« first day (1501 days earlier)      last day (3274 days later) » 
00:00 - 21:0021:00 - 00:00

12:00 AM
I got bored with the pattern matching language and started playing with new code for inca3. I've got some encouraging results in accessing the extended graphics characters with xterm. This should allow me to pack even more commands into a byte, at the expense of defining several more conversions (text-input format, disk-format, memory-format, text-output-format).
I also want to combine the ability to use longer names with the existing ability to use names without intervening spaces. I think I can do this by taking the set of decreasing prefixes of a string, returning the longest prefix that is defined and pushing back the remaining name fragments to the scanner.
12:13 AM
@Calvin'sHobbies only now did I realise the significance of the number 99 for your question today -.- (despite checking earlier whether question no. 100 would be your next)
I leave PPCG for a few hours and I come back to one of my posts being close-voted?
we do that all the time when you're not watching
12:31 AM
A: Sandbox for Proposed Challenges

da capo ariaChallenge about seals stranded upon an iceberg (temporary title; title suggestions always appreciated!) code-golf A family of seals are stranded upon an iceberg in the Artic Circle. There is a radio transmitter located on the iceberg which the seals can use to call for help. However, only the ...

12:45 AM
^ The backstory made me laugh. :D
2 hours later…
2:17 AM
The sandbox just got its 16th "More than 10 answers posted to this question in the past 7 days" auto-flag.
Yep, sixteen. I am slightly annoyed at this.
How often does it get an auto-flag?
Do a lot of PPCG questions get auto-flagged?
Every 7 days, I would assume.
Autoflags are about 80% of the flag workload here -_-
Well, it seems like the mods tools are simply making more work for themselves.
At that point, they're not really "tools."
Heh, occasionally they're useful. But usually it's just "oh, hey look, another 'too many edits' autoflag! declined"
On the website I "moderate", there's a flagging system, but all it does is send an email notification to the mods' (special purpose) email addresses.
I'm just listening to some music right now.
Waiting on some interesting KOTH entries to show up.
2:46 AM
I'm thinking of a challenge where everyone's name would be hashed to like 3 Turing-complete language on esolangs.org and you should have to do the given task in one of those languages.
That would be... interesting.
It's hard to find the interesting ones in esolangs and here the good ones could be voted to top.
Or there could be a challenge to stick in your username/userid and have the program spit out a deterministically random esolang (possibly making it change every day?), and the submissions to that challenge could then be used by users here who want to try new languages.
Yeah, mine would have a snippet with name input, language links output too.
might worth to add the daily change too
But in your proposal you don't get to vote so you never know which is a good language.
And it would be really disappointing to get a language like this one. :P
2:55 AM
I would restrict it to the Turing-complete ones: esolangs.org/wiki/Category:Turing_complete
Oh, whoops, right. I was looking at the "all languages" list.
2 hours later…
5:09 AM
A: Sandbox for Proposed Challenges

James MasseyCrack a Keyed Caesar Cipher Given the input below, write an algorithm to crack a keyed Caesar cipher. A keyed Caesar cipher consists of a word which acts as a key (eg puzzle) the rest of the alphabet is then shifted as demonstrated below: a b c d e f g h i j k l m n o p q r s t u v w x y z ...

2 hours later…
6:44 AM
@Calvin'sHobbies like I was not right :P
1 hour later…
7:46 AM
Is there a main meta post that talks about not using languages invented after the challenge? I can't seem to find one.
Interesting... I'm having trouble too
Is it just an unwritten rule? :P
Well it is written... just all over the main site itself
Maybe this post?
Closest I could find
8:05 AM
@Calvin'sHobbies kind of
8:26 AM
Q: Can I ever answer with a language invented after the challenge was posted?

Calvin's HobbiesA common rule here on PPCG is that programming languages invented after a challenge was posted cannot be used to answer that challenge. (I've had trouble finding a specific meta post that addresses this though.) This rule makes total sense because without it, someone could "invent" a language th...

@Calvin'sHobbies dude
I asked the same question like 1 minute from you
Hahaha, too slow! Sorry ;)
you should change the last few lines of your question
remove themm
they should be answers.
posted on March 10, 2015 by Optimizer

It has considered as an unwritten rule for a long time now. Lets have a written decision on this. This post targets all answers written in languages or versions of languages which were created/launched after the question was asked. Whether the newer language/version was influenced by the question or not, we should have a single stand.

wow ^ what a bug!
I have deleted the question long since. its not even a proper one box
8:57 AM
@Calvin'sHobbies there's this meta.codegolf.stackexchange.com/a/1420/8478
@PeterTaylor Do you happen to know the computational power of a Turing machine that can only perform an arbitrary but finite number of steps?
2 hours later…
10:50 AM
@MartinBüttner I'm not quite sure what you mean.
11:06 AM
I think he;s asking the time-version of "what's the computational power of a Turing machine with a finite memory"
Kinda like how a real-world computer can have almost an arbitrarily large memory, but never "enough" (infinite).
@PeterTaylor ^ ping :)
11:22 AM
I interpret it as "describe the class of those Turing machines that always halt".
12:14 PM
Just to add my thought: for me the "arbitrary large but finite number of steps" seems to be the definition of "infinite steps".
12:25 PM
Maybe this wiki article is relevant: Machine that always halts
12:58 PM
@Zgarb yes, although I think that should be equivalent to machines with finite memory, since you can solve the halting problem for those.
@PhiNotPi thanks that's what I was looking for
@MartinBüttner What do you mean by finite memory? Is it constant, or given by some computable function of input size (then it's not equivalent to the above)? Or is it given by some uncomputable function of the input size (then their halting problem is probably not solvable)?
yeah actually I think that total Turing machines might be more powerful that what I'm looking for
there will be an a priori limit to the number of steps the TM can take. that limit can be set arbitrarily (and can probably be at least a linear function of the input size, possibly even polynomial), but it will be bounded.
okay, the limit can definitely depend on the input size by a computable function, but probably not by any computable function.
Well, then you just get all functions in the respective time complexity class.
right, that makes sense
@Rainbolt you ninja-deleted your comment while I was replying :D
out of curiosity, did you get a ping anyway?
I deleted it within ten seconds of posting, and no.
I know why you replied without even reading
1:10 PM
That post is very pedantic
what, my comment?
The OP
oh okay
"I know I can't win with an answer written in a language that was invented after the challenge, but can I technically post my answer anyway and just not win?"
Answer is: who cares really
1:13 PM
people who want to answer old challenges and are afraid of down and delete votes for posting a technically invalid answer?
@Zgarb that is a really interesting challenge
@MartinBüttner This guy was't afraid - codegolf.stackexchange.com/a/32869/18487
Anyway it's been asked and now it'll have an official answer
I for one would like the disclaimer I suggested in the comment to be made mandatory for such answers
@MartinBüttner I started thinking about the reverse operation of merging one day, and was surprised that it is not trivial whether it can be done for a given list, but it admits a short algorithm (at least if you have no duplicates).
I don't know how much harder it is with duplicates. Depends on the tie-breaking rule, I guess.
I've been playing more with the xterm graphics charset. I've cobbled together an alternate keyboard which you can switch to/from with ctrl-N and ctrl-O.
A: Spiral Programming

Pietu1998Python 2, 290 It's probably really bad, but I tried :D The output contains trailing spaces, but that is not forbidden in the spec. m=x=y=c=0 l,f=1,[31*[' ']for t in[0]*31] for i in[0]*input(): k=m%4 f[14+y+(2<m<6)][14+x-(m>3)],x,y,c='|/_\\'[k],x+(k>0)*(2*(4>m)-1),y+(k!=2)*(2*(2<m<6)-1),c+1 ...

1:23 PM
The layout is not final. I just quickly defined everything so I could display it.
This does tie it to xterm or other vt220 terminal emulators.
so a windows port will be a chore.
Wondering if I can cut/paste these characters from the terminal. Do they translate?
Hm. kinda.
1:51 PM
@Zgarb I guess then I need to figure out what exactly sort of functions I can exactly compute for the time limit
2:11 PM
Running a Core War battle now.
It's pretty slow.
Well, I would expect that, considering how many battles are being played.
About 16000 per pair.
Dwarf beats Imp
It will probably be worth my time to optimize the controller.
Unofficial Leaderboard:
5 - CopyPasta
3 - Janitor
2 - Dwarf
2 - Imp
I'm going to ditch some debug text and see if I can run an actual full tournament his time.
Because I think the fact that debug text was generated each cycle (regardless of whether it will be printed) would slow it down a lot.
The above board was done with a small core size.
In the worst case, both opponents would be imps that can't win.
Then the full war between those two would be (8191 possible starting offsets) * (8 * 8192 max time per battle) * (2 options for who moves first each battle) = 1.07 billion ply (total instructions executed during the war)
It's about seven minutes per pair-off right now.
2:43 PM
Official pairwise results, hot off the press!
Dwarf Imp
CopyPasta = Imp
Janitor = Imp
CopyPasta Dwarf
Janitor Dwarf
All of my databases have entered the "Emergency" state. That sounds bad
On the bright side, I have two million READONLY rows.
Janitor CopyPasta
    5 - Janitor
    3 - CopyPasta
    2 - Dwarf
    2 - Imp
Is it just me, or does the phrase "dating back to the 20th century" make anyone else here feel old?
2:59 PM
It makes me feel old
I remember when Rock Band introduced a genre of music called 00s
To be fair, Rock Band wasn't around until 2007ish, so they only had to wait a few years for the decade to be over.
I still have a working Rock Band setup
I like this wildly-optimistic quote:
> In the article, guitarist Steven Van Zandt claims, "in the history of rock 'n' roll, Rock Band may just turn out to be up there with the rise of FM radio, CDs, or MTV."
Not quite....
Well, it was headed that way
Was it ever more than a fad? I mean, I know it was hugely popular for a couple years, but...
3:03 PM
Rock Band was responsible for many more get-togethers than Halo in my small town.
My Mom plays bass on Expert while singing on Expert.
And she doesn't play video games at all usually
must refrain from mom joke...
I mean she played Duke Nukem with my dad, my brother, and I when we were like 10 and 12
And I think she did Mario Party
Anyway I think she spent like $400 on DLC for Rock Band.
And all of my birthdays and christmases for those two years went to DLC as well
So we whip it out on holidays and play Rock Band again
Nice preemptive edit :P
3:07 PM
You know, drumming on a Rock Band set really teaches you how to drum well. I know it looks like a joke. Someone who actually drums thought I was joking
I could see that for the drums. The guitar is terrible for that, though (for both it and Guitar Hero).
You get the coordination of moving different limbs at different speeds and you get the repetitive movement down pat.
All that's left is learning how to hit stuff that isn't a pad.
And the hi hat with your left foot I guess
Nah, just get an electronic drum set. There are some really nice ones now.
Unless you're live on stage, nobody would know.
Oh I have. My friends and I wired my brother's trap set so we could play Rock Band with it
One of my friends is a robotics engineer
That sounds like more fun than playing the game :D
3:11 PM
Biggest challenge was getting the sensors to not register 20 taps because of the vibrations
If I could wire a guitar (in a way that made sense) up, I'd give that a shot.
So we tightened all of the drums to sound really terrible, and we set it to not register more than once every x ms
Peak detection probably would've worked better.
We were in highschool lol
Assuming you had good software control.
3:14 PM
We built a trebuchet right after we learned trigonometry and we thought we were awesome for using it. Didn't work. We launched a pumpkin like 20 feet using a 600 lb barrel of concrete
That sounds very disappointing. Obviously you just need ten times as much concrete ;)
Then we built a potato gun and launched a fencepost .35 miles (according to a four wheeler's mileage ticker)
We also painted a wall with it using some red solo cups filled with paintballs.
We built a golf ball cannon. Used an air compressor filling large PVC pipes as a pressure tank. Add some fast-opening elctric valves and you can do some serious damage.
That was ours as well
Our first one was powered by Axe Body Spray
That one sucked though
I have no idea how far they flew, but they were out of sight basically instantly.
Fast enough that it would blow a clean hole in a fence, though.
3:17 PM
Imagine if we had hit a cow with our flying fence post of death
Yea, spray-combustion ones just don't have the oomph :)
My dad would have been pretty pissed
Most likely, yes :D
Oh, so when we went to get the fence post back out of the ground, we're walking through the field and my friend (the now robotics engineer) says, "Run." with a panicked face and takes off.
That's... never good.
3:19 PM
So we are all running and after about 200 yards I finally look back and there's nothing there so I ask the question: "Why are we running?"
He just says "Keep running" and he looks panicked as hell still
So we get all the way back to the four wheeler and he finally tells us that he stepped on a coffee can that some hornets were living in
And they are still following us.
We drove off and nobody got bit thankfully
Turns out it was the coffee can that my dad uses to bang on when he feeds the catfish. He bangs on it and the top of the lake goes wild and then he throws out some dogfood and they eat it.
I always regretted not getting some of those GPS-tagged balls to find out how far we were launching them.
So that could have been bad
Nah, hornets just want to cuddle most of the time.
I saw one of them on a horse and didn't know what it was so I went to go get it off and my dad was like "Get the hell back!"
He threw something at it and it flew off and the horse was dripping blood
I've been scared shitless of hornets ever since
They're like two inches long. Very creepy...
@Geobits Think about it... Doorknob wasn't even born "back then"
3:24 PM
There are worse things to be scared of. At least that one's useful (from a survival perspective).
@MartinBüttner How old is he now, anyway? I swear he's been 13 for over a year :P
lol no, he's been 14 since December or January I think
"You kids these days just don't get it. Back in my day we had to stream our video games."
"back in my day, we had to study hard to play games"
3:28 PM
It's probably clean. How dirty can a shiny piece of metal get?
I don't care what it's made of, I have a general rule of not messing with teenage boys' socks. Who knows what it's been used for?
Great. I created 53 thousand files in ten minutes. And it's gonna take About 1 Day to send them to the recycle bin. Really Windows? Really?
I wouldn't have have discovered the true purpose of socks without American Pie.
@Geobits Luxury!
@MartinBüttner That is so strangely niche that I am amazed you managed to pull that out of the Internet.
3:32 PM
it's the oatmeal
not all that niche
Anecdotally, I agree that Japanese hornets are indeed a big "nope".
@PeterTaylor Oh, I suppose you got your start further back, on those old CD-based machines? ;)
I tried and tried the other day to find a way to play my 16 bit Dune 2000 disk
wow indeed
Sorry about that...
Freud is messing with you today :P
3:35 PM
Holy mother of Jesus. Wikipedia's page on Japanese hornets: "An individual hornet can kill forty European honey bees in a single minute; a group of 30 hornets can destroy an entire hive containing 30,000 bees in a little more than three hours. The hornets kill and dismember the bees, returning to their nest with the bee thoraxes, which they feed to their larvae, leaving heads and limbs behind. The honey and bee larvae are also taken to feed the hornet larvae."
@Geobits CD? Luxury! We 'ad to plug a cassette player into computer and load t' game from that.
@Rainbolt it's much more interesting how the bees fight back: theoatmeal.com/comics/bees_vs_hornets
(the Japanese bees, that is)
Someone brought out Peter's accent.
In text.
Ya'll seein' this?
3:54 PM
@PeterTaylor Ok, you win. My first computer had the floppy drive built in.
Looks like the solution to my problem is:
1. Install Dune 2000 on a 32 bit system
2. Copy the files to my 64 bit system
3. Fix the registry
4:17 PM
@Geobits So did mine, but the games were on tape. The disks were for serious stuff like CP|M.
Oh, floppy and cassette. Fancy! I never had any cassette games; didn't really start playing until '88 or so.
we had cassette games here till late 90s . They sell in high volumes even now.
4:43 PM
You crazy Europeans ;)
I thought Optimizer was from India
I don't know for sure. For some reason I thought he was in the UK.
We do have a state called UttraKhand
stupid java generics
I should totally be allowed to do this
interface ImmutableCollection<E> { ... <E2 super E> ImmutableCollection<E2> add(E2 e); ... }
oh, I think I see why now
4:59 PM
public interface I<E> {<A extends E> I<A> add(A other);}
it's because the type system is crap
Object everything
if I add a String to an immutable list of Number
Java should be able to figure out that I get back an immutable list of Object... but nope
Java in general is a big nope
@TheBestOne that kills the collection type parameter, which kind of defeats the purpose of having generics
Java's the least gross OO langauge that I know, so it's what I use
5:07 PM
OTOH, If I add a String to a list of Number, it's a sign I'm probably doing something wrong. If I need a list of Objects, that's what I should have declared.
if the signature was just add(E e), then I'm definitely doing something wrong
it should be add(LE e)
but the idea of the super thing is to suggest that you want the type to be lifted up
I guess I just don't understand why this is needed (or desired, tbh).
5:11 PM
dont listen to me, I am just random blabbering
no generics are really needed
it all works as <Object>
generics are fantastic when done properly
but then you forfeit the power of generics
5:12 PM
or rather, forfeit to the power of generics
which should include the ability to calculate a shared supertype at compile time
@Runer112 have you considered C# ?
Not really
(it doesn't have multiple class inheritance, should that have been implied by the timing)
that would explain why it isn't the nicest OO language you know
does C# allow this
5:14 PM
this yes
allow what, sorry? the weird covariance described above?
C# isn't exactly bursting with immutable collections (or if it is I haven't noticed and probably should have by now)
more like contravariance
I mean, java doesn't have them either
probably, yeah, I get the two confused
but just the ability to describe such a method
but if the list is is (co|contra)varient, then you should be able to cast down as you like, which I believe you can in Java also, so I'm probably missing the point
5:17 PM
that's probably what I'm going to end up doing
but I can't think of any reason why the compiler couldn't figure it out
probably because the type isn't (co|contra)varient, and it's horrible
it doesn't seem horrible to me, it seems quite well defined
add an object of type E2 to an immutable collection of type E, get an immutable collection of the most specific supertype shared by E and E2
anyways, pipe dreams I guess
I'll try the cast method and hope that it doesn't burst into flames when I try to add a non-String to a class that implements ImmutableCollection<String>
I can't say I'm a fan of most any implicit type conversions
Closest I can find: public interface I<E> {<A> I<? super A> add(A other);}
the notion of adding to an immutable collection seems iffy in of itself, is this in standard java libs?
5:22 PM
Apparently you can't have multiple lower bounds.
I think I'm using the terminology correctly
@VisualMelon Adding to an immutable collection should return a new collection. Adding to a mutable collection modifies the collection.
I was going with the interpretation that unmodifiable means you can't add to it
immutable means you can add to it, but you cannot mutate the object itself, so it has to return a new instance
@TheBestOne quite, but that doesn't sound very javary
it isn't
5:25 PM
Maybe naive to ask, but when is an immutable collection really useful? Besides declaring something like a constant list, I could maybe see that. When should you be adding to it?
when you're interpreting a stack-based language :P
is this collection your stack?
Why would it have an advantage over a mutable one?
no no, these are just lists in the language
this is an attempt to improve CJam stuff, for reference
which is basically a functional langauge, since all of its operations return a value
5:48 PM
6:00 PM
what are we paying our respects for?
functional language
indeed, RIP functional languages
6:41 PM
@MartinBüttner Remember that conversation about SQL indices we had the other day?
Well I've been pounding away on this query for two days now, trying to make it better. Then I had the brilliant thought to disable all of the indices just to see how much they were helping. Query ran instantly. Less than a second.
Turns out there were already 6 indices in place.
@Runer112, the kind of lattice type inference you want may be coming in some future version of Java, to judge by this recent Oracle patent.
what the hell
are they trying to patent type inference
@Rainbolt so much for "if in doubt, add an index" ;)
6:54 PM
If in doubt, remove the mess that's already there. If still in doubt, add an index and wink.
7:20 PM
@PeterTaylor that looks really interesting, but unfortunately, it looks dead as well
7:34 PM
Hi, Does this question fit at code-golf?
What's the shortest way to represent a datetime in alphanumeric characters?


I am printing a datetime stamps on my products. The ink I am using is expensive. By printing numbers AND letters it seems like I should be able to reduce the total number of characters printed while still encoding the datetime information.

Here's what I currently do...

I start with the date as an integer...

201503101436 (That's 3/10/2015 2:36PM)

Then I convert this integer from base-10 to base-36 and get...
How about base-64 ?
base 225 ?
then convert to character
trying to avoid characters that can't be read back over a phone easily
yes want to avoid those lol
7:36 PM
Base 64's probably your best bet then.
base 95
You could do 95 I guess, if you're just reading them back from a phone. Typing a lot of those is still a pain.
colon pound eks less-than caps-kay colon
true that could work pretty well
Any time I have to go into the second/third symbol menu I get irritated :)
7:38 PM
7 u9H7 - seven space you nine caps-eich seven
are milliseconds important ?
Of course, you could probably get it a bit shorter by starting with a unix timestamp instead of full datetime, if it's feasible.
you can skip the initial "20" also. depends though.
You could huffman encode the time :P
@TheBestOne overkill
Just count the number of minutes since an epoch, and base-36 it.
7:44 PM
Since any of these solutions are going to be incomprehensible to humans anyway, just use a tiny bar code. No characters and it saves ink. Read off as "skinny line skinny line fat line ..." :P
@Ypnypn base 95
dunno whats with you all stopping at 36
@Ypnypn 1 2 3 4 5 6 7 8 9 10 ...
still counting ?
Is the printing going to be in monospace? If not, you may (occasionally) have an issue with ''"'"'""'"" in base 95.
printer can't do barcodes :(
7:49 PM
@Optimizer I delegated the job to a program :P
Your printer sucks.
yes it's monospaced
hah yes it does lol
base-36 is still pretty dangerous for human communication
@Ypnypn wins I think that's 5 characters to represent every minute for the next 100 years
7:53 PM
@Runer112 I prefer base-52 when talking with people
You talk in caps and lower? I prefer to use volume and intonation to do that.
No spaces or periods?
but unfortunately, if you reduce to base-33 or so, it'll only work for about 70 years
i'll be dead then so no worries
Well if it only needs to work for your lifetime we can probably shorten it.
7:56 PM
I can imagine the person reading this off, reading calmly
then they hit an uppercase letter
and scream it at you through the phone, and then just continue calmly
(you said something about this being relayed over a phone, right?)
yes it will be spoken over the phone by non-tech people and entered into an input box by non-tech people
There's only ~37M minutes in a lifetime, so that should be four digits (in base 95), tops.
8B is another questionable mix-up
For some reason, I thought you would do it via dial tones.
8B could be bad actually. The text is not very legible.
oh sorry no it would be verbal
7:59 PM
Wait, 8B is your limit but you were okay with 0OIl|?
so you probably want to get rid of at least O, I, L, and B
those would be bad too
If people can pronounce APL, I don't see why we cannot do base 255 ...
verbal dial tones...
People don't pronounce APL. Those "people" are at least part machine.
8:00 PM
here someone actually did
"greater than x backslash bar right shoe 5 circle"
Yes. I'm saying they're not fully human.
On the internet, nobody knows you're a cyborg.
1 0 1 1 1 0 0 0 0 0 1 0 1 0 1 1 1 0 0
I mean here ------------------------^
Interesting. Users can close their own posts as a duplicate now as though they had temporarily gained mod power. meta.stackexchange.com/questions/250981/…
Here's the names of all UTF-8 letters: fileformat.info/info/charset/UTF-8/list.htm
8:10 PM
thanks all for the help... count minutes from epoch convert to base-33 which will work for 70 years. Total characters = 5
base-33? I thought we were getting rid of 4 characters: O, I, L, B
still gives you ~60 years
and you may as well base the times starting in 2015 or something
cool probably ok. I'm looking through TheBestO's list and I really don't see any that are going to print out that good.
yup i'm going to start 2015-01-01
print quality is pretty low
shame that the average person can't read monochrome bitmaps as binary
that would be pretty efficient
I guess that's basically the same as a barcode, though
8:18 PM
Shame that abaldwin won't shell out for a decent printer ;)
it's actually pretty much the best printer you can buy for the application we're doing.
apparently not enough demand for someone to make one that prints at a decent resolution :/
the recent meta discussion made me wonder: why do we accept answers at all?
8:35 PM
What's the difference between not being allowed to be chosen as the accepted answer and not being allowed to win
Shouldn't this cover the issue: meta.codegolf.stackexchange.com/a/1071/32700
@Runer112 its same
@TheBestOne does not include unintentional languages/versions
I'm having trouble discerning the difference between the responses in that meta discussion xnor posted, then
00:00 - 21:0021:00 - 00:00

« first day (1501 days earlier)      last day (3274 days later) »