell ;_
And only one stargazer: me ;_
 
--- Code ---
func main(s32 argc, ptr ptr u8 argv) to s32
end
--- Tokens ---
...
--- AST ---
...
--- C ---
typedef s8 signed char;
typedef u8 unsigned char;
typedef s16 signed short;
typedef u16 unsigned short;
typedef s32 signed long;
typedef u32 unsigned long;
s32 main(s32 argc,**u8 argv){}
yay
 
@TùxCräftîñg What is that
Please don't tell me it's your new lang
You've got like
>9000
 
ok i will not tell you
 
It's his new lang.
@TùxCräftîñg I told.
 
12:13 PM
:(
Too many langs
Too little time
 
@TùxCräftîñg y u use long and no int
 
becuz long is always 32 bit but int can be anything
 
32 or 64, so it's the same or better
 
a 64-bit type named s32...
 
Btw, u8**argv instead of **u8 argv
 
12:17 PM
ahye
 
@TùxCräftîñg
 
Let he design langs
ill code in em
 
True
We need more pictomemes in chat
 
@Qwerp-Derp ಠ_ಠ
 
@TùxCräftîñg ಠ___________ಠ
 
12:20 PM
@Qwerp-Derp ಠ__________________________________________ಠ
 
Wow u like Kannada language?
 
>_<
 
@betseg ಠ_ಠ
 
disapproval faces don't solve problems, they are like a hydra, make lots of other problems
 
ಠ_ಠಠ_ಠಠ_ಠ
 
12:21 PM
@TùxCräftîñg ಠ(_*)ಠ
Regexified ಠ_ಠ
 
ಠ_(100000)$*ಠ++>
straw code
you need a lot of memory tough
 
s/[Oo]/ಠ/g
 
I should give define hell to the mass. :I
 
--- Code ---
func main(s32 argc, ptr ptr u8 argv) to s32
    a.b.
    set a [b 10].
end
--- Tokens ---
much tokens
--- AST ---
such ast
--- C ---
typedef s8 signed char;
typedef u8 unsigned char;
typedef s16 signed short;
typedef u16 unsigned short;
typedef s32 signed long;
typedef u32 unsigned long;
s32 main(s32 argc,u8** argv){s32 a;s32 b;a=(b=10);}
 
@zyabin101 maybe char* and char[] to string; and int* and int[] to array?
@TùxCräftîñg you don't need the parentheses
 
12:25 PM
precedence problems
 
Good idea, make an Issue and/or Pull Request and I'll put them in the project.
 
I'm outside, I've just had a flushot, this'll have to w8
 
--- Code ---
func main(s32 argc, ptr ptr u8 argv) to s32
    n.
    * + [n 10] 20 30.
end
--- Tokens ---
( ͡°tok ͜ʖens ͡°)
--- AST ---
ಠastಠ
--- C ---
typedef s8 signed char;
typedef u8 unsigned char;
typedef s16 signed short;
typedef u16 unsigned short;
typedef s32 signed long;
typedef u32 unsigned long;
s32 main(s32 argc,u8** argv){s32 n;(((n=10)+20)*30);}
 
You can just strip the tokens and AST with a compiler flag to implement.
 
@zyabin101 pred
 
12:37 PM
@zyabin101 you didn't assign me
 
I will now.
@betseg I'll send the link in a privnote, the password is opendefhell
 
how to add a commit to a pr?
 
@TùxCräftîñg if the PR is open, the commits will be added automatically.
Just push to the correct branch.
@betseg ^^^
 
here, it's closed ._.
 
12:52 PM
Woah
 
@betseg did you read the privnote?
@TùxCräftîñg Uhh, I'll check the fork directly ._.
 
i pushed a fix
 
The link there is user specific
i.e. tux wouldn't see an invite there
 
^^
 
12:54 PM
So why privnote
 
Please, no more huge xkcd oneboxes for now.
2
 
-._(._.)_.-
^^
@TùxCräftîñg strange, the fix didn't appear in the PR's discourse page.
 
Now I see it :3
 
12:56 PM
@betseg The bottles produced vs bottles recycled makes me sad.
 
@TimmyD yeah :/
 
@betseg now, did you accept the invite?
Yes!
 
Yup
 
Now, you should learn about GitHub's Pull Request Review.
 
...i do
 

« first day (2063 days earlier)      last day (2783 days later) »