« first day (1284 days earlier)      last day (1949 days later) » 

12:05 AM
the way I'd do it the DFS would be unbelievably slow, but it would save me having to manage a list of active nodes (as I'd have to do with the BFS)
the main issue here is lazyness, closely followed by home much code I might have to right (which is totally a legitimate concern 'cause it's codegolf)
 
If running time doesn't matter then go for the depth first. Unless people need to be able to run your code to check it works...
 
yeah, it would also be pretty bitty, and frankly I've written so many BFS pathfinders I'm a bit ashamed to be considering DFS
 
Breadth first is quicker if not all the potential paths are the same length?
Can you guide it in some way, to encourage it to check short paths first?
 
you can, but that adds a lot of complexity if you still want an optimal result
speed won't be an issue for this
 
If you could sort the paths by length before considering them, and then process them depth first in increasing length order, you could stop as soon as you find a valid path.
 
12:11 AM
I've spent a lot of time staring at pathfinders back when I needed to traverse a few thousand nodes in milliseconds, this problem is tiny
 
Depends if there is a way of sorting by length without actually processing them first...
 
you use a heuristic, so you could rank them by the linear or Manhattan distance to the target node
then you can follow the "closest" one, and then keep following it if it doesn't go wrong, and hope you find the target
 
You'd have to be confident there was no shorter path than your heuristically suggested one. Either of those measures sound like they'd guarantee that.
 
generally you sacrifice the optimal solution for a not so good one by doing this, but they are good enough, and because they seem to start off in the right direction can seem more natural than the perfect path
 
But this is only buying you time, at the cost of more characters, right?
 
12:14 AM
yeah, it's a non-issue
I just need to get on with it when I'm more awake
 
a night's sleep is more effective than a week's hard work
everything makes more sense after sleep
 
a good run helps as well
"clears the mind" as my father would say
right, sleep for me
g'night
 
oooh, another pacman!
 
indeed - and it's got a few votes already
 
12:48 AM
I'm thinking of making some more Manufactoria challenges? Anybody interested?
 
making things?
do you mean images or solutions?
 
Manufactoria is a game centered around a 2D programming language.
 
Ah I see now. New to me...
 
I think it's a fun game.
 
@PhiNotPi have you played other games like manufactoria?
like spacechem?
or the codex of alchemical engineering?
does this look venn-diagram-y enough for the venn challenge?
# echo "1,2,3,7,13 2,3,8,11,13 3,4,5,11,12,13,14,0,6,9 15" | perl venn.pl ;echo
+---------------+---------------+---------------+
|            1 7|              2|              8|
+---------------+---------------+---------------+
|               |           3 13|             11|
+---------------+---------------+---------------+
|                                0 4 5 6 9 12 14|
+-----------------------------------------------+
10 15
 
1:01 AM
no, but I might now that you have mentioned them
 
That is technically a Venn diagram, but it takes a bit of staring to establish which boundary belongs to which set
Could you offset the edges so each set boundary is separate from the others?
Since it's code golf that's not essential...
 
sure, I COULD :)
but it would cost a few dozen bytes
and I'm winning right now
 
Does the size of each section accommodate the maximum number of potential occupants? Or do the sections grow?
 
the sections are all as wide as the biggest section needs to be
or a bit wider than that
I got really golfy with the cell width calculation
 
1:18 AM
A possible Manufactoria challenge: link
 
hmm
how are you thinking of making this appropriate for codegolf.se?
 
In the past, the challenge has been to use the fewest parts.
It's a language-specific code golf challenge.
 
ahh
I thought you meant to write a program that solved manufactoria problems
 
metamanufactoria
 
1:45 AM
0
A: Proposed Question Sandbox - Mark XIV

chilemagicI got this idea after answering Calvin's Hobbies question which reminded me of a cool programming trick I learned a year ago that can be used to satisfy program number 2. Definitely open to feedback, different formatting/wording, or a different title. Please let me know what you think! This woul...

 
I've written out the challenge, should I sandbox it or post it to the main site?
 
If in doubt, sandbox it.
 
I don't really see much room for improvement.
 
Up to you
 
I think it's pretty straightforward.
 
2:01 AM
user image
2
@MartinBüttner Congratulations! :D
 
@MartinBüttner good work! Did you hit 10,000 exactly???
 
@githubphagocyte Yep, exactly 10k! (I may or may not have upvoted a random question and answer of his at 9985 :P)
Now don't go and ruin it. :P
 
lol
I'll wait until it's had a chance to be seen before I do any voting...
 
2:59 AM
0
A: Proposed Question Sandbox - Mark XIV

PhiNotPiPrimality Testing in Manufactoria manufactoria code-golf Background Manufactoria is a game about programming. The player must use a form of two-dimensional programming language to complete tasks. If you've never heard of it, the easiest way to learn is to try out the first few levels of the g...

 
@PhiNotPi have you confirmed that there is a solution at all?
 
 
1 hour later…
4:17 AM
hey guys, anyone here? was hoping to get some insights.
 
sure
 
i'm currently in the processing of planning a koth challenge
what's unique is that i plan for it to be a 3v3 match
each character is one submission, and a player can team up with other players to make that 3 man team
do you guys think the code golf community would respond well to that? i was hoping to avoid 1 man creating 3 characters to create 1 team, to encourage working with other people.
they don't have to find teammates before they code. they can create their character then put up in their answer that they are looking for teammates.
 
there's a discussion up on meta about having team challenges
more generally about half the entries vs the other half, but 3v3 would probably fall into that category too
 
let me have a look at it
brb
hmm. the community seems enthusiastic of cooperative challenges.
which is good
thanks sparr
 
np
join the discussion on that thread if youd like
 
4:30 AM
Hi all! I've not been here for quite a while :o
 
howdy
quiet time around here, you should try back in 6-16 hours if you want to find a lot of chat
 
I know, right. 24h activity waves cover the whole internet
I wonder if @Doorknob is around :3
 
@mniip Nope
I'm definitely not here
 
Sup
 
@MarkGabriel I'm not sure how much "Working together as a team" you'll get
there doesn't seem to really be the framework on SE for that
there may be 1 team that arises out of chat
 
4:41 AM
@Doorknob Sup? Anything new? :P
 
@mniip Uh, not really. Nothing very interesting happening around here
I am currently being bored and designing a new wallpaper for my computer even though I almost never see the desktop anyway. :P
 
Heh. I've heavily affilliated myself with dogecoin :P
I've replaced their IRC tipbot that was dying and since a while ago I also moderate their IRC channel
 
Lol, very wow
 
@NathanMerrill, right now, one of my design philosophies is that you could code a generic behavior for your character, which would work with or without teammates. and when you do get teammates, you could adjust your code a bit to synergise more with them.
 
@Doorknob You should check it out sometime :)
 
4:48 AM
Heh, I've heard of it before, but I've never really look in to cryptocurrencies in general
 
It's fun \o/ especially dogecoin
:O
I wonder if I can use UTXO script for golfing
 
 
3 hours later…
7:51 AM
@Doorknob Thanks for the screenshot! :) When I woke up I was already at 10.1k.
 
Congratz @MartinBüttner
 
thanks :)
also I never had 10 chat stars on anything :D
 
Did you screencap that yet?
 
@Sparr I have tried all of his games except Ironclad Tactics
("him" being the guy who made Codex and SpaceChem)
He's pretty good at designing these kinds of games
@overactor nope, but I feel like I have a few days to do that :D
even 6 or 7 star messages tend to stay around for a week or so
 
I think it might not stay at 10
 
8:00 AM
@MartinBüttner thanks for the hints at what was wrong with my random numbers. I've got it all working now, just putting together some frames for a GIF then I'll be posting it.
 
sweet :)
@overactor ah, a screencap of 11 stars isn't bad either :P
So until yesterday we didn't have set-theory tag. That can mean one of three things: there is a distinct lack of set-theory challenges; they have all been tagged with something else (like combinatorics); set theory just doesn't provide a lot of interesting golfable problems.
If we figure out which one it is, there might be a whole lot of new decent code golfs to ask.
 
Either 1 or 2
 
okay I think I've got a retagging project: codegolf.stackexchange.com/search?q=set+operation
 
210!
Not as bad as it looks though - a lot of those are answers rather than questions (I took a moment to realise why so many of them had no tags...)
 
yes, but I searched specifically for "set operation"... I believe there would be a bunch of others that don't have the word "operation" in them
there's also a way to filter out answer, but I forget what it is
oh there's a help for the search... never noticed that
it's still a lot of false positives though
 
8:17 AM
Do the false positives have anything in common?
 
 
1 hour later…
9:21 AM
haven't checked yet
 
Posted my diffusion limited aggregation image
Going to add another image and gif later with the particles that haven't yet aggregated visible too
 
on a related note, random paintings have finally taken over mandelbrot tonight
(thanks to a downvote on mandelbrot I think :D)
just saw the DLA answer
nice!
can you make them fill more of the plane?
o.O ++d%=4 is valid?
that's amazing
 
@MartinBüttner I just saw that you reached 10k. Congrats!
 
@MartinBüttner yes it can fill up as much as you like but that would require looping higher than max for int. I could change it to long long but I'd have to find 6 extra characters from somewhere...
 
@ProgramFOX thanks :)
 
9:31 AM
You can also fill up more of the space by starting with more seeds but then you get lots of little ones - I experimented a bit and settled on this size.
 
@githubphagocyte can't you use a double for a loop somehow?
 
I too was amazed by ++d%=4
I only tried it out of desperation - I really didn't think this was going to golf down enough
 
so apparently my swirly answer had an ancestor in the 90s
 
did you find it?
 
I also tried #define D #define but no such luck
 
9:33 AM
lol
 
Does #define D define work? Then you can try #D as a define.
 
@ProgramFOX I tried every combination I could think of - I really was that desperate...
It won't recognise macros in the preprocessor directives.
 
I was hoping so hard
 
@githubphagocyte you've got a lot of defines, you might be able to save something with #define D(i) !i?firstdef:i<2?seconddef:i<3?thirddef:fourthdef, and then use an integer parameter to switch between the macros
 
9:36 AM
i actually tried that a few months ago, would've been great for the ioccc
 
@MartinBüttner that would be incredible. I think I need to let this one go now that I've posted it or I'll break my brain. I want to move on to the next idea though so maybe I'll end up trying it on that one...
 
@MartinBüttner hey man. got an email from a newbie who'd like to submit a challenge but he's new to PPCG and doesn't have enough rep to submit an answer now that it's unprotected. what's the policy on un-protecting a Q?
ref: the RPSLV challenge
 
i only recently learned that you could upvote multiple answers on the same question...I've been on an upvoting spree xD
 
@EoinCampbell damn... I guess I'll just unprotect it and leave it unless you start getting a lot of spammy submissions
@cjfaure o.O
 
9:47 AM
I actually don't see why it is protected. There was no spam, and just one of the deleted answers came from a 1-rep user.
 
@MartinBüttner cool. I figure this guy is legit, he send me his code via email and it's fine so YMMV :-)
 
@EoinCampbell done
 
@MartinBüttner and by that, i mean 2 or 3 per 15
i upvoted 4 things on the tweetable art popcon ;_;
 
@Trimsty you can upvote all 15, but then you won't get many votes for the rest of the day...
 
@MartinBüttner thanks dude
 
9:49 AM
@githubphagocyte ;3
 
Did I say 15? There are 23 answers now...!
 
81
Q: Toilet flush buttons

BoranasThis question has been bugging me ever since I've started learning usability. I am talking about the two buttons for the flush : Full and half container. On one hand, the button shape has to represent its function. Thus a bigger button must mean the full container. On the other, dangerous/hea...

is ux.se going down the drain?
 
10:25 AM
@cjfaure so how do we figure out that gravity language :D
 
@MartinBüttner i dunno :P
 
I was thinking if it were actually the variables that fell through the grid and determined the control flow that would be quite horrible
(or at least harder to realise)
 
@MartinBüttner i was just thinking they'd be values and there would be no variables
 
because then you'd need grid cells that wait until two values fall in to throw out something else (binary operators basically)
well yes, values
 
yup
 
10:27 AM
instead you could just make it Befunge-like just that the cursor movement is controlled by gravity (and portals)
 
also just had an earthquake here. scary stuff ;_;
 
just a little shake I hope?
 
@MartinBüttner well anything is scary when there's a giant geyser above your room on your (weak) ceiling
 
o.O
well...
 
xD
 
10:28 AM
I'm glad your ceiling is still where it's supposed to be, then ;)
 
@MartinBüttner indeed d:
 
btw, I think the language would work better if each cell was two characters wide
 
@MartinBüttner interesting
 
you could use better mnemonics for the symbols and you could enumerate portals to pair them up
 
@MartinBüttner why not let cells be arbitrarily sized?
 
10:30 AM
because then they don't line up nicely
 
@MartinBüttner they don't have to, they just have to take up X units in shape Y
like html tables (ew)
 
but it has look nice!! ^^
 
@MartinBüttner Ooh! Let's do that ;D
pretty esolangs are rare
 
how would branching work?
 
@MartinBüttner eh?
 
10:32 AM
well we've got loops with portals
how do we realise conditional execution
 
@MartinBüttner Ah...pistons?
 
yeah but what determines if they are triggered?
(you could also have a \ or / which can change its orientation somehow ... or a combined /\ )
 
when they're passed a value they toggle to that value
hm
 
I'm looking forward to seeing a 3d animation of this language one day...
 
3d? o.O
 
10:46 AM
Maybe I've pictured it wrong
I was seeing a 2d grid with values falling onto it
 
it was just supposed to be a 2D grid
 
well yes, but you're not looking onto it from the top
 
oops...
 
but from the front
 
10:49 AM
i was also looking at it from the front o:
 
@githubphagocyte love the new version with the particles!
 
@MartinBüttner I just pressed post...!
Good timing...
Thank you :)
 
btw, do you have enough characters left to invert the colours?
 
Which ones? Do you just mean a negative image? Or blue particles and yellow aggregates?
I'm very short on characters, but some colour changes can be made simply by switching the code between the functions, at no character cost.
I don't understand how the macros work. They seem to be able to use substitution rules that are defined later than themselves. This does mean swapping the functions around is trivial to do.
 
whew you have exactly 3k network rep
 

« first day (1284 days earlier)      last day (1949 days later) »