« first day (1492 days earlier)      last day (3352 days later) » 

1:34 AM
I'm installing an Arch Linux VM right now.
not as retro, but pretty techy
 
 
2 hours later…
3:45 AM
Arch is nice, my 2nd favorite distro
Quite annoying that they switched to systemd though :(
 
 
2 hours later…
5:17 AM
@TheBestOne It's trying to tell you to switch to Vim :D
 
 
3 hours later…
7:59 AM
@aditsu but its super easy
the code is already there in the ticket
 
8:31 AM
@Nabb Must've been years since you were last on or something?
 
8:49 AM
@Sp3000 I actually drop by and read things sometimes, just don't comment/answer nowadays
 
Ahaha I see :)
 
 
2 hours later…
10:27 AM
@Optimizer that code covers maybe 3% of what I want to do, and is not usable directly
 
10:55 AM
@aditsu why ?
what else is there to regex ?
I think that is a good start and covers 90% of the use cases
I am not sure how you are quantifying it as 3%, but its definitely much more than that.
 
11:08 AM
Moreover, you should think more on terms of what users of the language want, rather than what you wish to see for the language and then compromise/improvise :)
 
u n me
and SP3000
and so many more trying out cjam.
 
As much as I like regex I wouldn't force aditsu to compromise anything though
Implementing new features takes a fair few design choices/iterations, I'd imagine
 
I still dont see what is being compromised in that code ..
 
@Optimizer and who says Sp3000 and I want the same regex features for the language as you do? :P
maybe I agree with aditsu?
 
11:14 AM
what else is possible ?
 
regex replacement?
regex split?
 
Overlapping matches?
 
just check the issue tracker for Retina :P
 
2 things
what else ?
 
If you want to get even more golfy, maybe there should be a charset for vowels/consonants?
 
11:15 AM
I want to see how its 3%
 
Does Java have recursive regexes?
(not that I necessarily think CJam should have any of these, but they're things that could be considered)
 
even if it had, GWT wont support it because JS does not have
 
@Sp3000 no
it doesn't really have any fancy features, except sort-of variable-length lookbehind due to a bug
 
@Sp3000 iterations is precisely what leads to landing of basic feature first and then extending the feature in next iterations
 
@Optimizer that will be tricky... because I don't think Java can match ECMAScript flavour
 
11:19 AM
I don't know how GWT does it
but it has to use JS based regex
 
so unless aditsu reimplements ECMAScript regex in Java, it will be impossible to make the online interpreter and the Java interpreter use the same regex flavour
 
as I said, maybe GWT does something . I am not sure
 
 
1 hour later…
12:43 PM
I found an eval-free quine in CJam that's only one byte longer than the standard quines
"_`\\"_`\
actually, I can do it in 8 bytes
"_`o"_`o
o.O ... this can be done in 7
"_p"
_p
@aditsu new minimal non-trivial quine? ^
this also works in GolfScript (with a trailing newline)
 
I think this is on the same lines of {"asd"}
making use of toString() and printing it.
 
umm
less so than any other quine we've been using
at least it uses p once instead of relying on autoprinting at the end
and for 9 bytes I can get rid of any autoprinting at all (just add two os)
in any case, this is closer to a normal quine than any of the others, because it doesn't use either stringification of a block, nor does it use eval.
 
Damn ><> using defaultdicts :/
 
@MartinBüttner cool
 
12:59 PM
it uses stringification of a string. I am not sure how you rank stringification of a block farther away from a normal quine as compared to stringification of a string (or use of eval)
 
most of the standard quines use stringification of a string (usually with format strings)
also I don't know a CJam quine that doesn't use stringification
and stringification of a string seems a lot more legit than stringification of a block
 
on what grounds ..
 
because that's how quines work...
stringifying a block is a lot more like printing the source code directly
 
do you have a link which says so ?
 
what part?
I can link you to the former
 
1:03 PM
ok
 
> So to make this possible anyway, we write the build the program from two parts, one which call the code and one which we call the data. The data represents (the textual form of) the code, and it is derived in an algorithmic way from it (mostly, by putting quotation marks around it, but sometimes in a slightly more complicated way).
also, see the C example in the section above it
 
 
1 hour later…
2:05 PM
I like how this question is one where Pyth loses easily to CJam, due to the lack of a lower function
 
Which question?
 
he… it's funny
I saw @ngn's most recent answer where he claims to got help by Roger Hui
Which is weird because Roger Hui is a very famous APL specialist and designer of J.
I just found out @ngn works for Dyalog, the same company Roger Hui works for.
Now all of that makes sense…
 
2:36 PM
@MartinBüttner Prelude golfing again, I see :P
 
only a little bit... :D
thanks for the bounty :)
 
hm...
Is there a better way in J to get the minimum of magnitudes than (>&| { ,)"0
 
@Sp3000 else is annoying in Prelude
 
it's important that it works on arrays and not just on scalars, thus rank
 
@MartinBüttner I'd do the alphabetical question in Prelude/BF/Pancake stack if one of them actually had a comparison operator
That's annoying too :P
 
2:40 PM
haha, yeah that's exactly what I thought
 
I was almost going to Piet, but ... Piet.
 
if I ever write a better Prelude it'll allow ( and ) on the same column for if/else constructs
or maybe just |
 
2%|2/
  |3*1+
Like that?
 
Wouldn't that make it a while-else statement instead of an if-else statement?
 
@Sp3000 one of them is enough
@PhiNotPi yeah, it would be a loopy if/else
the else would also loop
@Sp3000 oh wait, I misinterpreted your example
that looks interesting, but that' not really Prelude like
I was thinking more like:
2%(3*1+0|2/0)#
 
2:54 PM
Oh, right
 
but you could pull the else into another voice if you wanted
so by using multiple | on different voices you get (loopy) else-if as well
but in vanilla prelude if/else or even if (!...) is really annoying, and you need at least two voices for it
I wonder if it would be useful for a golfing language (like CJam) to be able to switch between stack and queue modes
 
I think prelude would benefit from both while-else-while and if-else-if constructs.
 
It might be more useful to just have a reverse stack like ><>
 
@PhiNotPi for golfing purposes yes. for slightly increased ease of use, while-else-while seems enough
 
An interpretation of multiple )s or |s on the same line could be a logical ANDs.
 
3:02 PM
@Sp3000 I think I'm more generally wondering how well a queue-based golfing language would work
@PhiNotPi yes, I've been thinking about that for ages
AND or OR?
likewise, you could make the position of the ) significant, but it's good for golfing to be able to place it anywhere
 
3:36 PM
Installing PC-DOS v1.10
Once I heard that PC-DOS v1.10 was floppy-only and couldn't even be installed on a hard drive, I was sold.
 
4:28 PM
you get sold on weird things
 
4:50 PM
I'm trying to update IE on Win2000. Any tips?
 
@PhiNotPi Set yourself on fire, I believe that's less painful.
 
The problem is that IE 5.5 (?) seems to break on any page with any real functionality.
 
Argh… my Paeth-challenge verifier has a bug
argh argh argh
 
Is it pirate day?
 
no. I'm just unhappy with the situation.
 
4:58 PM
I'm trying to install Guest Additions.
 
I remember I worked on a project for a bank once, and they were having problems with it; I went to visit them, and found out that their systems guy had a very strict policy of using only IE3 on all computers (while IE5 was already out at that time)
had to modify some javascript and make it even clunkier :p
 
I managed to load the PPCG website, but the login page is broken.
 
hmm I wonder if I have an ie6 installer somewhere :p
but it would probably take me too long to find it
 
5:17 PM
I have guest additions working, so I can transfer files in and out of the VM
I'm downloading a browser outside of the VM and transferring it inside.
 
5:41 PM
0
A: Paeth-transform arrays

FUZxxlJ, 102 characters Let's get this started. Notice the absent exit; J prints a prompt (three blanks) after all the output has been written, but then gets an EOF on stdin, causing it to terminate. I believe that quite a couple of improvements are possible in this code. 1!:2&4":(|.@$,,)256|(-p+[:(>&...

 
5:56 PM
Well, I deleted all of my Windows VMs.
 
Could none of them read punchcards?
 
maybe he was not sold enough
 
6:10 PM
If I wanted to make an operating system, how could I do it?
 
52
Q: How do you write a basic operating system?

Agusti-NLet me know how I can create a basic operating system. What books do you recommend?

 
I wasn't serious, but now I might be.
 
6:30 PM
Having read the articles, I am no longer serious.
 
 
1 hour later…
7:41 PM
holy crap, where are those votes for the sorted word challenge coming from... o.O
 
from the answerers I guess
if I would post a sum(a+b) challenege and managed to not get it closed I could get a lot of upvotes..
 
yay, 222 bronze pixels :)
 
what?
 
@randomra oh I didn't mean "who on earth upvotes that question"
I meant upvotes for answers on that challenge
 
well people can understand CJam and such easily here which helps
I will post this
2
A: Sandbox for Proposed Challenges

randomraAnalyzing Collatz-like sequences code-golfnumbersequencemath We define a Collatz-like sequence s with 4 positive integers: n starting value d > 1 divisor m > 1 multiplier i increment (In the original Collatz sequence d = 2 m = 3 and i = 1.) Given these integers s will be created in the fol...

is the output spec clear?
 
7:49 PM
looks good to me
 
What do you think?
 
@MartinBüttner great, thanks
 
8:08 PM
@FUZxxl I don't think I will analyze it now.
 
@randomra Ok :-)
 
hi all
 
hi @Lembik!
 
@FUZxxl how things?
 
@Lembik thingish
 
8:17 PM
I was thinking about adding a bounty to codegolf.stackexchange.com/questions/47163/… .... or should I modify the question first and wait a bit?
 
"modify the question" sounds like a bad idea
what did you have in mind?
 
@MartinBüttner I meant if people can see something wrong with it which is putting people off
I could for example allow people not to actually compute the optimal answer and still get some score
 
nah it looks fine to me
 
@MartinBüttner ok thanks..now the usual question.. how large a bounty? :)
 
Man… I should really study.
 
8:20 PM
depends on how much it's worth to you :P
 
@Lembik Over 9000
 
@MartinBüttner well it's all just for fun but I do love it when I get great answers :)
@FUZxxl how hard would you work on it for 9000?
 
@Lembik All the time.
 
interesting :)
 
until it's done and golf'd of course.
 
8:21 PM
and the more important question: where is Lembik gonna take that kind of rep from? :P
 
actually .. do you find the question interesting at all?
 
It's interesting but I have no idea how to solve this efficiently.
 
I was thinking of asking a pure golf question about computing the longest common substring in linear time. Is it still code golf when you specify a time complexity?
or does it then become a code challenge
 
It's okay to specify a time complexity
but the complexity shouldn't be über hard to achieve.
 
@FUZxxl ok that's cool.. the other thing I am not clear on is libraries
 
8:24 PM
@Lembik if you just bound the complexity it's code golf
when you start pulling the complexity into the winning criterion, it becomes a code challenge
 
I would like to say you can use any library that you can get using apt-get install on a standard ubuntu intall
 
@Lembik General rules forbid plugging in a library that does (almost) the whole task.
 
but i realise some people don't use linux :)
@FUZxxl right but I also don't want to disadvantage the C people who need libraries to do anything
so any library you can install from a default repo on ubuntu would be fine
 
@Lembik The trick is to avoid things that are cumbersome to do and not really part of the challenge.
 
@FUZxxl what do you have in mind?
 
8:26 PM
E.g. if you do a challenge that involves generating images, do not specify an image format.
 
ah ok.. got you
 
or specify something like PNM which is very easy to generate.
 
how do you get people to answer in different languages? I would like to pose a graphical challenge so that we can see the shortest code in many languages
I think that could be really interesting
is there a rosetta tag?
 
include a by-language leaderboard
 
or something similar
 
8:28 PM
the rosetta tag is for a single person answering in as many languages as possible
 
ah ok
 
and of course, try to be as liberal as possible with the what you allow for graphical output
screen, file, raster, vector, ...
 
codegolf.stackexchange.com/questions/10656/… .. basically that but for code golf where I really want answers in different languages
@MartinBüttner I am interesting in golf'ed interactive graphics...like in that question
maybe pong would be better
maybe I should start with just drawing a line on the screen :)
god knows what the shortest C code is for that, for example
what is the simplest possible interactive graphical game?
 
@Lembik putc('|');
 
not graphical enough!
 
8:33 PM
;)
 
or interactive enough :)
there used to be a skiing game where each line had a black square , a few grey ones then a black one. Consecutive lines could be shifted right or left and you have a little icon you had to navigate right and left so it didn't touch the borders :)
it might have been horace goes skiing for the ZX81 :)
 
9:07 PM
0
A: Sandbox for Proposed Challenges

Martin BüttnerPrelude Syntax-Checker code-golfstringsyntaxprelude Prelude is an esoteric programming language, which has very few, but unusual, restrictions on what constitutes a valid program. Any block of printable ASCII text is valid provided that: Every column of text contains at most one of ( and ). I...

 
@FUZxxl thanks! :)
 
:-)
3
A: Random Golf of the Day #3: Integer Partitions

ngnDyalog APL, 67 59 51 bytes p←{⍵,⊂1,⍨+/¨⌽⍵↑¨⍨⌽⍳⍴⍵}⍣⎕⊢⍬⋄f←{⍵=0:⍬⋄a,a∇⍵-a←{1++/(?+/⍵)>+\⍵}⍺↑⍵⊃p}⍨ (67 bytes) p is a vector of vectors in which p[n][k] is the number of partitions of n into k summands, or equivalently: the number of partitions with greatest summand k. We build p by starting with t...

That's a neat strategy to get a job.
 
heh, nice
SE used to handpick their employees from the SO user base
 
used to?
Last time I checked they didn't hire Jon Skeet
 
maybe Jon didn't want to be hired? :P
 
9:16 PM
maybe...
 
10:00 PM
hmm, if a question says the input will be a string than e.g. in J you can expect a 1-char string to be inputted as 1$'a' because 'a' wouldn't be a string just a char
is this the common practice?
 
I think you can assume that.
Would be better if you didn't have to, though.
 
10:29 PM
weird that you can't compare characters in J but you can sort them
 
it's funny, isn't it?
I dislike this, too.
 
maybe the simplest compare is with sorting :D
'a' (,-:/:~@,) 'b'
 
hehe
 
The Fortnightly Challenge spec now contains over 10000 characters.
It's going to be epic.
 
I'm not sure if I should be impressed or frightened
 
10:41 PM
I don't know how much you've been following it, but the plan right now it to create a large set of example problems that the 2D pattern-matching language should be able to match.
 
o.O
linky?
 
And the entrant creates a language and implements it, and then solves as many examples as possible to demonstrate the language.
 
ok
 
1
A: Sandbox for Proposed Challenges

Martin BüttnerFortnightly Challenge #6 - Language Design Join us in the Fortnightly Challenge Chat to work out the details of this challenge! 2-D Pattern Matching popularity-contestlanguage-designgridstring And now for something completely different... In this challenge, you will be designing a language! ...

 
I'm already preparing my answer :)
 
10:44 PM
@TheBestOne You are the reason why I don't post my challenge to the sandbox first.
4
 
The parser will be the hardest part :/
 

« first day (1492 days earlier)      last day (3352 days later) »