« first day (1391 days earlier)      last day (3467 days later) » 

12:17 AM
@TwiNight Have you tried ChatSEy?
 
@TwiNight You can chat on the mobile version of the website if that helps?
 
The mobile chat website is terrible, though :P
 
@Geobits indeed it is :) Glad to see you have a dedicated app.
 
12:29 AM
@Geobits Thanks
 
 
2 hours later…
2:08 AM
i had a crazy idea
what if we had a "Tips for Writing a Code Golf Question" as a tips page?
analogous to "Tip for Golfing in X", but for question writers instead?
the help page only covers the very basics, IMO
of how to write a valid question, but not how to write a good one
 
Sounds like it'd be better as a meta question, from which we can put the better answers into the help page
 
"Tips for writing meta proposals" - First question on the the meta-meta.
2
 
"Include your opinion as a separate post on your meta discussion question. That way, people can downvote your views while upvoting the request to discuss the issue."
 
"Resign yourself to disappointment in the final decision, no matter which answer is chosen."
 
"Downvotes on meta don't decrease your rep, but they decrease your self-esteem twice as much."
 
2:17 AM
"Upvotes, however, are as empty and fake as Twinkies."
My god, the chat's full of stars!
3
 
@FryAmTheEggman "Tips for writing meta-meta proposals"
 
"Write your meta-meta proposals in chat, so you can be rewarded with stars."
5
 
Rejected for being too meta
 
0
Q: Tips for writing good code golf questions

xnorI think the code-golf tag wiki doesn't have enough on writing a good code golf question. I'd like to collect advice on these, and perhaps the most well-liked can be incorporated. What makes a code-golf question good? What are some pitfalls to avoid?

 
2:42 AM
Trying to think of code-golf questions with weird scoring methods, and the only one the comes to mind is Martin's meta regex golf one
(aside from bonuses being awarded)
"your code in number of bytes * 5 + sum of the 10 generated BF codes"
Oh, I guess there's stuff like that too
 
there's the usual "-20 bytes if you do X" which all the golfscript/cjam solutions take but nobody else does
i think -X% works better for this
 
Yeah that's what I meant by bonus
I've thought about doing -X% myself, but have never been able to figure out a good X
 
 
3 hours later…
grc
5:26 AM
Golf Practice: Python has been closed
 
... oh
Controversial, I guess
 
5:50 AM
@Sp3000 I found a work around for the regex golf. Haha
 
Oh?
 
Obviously it does not count because you're not allowed to cheat like this. regex101.com/r/nQ5bG9/2
 
How's that cheating?
 
I just figured it would be since I match the non-matching strings and fail them (which only works if they are in that order)
 
Oh right ... yeah you're suppose to match/fail each one individually :P
 
5:55 AM
=P lol
 
Nice try :P
 
I still do not see how you got it in 29 chars.
I have torn every subpattern in each string apart literally
 
I got a 28 actually, although if I'm really mean I could rearrange the chars so that the challenge wouldn't be able to go on for much longer :P
 
Now now... calm down =) I will figure this out.
 
:) good luck
Hmmm maybe I should throw an incentive into the mix :D
 
6:06 AM
Should I reveal the answer to this (Unscramble source code): codegolf.stackexchange.com/a/41228/6638
I intended to leave it unanswered for as long as possible, but there doesn't seem to be anyone who attempted it
 
Yeah, unscramble cracking's died down a bit...
 
6:25 AM
Any Java experts here?
nvm then
 
6:59 AM
This is brilliant :D
2
 
 
2 hours later…
9:03 AM
@Sp3000 oh, you offered a bounty
 
:D
Write a regex gogogo
 
:P
I was just considering to post some comments to poll if people would prefer if I relaxed the scoring a bit to keep the challenge going
 
Oh... right... I forgot about that :/
Well... if you want to then go ahead :) There's a previous answer which I think would deserve a bounty anyway, if it comes to that
 
no, with the bounty it kinda feels wrong now ^^ ... I might do it after the end of the bounty... although if I do it would be nice if you awarded it to user23013 in case he loses his first place due to the rule change
(he won't immediately, but who knows what happens in the long run if I do that)
 
... that was silly of me :(
 
9:14 AM
no no, I'm still not even sure that changing the rules is a good idea. and having a bounty awarded before the rule change is actually not a bad idea
 
Ahaha k, hopefully someone finds an appropriate regex then :)
 
 
2 hours later…
11:27 AM
My rep graph is sad for the last 10 days :(
 
12:08 PM
@MartinBüttner How are you planning on relaxing the rules? Increasing the size ? (and I'm not just asking because I'm stuck since three days on a 31 bytes regex :) )
 
Haha, Martin has a really nice way of getting 30 ;)
 
12:32 PM
@plannapus, allowing another character every 5 or 10 answers and changing the denominator of the score to 30+N/5 or 30+N/10
 
12:56 PM
that's actually not a bad idea at all.
 
Did anyone clearly understand the bag of letters and words question ?
 
1:21 PM
I was able to get a regex which matches all from non matching list and does not match any from matching list, in around 28 bytes, (don't remember right now). But its of no use :D
 
2:05 PM
@Fry and @xnor are on a roll today. Way to bump me off the transcript!
 
Somebody starred all of our 'meta-meta' proposal :P
 
@Optimizer that would have been a nice idea too (being able to decide which set to match and which set to fail)... but that's a too crazy change to make now
@Rainbolt I'm surprised you haven't close voted the meta-golfing tips question yet :P
 
@MartinBüttner I wish you could have seen the reaction on my face when I at first didn't believe you and then I found it.
 
I said my piece and that's all I have to say about that facet of the question, but I do want to talk about something else
If stuff starts getting extracted from these tips and added to the tag, beware of making the tag too large
Short and simple is probably easier for newbies.
 
2:15 PM
I think the tag is already pretty exhaustive... it doesn't cover all pitfalls but it does cover everything the spec needs
 
But if someone does come up with an awesome tip (which I admit is a side effect of asking for them) then you might want to collect it. You never know
@FryAmTheEggman You helped me verify that I am indeed not very smart.
 
Uh, you're welcome? :P
 
After spending the greater part of my free time yesterday thinking of rules for generating fragile primes, I noticed that my observations were labeled "trivial". And then that guy pointed me to a page on Mathematics where a collection of smart guys has basically written a research paper on the subject.
 
@MartinBüttner - another cool idea, the current answer decides which set the next one matches
 
Well, anything anyone can come up with in < 1 day is 'trivial' in my book :P
 
2:29 PM
leading to 4 possibilities for each question, where to put the regex and what to choose as matching for next one
 
It just occurred to me that all I have to do is translate an existing Python entry into Java or C in order to beat them
 
2:48 PM
That probably would've worked on me, but checking all the permutations won't be much faster in just another language.
Unless you go crazy with C/C++ and do everything with bit-shifts ;)
 
3:02 PM
@FryAmTheEggman Python can bit shift as well
 
@Sp3000 o.O (the new regex)
 
I was referring more to doing something like this
 
@MartinBüttner wasn't that answer already there in deleted form ?
 
@Optimizer no, the deleted one was completely different
P[.$?[]|w\^|^[^|C\\]*$|^[!P]
but he posted a very similar one a bit up the chain:
^..(.[!G)(3w^]|.{7}$|$)|\$\?
 
I can see deleted answers ..
but I think I remembered that there were two deleted answers
and now there's only one
 
3:06 PM
there are 3
but none of them are similar
 
oh wait, there are still 2 deleted :D
I am not used to order by oldest yet :D
 
@Unihedron @hwnd we have a new regex :)
 
I wonder why he has written "This answer still works..."
 
It's the same as an old one, with the | values swapped around
 
oh
 
3:15 PM
not quite, but almost
 
3:25 PM
@MartinBüttner ?!
 
2
A: A Different Kind of Meta Regex Golf

user2301329. PCRE flavour - 28 ^..(.[!)3G^w]|$)|\^.{7}$|G\) Tested on Regex101 This answer still works... The next answer has to match the following strings: PPCG ^P ^[P^]P [^?][PG]$ (?<!\\..)(?!]).$ ^[\w^]*$|!|]P|G]\$$ !|[^?]P(CG|G..)?$ [^])]\$|^\^?P|P.\].$ ([.$?]|G\])\$$|^\^?P|\]P$ ([P.$?]\$|[]^]P...

 
.............
GG
 
I wonder if Sp3000's solution still works now
@Unihedron btw, I'm considering to relax the rules a little once the bounty is awarded. so stay tuned ;)
 
@MartinBüttner :)
 
3:51 PM
@user2179021 what about unary -? does it have to be surrounded by parentheses? or can you write n*-15?
 
4:30 PM
I've made a first pass at the test case generator
first, here some example output: pastebin.com/gXwZtwQW
It's pretty naive so far. It just throws together a random syntax tree up to depth 4, taking care not to nest factorials, binomial coefficients and exponentials (the base of exponentials is fine, the exponent isn't). Still numbers some times get ridiculously large, and you get stuff like division by 0 and 0^0, but Mathematica just ignores stuff like that. Then I chop off any ends of the series that get rational/real, negative or greater than 10^20.
if at least two (distinct) elements remain, I add the series to the results.
and here is the code: pastebin.com/77ByGQr9
it should be reasonably understandable even without Mathematica knowledge
oh, and I select numbers randomly from 1 to 100, currently skewed towards larger numbers. I actually wanted to skew it towards smaller numbers, but I did that wrong, will look into fixing it.
I suppose a better idea might be to build up the syntax tree from the ground up, evaluating the first twenty entries for each subtree, so you can abort early if you get division by zero (or similar), throw the tree away, if things cancel to a constant, and avoid factorial/exponential when the values exceed some hard limit that makes sense
Okay, here is 100 sequences for fixed code. It gets reasonable sequences much quicker now: pastebin.com/wNJuHgtB
(fixed meaning, it skews towards smaller number literals. I'm still building the syntax tree randomly top-down)
 
4:55 PM
@MartinBüttner thanks for the comment.. can you tell me how you find the sandbox?
when it isn't linked from the front page
it is now so it's ok ;)
@MartinBüttner I added smoething about your example
it is n*(-15)
 
@Geobits I think Linda got a bye the second time because otherwise Steve would have had 2 byes in a row. Then Linda's seed was higher than kyle's, so he had to play (?)
 
Sure, that's why I'm asking if it matters and how exactly it needs to be handled. If I don't have to track who had a bye in the last round, that's one less thing to code.
 
Asking is definitely a good idea, I was just saying what I thought he meant.
 
Yea, I pretty much assumed it was so Steve wouldn't have two in a row. I've also seen formats where the lowest ranking teams got the byes each time, just to throw a weird unpredictability into the mix (a barely-squeaked into playoffs player makes it to the last few rounds, etc). People are weird :D
 
^Basically the plot of all sports movies ever :)
 
5:11 PM
You know what's better than watching sports? Watching fake sports :D
 
@user2179021 did you realise that the 15 messages after that were addressing you as well? :P
 
casually listens to lorde while writing FFT visualizer
 
Weird Al does Lorde songs better than Lorde IMO ;)
 
@Geobits hence i'm listening to remixes specifically :P
 
5:30 PM
@Geobits man, I loved Foil :) The innocent lunch-packing to the illuminati was pretty great :D
 
i can't find a good remix :/ time to make one
googles "how to get song stems as a person that's too shy to email lorde"
 
5:56 PM
0
A: Sandbox for Proposed Challenges

QuadmasterXLIIShortest Program that May or May not Terminate: Write a program such that whether or not it terminates depends on the answer to an unsolved question in Computer Science or Mathematics. For example, your program might test the Goldbach conjecture for every N and quit if a counterexample is found,...

 
 
1 hour later…
7:19 PM
@MartinBüttner Funnily enough, that's exactly what I had :)
 
7:42 PM
How do you guys think a Kolmogorov complexity challenge where you have to output the nth Rule of Acquisition would go down?
 
7:57 PM
@BetaDecay Would you provide a program to test the entries?
@BetaDecay Before I followed your link I was imagining a series of Acquisition Rules starting with RAII
 
@githubphagocyte Yeah probably. I don't think there's enough Star Trek on PPCG :D
 
@BetaDecay Based on TOS, write code that causes the computer to burst into flames when asked to divide by zero
 
Reminds me of this ;)
 
8:13 PM
@BetaDecay There's an xkcd for everything... :)
 
That would be a good challenge (although almost impossible :) where, given a piece of text, it produces a valid xkcd
 
I can't tell if this guy is getting smart with me or not boardgames.stackexchange.com/q/20785/6692
The first couple of times I read his latest comment, I decided he was seriously asking if he had the freedom to chose. But the more I read it the more I think he is upset that I edited his post.
Am I reading way too far into it?
 
Which guy is it @Rainbolt?
 
The OP
"Do I have a choice to decide between these options, or do you decide it for me? "
 
Yes, he's upset. Source: telepathy
2
 
8:19 PM
^ Seems pretty conclusive to me.
 
@Rainbolt I think he's just one of those people who likes to act like that
 
Hmmm I guess you guys can't see that he has broken English because you didn't see the original post.
So I'm reading his comments in a difference voice.
 
@Rainbolt although the wording could be used sarcastically, in this case I get the impression that this is someone so new to SE that they regard you as a moderator due to having the power to edit. So my guess is that it's more asking for your advice on how the site runs than challenging you.
 
Hmm. I think he might be like the guy who commented on my post. Maybe he is upset... I have no idea now.. :/
 
@BetaDecay Did you delete your comments?
 
8:23 PM
@Rainbolt Yeah, I think I said something along the lines of No need to be so facetious
 
Ok, this is just confusing:
it was so long ago since i coded in python... and so little did i know of its library that i've never seen "zip" before... i'll soon look into the documentation for this. and by "soon look into", i mean look into it after a few days (that's just how i roll)... but there certainly are things i could pick up from this. thnx. — Amnat Heer Sep 1 at 1:22
 
@Rainbolt Could they be asking about the ambiguous cases you mentioned? Asking whether the site ("you") chooses which interpretation, rather than asking if you personally are imposing link style?
 
@githubphagocyte I actually was going to ask that, but I decided that the answer was the same regardless. It's his choice.
 
I have no idea about that comment...
Maybe he thinks that zip is a library?
I mean it is, but not that zip
 
@Rainbolt This is the problem with plain text communication. Without an indication of sarcasm it's always a question hanging in the background. I have the same problem in spoken language because my sarcastic voice sounds identical to my serious voice.
 
8:28 PM
I never have that problem.
 
@BetaDecay maybe they are using "library" very loosely to include the built ins too?
@Geobits Do you mean you are reliably always sarcastic...? :P
 
@githubphagocyte Yeah probably...
 
@githubphagocyte That's one interpretation. If read sarcastically, it's a humorous way to reaffirm your point. The problem is, which is it?
 
For future reference, I find that a parenthetical exclamation mark works well for avoiding ambiguity in intended sarcasm(!)
 
Inability to transmit sarcasm is why <sarcasm/> tags were invented ;)
 
8:31 PM
Please fix your sarcasm tag. It offends the eyes as it is.
@githubphagocyte Is that sarcasm?
 
</sarcasm>?
@Geobits You mean it didn't remove the possibility of misinterpretation??
 
Okay, so sarcasm is officially starting to not look like a word any more. I need to step away from the computer :P
 
I now have another xkcd in my head...
@Geobits What about sarcastical?
 
@Geobits I had that problem with "good" yesterday.
The longer you look at a word, the more it starts to sound like a person with Down's Syndrome is reading it in your head.
 
@Geobits Serious question, did you add the second "u" in "humorous" for my benefit, or is that an exception in US English that I haven't seen before?
 
Humourous? That's weird...
 
@githubphagocyte OH wait... you asked if he added the second u. Do you usually spell it humoros?
 
@githubphagocyte That's the only way I've ever seen it spelled. 'Humoros' doesn't seem right at all to me ;)
Now if I wrote 'humourous', that would be strange ;)
 
"humoros only yielded one result"
 
@Rainbolt I thought that was the US spelling until you spelled it out and now I feel uncomfortable looking at text in general following how stupid that makes me feel :)
 
8:38 PM
@githubphagocyte Well to be fair, we are on the topic of how staring at a word for too long makes it look weird.
 
Let's have a poll: aluminum or aluminium? I'm aluminium
 
You're insane.
 
ALUMINUM
 
It's long enough at four syllables, why add another useless one?
 
Well you have sodium and lithium. -ium just looks right for an element
 
@BetaDecay I think "ium" is the ending indicating a metal. It's not right to have metal names not ending in "ium". We need new names for gold, silver, iron...
 
Here's the British one:
 
@BetaDecay I'm in the UK and say aluminium myself, but I still think it's a silly argument :)
 
@githubphagocyte Rainbolt is the best.
 
Lanthanum, Platinum, Tantalum...
 
8:43 PM
@Rainbolt that's an excellent point. However, out of context it sounds crazy, so I've starred it... :)
 
That's really the best reason to star something, but you should always wait two minutes :)
 
@Rainbolt genius
 
The elements are already about as bad with exceptions as the English language, it doesn't really matter what you call alumini?um
 
Yes it does, because I'm right :P
 
8:45 PM
(Incoming xkcd reference) Wait, someone is wrong on the internet!
 
You can't strip away too many syllables since alum is already a thing, so how about alumm for the metal?
 
If it helps, the Welsh is Alwminiwm which is pronounced the same as British English..
 
In my head that reads AHW-loo-MIH-nee-wum
Ugh I can't get it right
 
There's also the obsolete 'alumium' version.
 
@Geobits Ok mr. wikipediahead. I can has google.
 
8:48 PM
@BetaDecay pronounced aloominioom - close to the English pronunciation but longer
 
@Rainbolt Haha yeah, just wait until you see the town name: Llanfair­pwllgwyn­gyllgo­gery­chwyrn­drobwll­llanty­silio­gogo­goch... The locals call it Llanfair P.G.
 
@BetaDecay it's so long since I've seen that written down - does it really have a DOUBLE double L in the middle??
 
YAHNTEE SILLIO GOGO!!!
 
@githubphagocyte Wow, I've never noticed that... That's just going to spray saliva over everyone...
 
@BetaDecay I might have guessed you didn't type that in by hand... ;)
 
8:52 PM
@githubphagocyte Oh god no! I doubt anyone can do that... ;)
 
@BetaDecay I had to in school but I definitely couldn't now...
 
@githubphagocyte Really? That's just cruel...
I suppose you were in Wales then
 
@BetaDecay Lol yes - otherwise it would be REALLY cruel
I can't remember the English translation
 
@githubphagocyte Something about a church and a waterfall IIRC...
 
@BetaDecay Just looked it up on Wikipedia and the pronunciation guide is full of alien symbols.
Llanfairpwllgwyngyll is a large village and community on the island of Anglesey in Wales, situated on the Menai Strait next to the Britannia Bridge and across the strait from Bangor. It is alternatively known as Llanfairpwll, Llanfair PG, or Llanfair­pwllgwyn­gyllgo­gery­chwyrn­drobwll­llanty­silio­gogo­goch (pronounced [ˌɬanvair­ˌpuɬɡwɨ̞n­ˌɡɨ̞ɬɡo­ˌɡɛrə­ˌχwərn­ˌdrobuɬ­ˌɬantɨ̞­ˌsiljo­ˌɡoɡo­ˈɡoːχ] ( )). At the 2001 census the population of the community was 3,040, 76% of whom speak Welsh fluently; the highest percentage of speakers is in the 10–14 age group, where 97.1% speak Welsh. It is the sixth...
 
8:58 PM
@githubphagocyte That was just scraping the bottom of the unicode barrel...
 
The character/letter mismatch seems at home on PPCG:
58 characters (51 "letters" since "ch" and "ll" are digraphs, and are treated as single letters in the Welsh language)
 
whats with the gogogo part
 
@Rodolvertice I think the final goch is a mutation of coch, meaning red. I'd have to look up the rest...
> The name means: [St.] Mary's Church (Llanfair) [in] the hollow (pwll) of the white hazel (gwyn gyll) near (go ger) the rapid whirlpool (y chwyrn drobwll) [and] the church of [St.] Tysilio (llantysilio) with a red cave ([a]g ogo[f] goch).
 
I only hear throat noises when I see "go go go goch"
 
That seems like an unnecessarily specific name.
 
9:01 PM
i read it as silly go go goch
 
@FryAmTheEggman I've just discovered it was a publicity gimmick back in the 1800s...
 
Haha yeah. North Wales is full of made up stuff for tourists. Beddgelert (Gelert's grave) is an example
 
Yeah I remember reading somewhere that the long names attracted tourists
 
That makes much more sense XD
 
@BetaDecay Beddgelert isn't real??
 
9:02 PM
gelert sounds like some sort of icecream
like gerbert
 
gelato!
 
haha
 
@githubphagocyte Nah, I think a farmer made it up.
 
@Rainbolt The final "ch" is pronounced as a throat noise :)
 
@Rainbolt It's the same as H in Klingon, if that helps
 
9:04 PM
That helps.
 
@BetaDecay but it's such a horribly sad and heart wrenching story - oh wait - that does now add up...
For help in pronouncing Welsh, refer to a more widely used language like Klingon
4
 
@githubphagocyte Yeah.. There would be at least some remnants of such a castle.
 
"Students, pronounce the h in goch like the h in pahtak."
 
I felt so sorry for Gelert when I heard that story as a child
 
@Rodolvertice To be technical, it would be paHtak. The ah is pronounced ay like in Kahless, not ach
@githubphagocyte What kind of dog was Gelert?
 
9:11 PM
Sorry, I'm no Klingon expert. Judging by your avatar I guess you are haha. Thanks for the correction.
 
Haha no problem. I'm no expert either, I've just watched a lot of Star Trek and read the pronunciation chapter of the Klingon-English dictionary :)
 
I had a Klingon bus driver once. I almost greeted him "Ka plah" (I don't know the correct spelling) but I chickened out.
 
How does one get a Klingon bus driver
 
@BetaDecay Wikipedia just says "a faithful hound"
 
@githubphagocyte I imagine a wolfhound or something
 
9:17 PM
@Rodolvertice it was just for one day so I assume either a dare/bet or on his way to a fancy dress party. It didn't look like fancy dress though. It looked like full television makeup - no joins. Nowadays I might just ask, but at the time I was too quiet.
 
I can't believe there's a Wiktionary page for Qapla'...
 
@BetaDecay Something large enough to kill a large wolf
 
Also, if you want to great someone, use nuqnaH (What do you want?).
 
@BetaDecay Odd greeting to the driver when getting onto a bus though...
:18708277 you can go back and edit
 
@githubphagocyte Didn't know that. (thanks, i'm kind of new here.)
 
9:20 PM
(that's why it says "Rainbolt is the best" in the star list to the right...)
 
Haha
Say something smart, get a star, then edit...
 
@Rodolvertice Usually when I tell someone that they already know - it's nice to finally be helpful :)
 
Nah, it's easier to get starred with funny rather than smart ;)
 
Don't forget witty ;)
 
Funny things require wit to come up with
 
9:22 PM
Oh yeah. I suppose just standalone jokes don't get starred...
Did you here about the constipated mathematician? He worked it out with a pencil!
 
@BetaDecay you missed a word...
 
I'm not sure what you meant by that haha
 
Well then. That's nasty.
 
@Rodolvertice Haha it was a little experiment of mine
 
9:25 PM
@BetaDecay I'm not starring it but I did laugh out loud
 
Haha yeah. That's one of my dad's favourites
 
@Geobits I'm sure nowadays that would get more than 6...
 
I wonder if the mathematician used pencil and paper to solve his problem...
 
Ugh. We have a requirement to respond with a 500 error code if "something goes wrong internally".
So now our customer is testing our product, and they are asking us how to produce a 500 error code so they can test that we did it correctly.
 
Pull out some cables somewhere and see if that works... ;)
 
9:32 PM
I guess I could wrap the entire service in a try catch and respond with a 500?
 
@Rainbolt so they are asking you what bugs you are already aware of and decided not to fix?
 
@githubphagocyte You have obviously worked in development before, haven't you? ;)
 
@Rainbolt lol no but interacted closely enough...
@Rainbolt Tell them they can test it by putting a slice of processed ham in the cd drive
 
@githubphagocyte my friend once did that, he put 4 slices of ham on the tray, pushed it until it closed, and nothing happened. The computer worked fine, but the cd tray never worked again. (He couldnt get the ham out)
3
 
@Rodolvertice I particularly like that the starred message cuts off the vital last phrase, making it look like good advice...
 
9:40 PM
Haha
It fixes windows 98 blue screens!
 

« first day (1391 days earlier)      last day (3467 days later) »