« first day (1506 days earlier)      last day (3340 days later) » 
00:00 - 19:0019:00 - 23:00

7:00 PM
if you make TEST a static inner class, it will probably make a lot more sense
non-static ones are a bit trickier
 
I was really hoping for a direct translation of my C/C++ code, but I'm not comfortable with having these file naming conventions and having all of the procedural code wrapped in an object.. yet.
I'll do some more reading, thanks for the help
 
make TEST static, trust me :) also you're welcome to discuss more on irc (##friendlyjava on freenode)
 
hi.. I would like to copy codegolf.stackexchange.com/questions/47759/… but change it to something that can fit in a tweet which outputs consecutive digits of pi forever
 
okay, I think I have something in 63 bytes for Pith
 
is that acceptable? It seems a little like plagiarism
 
7:06 PM
@aditsu thanks, I'll probably join there later. I want to get this running tonight. By the way, is there anything anologous to JSFiddle for Java?
 
if you want it to run the code too... maybe ideone?
 
@Lembik I guess that's not Monte Carlo so isn't the only common part is pi?
 
done :)
 
I mainly ask in case I would like to show someone code later, it's more polite than spamming. I guess as long as it's not running, pastebin would work
 
@randomra I quite like this challenge .. I hope other people do too
 
7:09 PM
shortest tweetable code is an alias for code-golf without java submissions?
 
@randomra why not java? Because it won't fit?
the tweeting part is basically irrelevant..it's just there for fun
 
@Lembik "must never terminate" - I'm pretty sure any such program will either terminate after a finite time or start taking longer time per digit
 
@Lembik what about built-ins?
 
@Lembik yep
 
and most likely both
 
7:11 PM
@MartinBüttner good point.. I didn't think that some languages might have infinite precision.. editted
@aditsu I am not sure that is right
 
@Lembik you should also disallow trigonometric functions then
 
@MartinBüttner infinite precision trig functions!
argh :)
 
I have a pi challenge in mind too
I'll sandbox it
 
are we allowed to print newlines between the digits?
 
7:13 PM
@EricTressler very cool!
 
Borwein's algorithm is great because it looks like it uses magic numbers, but it doesn't
 
seems a very good candidate
 
the numbers are way too long...
 
@EricTressler am I right in thinking that computing the digits of pi doesn't have to slow down?
 
look at some lower examples on that page
There are recurrences for pi that quadruple or even more the number of correct digits each iteration
each iteration also takes longer than the last, of course, but...
 
7:16 PM
ok I updated it
 
there are various google results about "nearly linear time" and "quasi-linear time"
 
This is code golf, so the shortest submission (in bytes) wins except that it must output at least one digit per second on a reasonable modern PC for the first 1000 digits and it must never terminate.
 
ugh, I've got 15 minutes for my presentation and the first attempt just took 22...
I hate it when that happens
 
@Sp3000 an empty 1() counts as valid ?
 
heh, you can easily get 1000 digits in a second :)
 
7:18 PM
@aditsu good :) But your code shouldn't stop then
 
@MartinBüttner you need to golf it
 
@MartinBüttner man! golf it!
too long as is right now
 
yeah, even with a basic series. Monte Carlo approximations for pi have bad convergence properties
 
ok last edit done :)
 
7:19 PM
damn, I forgot to take a screenshot yesterday
 
@Lembik yeah, now go find one for base 10 :p
 
8 mins ago, by Martin Büttner
are we allowed to print newlines between the digits?
 
@aditsu :)
 
@MartinBüttner ^
 
7:23 PM
lol not bad
 
Huh, cool, looks like SE adjusts the number format to your region, because I see the rep on the profile pages with a comma and not with a period.
 
that's basic I guess
;)
 
perhaps, but I never noticed it before :p
 
you are not srsly thinking that's a dot, right ?
 
... oh, wait
facepalm
 
7:26 PM
@Lembik yeah, the binary spigot algorithm is neat; too bad it's more expensive to convert to decimal than it is to calculate the digits directly. Also, nobody knows if spigot algorithms for other bases exist.
 
I really saw that as a dot...
 
:D
 
@EricTressler that is very interesting
 
Its just cropped
 
@MartinBüttner how would you use built in Pi to help with my challenge? I just realised I don't know
 
7:29 PM
an arbitrary-precision built-in pi could help :p
I expect Mathematica to have it
 
@aditsu how would you actually use it? Compute pi up to 1000 digits, output one at a time, compute pi up to 2000 digits, output digits 1001-2000 etc?
 
or you could do it in bigger batches
oh wait, that's kinda what you wrote
 
@aditsu My limit of 10 seconds is the only part I am still not sure about
@aditsu do you think it is too generous?
maybe I should have made it 10,000 digits
hmm....
 
maybe, but it doesn't have to be changed
it weeds out inefficient algorithms anyway
 
I'll change it to 10,000 :)
and then stop editting
@aditsu do you think you will try it?
 
7:38 PM
not sure, I'll think about it
 
hmm.. my first criticism :(
someone who likes speed-golf :)
-1 !!
oh dear
 
@Lembik you can ask Mathematica to give you the nth digit of pi
 
@MartinBüttner oh! If you do one after another is it quick?
 
of course not
but that was no requirement
 
@MartinBüttner well it has to be not very slow :)
that is a requirement
 
7:45 PM
oh it was
I overlooked that
but I'm sure it can meet that requirement
and yeah if not then batches definitely work
 
one of the commenters wants " display as many Pi digits as possible in a tweet (140 chars) AND in less than 10s".. I am tempted to ask that for E
is that too similar to this one?
I don't want any more negative rep!
 
@Lembik I don't see the difference
 
@MartinBüttner it's a fastest-code challenge now with a length limit
it is optimising speed with a code limit as opposed to optimising code length with a speed limit
 
oh I see
 
@Lembik By E do you mean e (the base of the natural logarithm)? If so, it's a much less interesting number to compute than pi. Every answer will just use the Taylor expansion.
 
7:58 PM
@PeterTaylor I did yes. I won't do it in that case.. I hope I get above -1 for my question :)
@PeterTaylor numbers.computation.free.fr/Constants/E/e.html has some different methods in a slightly broken looking web page
if you look at numbers.computation.free.fr/Constants/E/e.ps , section 6.2 is amazing!
 
8:14 PM
I don't really know that it's amazing... the square in the one is pretty significant
Here's a constant that's more interesting to try to compute, maybe
 
@EricTressler that's up my street too :)
 
That one's notable because the hard hexagonal entropy constant is known to be algebraic, and there's really no evidence that I know of regarding whether the hard square entropy constant is or not
so it's either really expensive or really cheap to compute, nobody knows.
 
hmm, my challenge got some interesting comments: codegolf.stackexchange.com/q/47801/7416
 
Isn't it "Lena"?
 
I have to admit I agree..it's not really appropriate
 
8:21 PM
I'm pretty sure it's double n
 
"Lenna or Lena is the name given to a standard test image..."
 
well I only really agree about Lenna
it is basically part of a porn shoot
 
i know what it is
oh, you were talking to aditsu
 
yes :)
to the extent that PPCG wants to be a grown up place for the public to take part in challenges
parts of porn shoots don't seem appropriate
 
@aditsu not pullign in any feminism here, but all three pics of girls is a bit partial. We have had a lot of reference images here on many questions and they inlcude all sorts of images, not just all girls
 
8:25 PM
just women is OK but maybe not just images called things like "blueeyes" :)
 
alright, looks like most would agree to replace Lenna, any suggestion of an old gif?
 
researchers used that image for years. it's basically the stanford bunny of reference images. personally, i don't think the cropped photo should offend anyone, but i don't like to argue with people
uh.. i would just go with wikipedia animals or something
because of the licensing
 
@EricTressler I think the issue is the word "offend" ... It is used too broadly. I am not offended but I still don't think it is appropriate
 
fair enough. in any case, i would just take a few cute pictures from wikipedia
 
8:28 PM
is their license compatible with CC-BY-SA 3?
 
google images has a tool for this
to search by rights status
 
I would prefer images that circulated in the 90's (ideally in gif format)
 
I vote for Bosch as new reference image
 
^ has more nudity than most of the A rated Hollywood movies out there
 
A standard test image is a digital image file used across different institutions to test image processing and image compression algorithms. By using the same standard test images, different labs are able to compare results, both visually and quantitatively. The images are in many cases chosen to represent natural or typical images that a class of processing techniques would need to deal with. Other test images are chosen because they present a range of challenges to image reconstruction algorithms, such as the reproduction of fine detail and textures, sharp transitions and edges, and uniform regions...
Which contains Lena/Lenna, btw
but also lots of more neutral stuff
@aditsu I'm pretty sure this is exactly what you want
 
8:38 PM
@EricTressler oh cool, I'll check it out, thanks
 
I think what I linked will involve less hassle
 
I think I'll use the peppers one from that page
 
9:28 PM
done, I hope that is satisfactory
 
9:47 PM
I wonder if it is possible to estimate the percentage of valid Pith programs.
(Pith, as in the recent challenge about it. It's not a real language.)
 
0
A: Sandbox for Proposed Challenges

aditsuCalculate π using the Gauss–Legendre algorithm Write a program that takes a positive number n and performs n steps of the Gauss–Legendre algorithm, then prints out 5*2n-1-2 digits of π (including the beginning 3). Examples: input 1 → output 3.14 input 2 → output 3.1415926 input 4 → output 3.14...

 
Or, perhaps, the "golfability" of the language, which I will define somehow.
 
Well a naive approach to estimate it would be to just generate all possible programs up to length n, then check each one and divide the number of those that work by n. By doing this for many small n we may make a somewhat accurate prediction of bigger n.
 
Perhaps (number of valid, functionally unique Pith programs of length N) ^ (1/N)
 
@aditsu I wouldn't see a problem with all women, but I suspect that "all female models except for one token unrelated image" is just as likely to seem alienating
 
9:56 PM
I don't think it's a strong point when "all except for one" amounts to 2. Also, the Len[n]a image was more problematic because of the Playboy origin.
 
@PhiNotPi I think technically this also depends on how many of these programs are functionally equivalent (which is not defined for Pith)
 
Yeah, it's not actually defined.
I could just say "valid" and make it easier.
To determine functional equivalence, the commutative/associative/etc. properties would have to be known for each operator.
 
10:30 PM
@Optimizer 1() is valid, yes
 
ok cool
although, can't seem to golf it below 60 bytes :(
 
:(
@MartinBüttner Yeah, probably do. I'll might add some more cases later when I get time
@Optimizer "and add lots of ) to compensate for unmatched (". Heh.
 
10:52 PM
somewhat related to the image choice issue, I thought this was pretty funny: notalwaysright.com/not-a-picture-perfect-way-to-advertise/42080
 
0
Q: Gender of example images

trichoplaxThere was some discussion in chat and in the comments on this question about the choice of example images. Initially the images were criticised for being exclusively of women. The image of Lena Söderberg also had its origin highlighted. The question has since been edited to replace this image wit...

2
 
00:00 - 19:0019:00 - 23:00

« first day (1506 days earlier)      last day (3340 days later) »