« first day (1298 days earlier)      last day (3531 days later) » 

12:25 AM
0
Q: How can we attract more questions?

isaacgIt's pretty clear that writing a good question for PPCG is pretty hard - we close a lot of questions, and we have a low rate of questions asked/user. What are some methods we could employ to attract more good questions? Ideas: A standardized question-making form, for inexperienced question ask...

 
12:39 AM
I propose we grab @isaacg in here and discuss his ideas further. :-P
I'm especially interested to know why his proposal made no mention of the sandbox at all. :-)
 
Sure, I'll comment on his post
 
I've already superpinged him, so if he doesn't respond to that, not sure if he'll respond to your comment. :-)
 
Oh, that was a SuparPowarzMegaPing™? Didn't notice that. :P
 
Yeah. :-P
 
What does that do? Does someone show up at your house?
2
 
12:49 AM
@EricTressler Shhhh don't tell anyone; that's top secret SE technology
 
1:01 AM
I wonder how many questions are in the sandboxes....
 
@PhiNotPi you mean unposted questions in all the sandboxes since the start?
 
Ugh. I had a programming interview earlier, and I just realized that my binary search base case used a strict comparison instead of <=, and that it can lead to an infinite loop. Guess I should write that one off :(
 
@EricTressler Post your code, or it didn't happen. ;-)
 
int Inventory::_findWeightIndex(int first, int last, double weight){
    if(weight <= weightList[first]){ return first; }
    else if(weight >= weightList[last-1]){ return last; }
    else{
        int mid = (first + last)/2;
        if(weight < weightList[mid]){
            return _findWeightIndex(first,mid,weight); // [first,mid)
        }
        else{
            return _findWeightIndex(mid,last,weight); // [mid, last)
        }
    }
}
Lines 2 and 3 had < and >. I didn't include adding two equal weight things to my vector during testing.
Time constraints usually lead to me writing bad code, unfortunately
 
@githubphagocyte Yes.
 
1:04 AM
Under normal circumstances, it's quite likely I would have made the same mistake, but I would almost certainly have caught it right away
 
@EricTressler This is why I <3 Scheme. Even under time constraints, I can still write decent Scheme code. ;-)
 
I imagine they expect the time constraint to result in imperfect code. They'll probably be judging your approach more than correctness, unless testing was part of the interview...
 
@githubphagocyte Here is some data from December: meta.codegolf.stackexchange.com/questions/745/…
 
@PhiNotPi you could ask data.stackexchange.com...
 
That is from when we were on Sandbox VI
 
1:08 AM
@PhiNotPi I don't know how that was compiled, but if we wanted it done ongoing ideally it would be automated
 
I compiled it. By hand.
 
@EricTressler I actually dislike C++ interviews, because C++ is such a big language that people have different expectations of what good C++ code will look like.
I <3 C++ as a language, though.
 
C++ is the only language I'm proficient in at the moment
 
The only experience I have with C++ is writing KotH controllers with it. :P
 
As far as writing an actual decent program without extra hours spent googling
 
1:10 AM
@Doorknob I've worked full-time with C++ for 3 years.
 
@PhiNotPi I think it's worth finding out how much of it can be automated with data.stackexchange.com and then you'll know how much needs to be done by hand. Even the data provided by the site can be manipulated further once it's retrieved if we get a program together. Maybe a meta post to establish what information people want to see, and then we can collaborate on a program for it
 
@EricTressler For me, Perl, Ruby, and Scheme all qualify for that. I can write C++ and Java programs from scratch without too much Googling too, but I find that tedious without an IDE.
 
@githubphagocyte You could always download the data dump and just work on that.
 
I definitely have to google things like "STL map" to get syntax right, virtually every time I leave C++ for over 2 weeks
 
@Doorknob you mean to bypass the queries on data.stackexchange.com and just feed the entire data dump to a program, so there only needs to be a single piece of code?
 
1:13 AM
@EricTressler What on earth? #include <map> std::map<int, std::string> foo;
 
@githubphagocyte Ah, never mind, just realized you were only looking at sandboxes. Data dump would probably be overkill. :P
 
shouldn't that be map<string, int>?
Also, I can't tell if you're serious
 
@Doorknob does the data dump only cover whole site or nothing? Or is there one just for meta?
 
@EricTressler That works too.
@EricTressler I'm seriously curious what you're struggling with.
 
I struggle to remember what the member functions of all the STL containers look like
 
1:15 AM
@githubphagocyte There's separate data dumps for meta and main, but again, it doesn't make much sense to download the entire data dump just for a few posts. My mistake; I wasn't paying attention to what you were trying to achieve. :P
 
@Doorknob yes that would probably still be overkill. Thanks for thinking anyway :)
 
@EricTressler find, insert, lower_bound, upper_bound are the main ones you need to remember. ;-)
 
Well, I remember them now, because I just looked them up earlier. But I also used them extensively about 8 months ago, and didn't remember them today
 
*nods*
Yeah, I've used C++ for a few years, so a lot of it is fairly ingrained.
Also, if you love C++, you might like some of the more fun (to me, at least) things that you can do with it:
9
A: Bind pointer to member operators in C++

Chris Jester-YoungA PMF (pointer to member function) is like a normal (static) function pointer, except, because non-static member functions require the this object to be specified, the PMF invocation syntax (.* or ->*) allow the this object to be specified (on the left-hand side). Here's an example of PMFs in us...

 
I've used it for a long time, but I'm always writing quick prototype code, never production code. And I also usually have time to iteratively improve my code
 
1:24 AM
Yep.
 
What we could do with a lot of data is determine the exact effect of "early answer" -> "most votes"
 
@PhiNotPi As a long-time Stack Overflow user, I know FGITW is a thing.
 
Everyone knows it's a thing, but has anyone ever measured it?
 
@PhiNotPi definitely. I think we're pretty much agreed that the current method favours early answers, but being able to put some definite numbers to it would help with arguing our case
 
Look at four-man standoff. Martin's answer was #1 last time I looked, though it's not faring all that well in the KoTH rankings
You could just use KoTH contests for measurement, and look at the correlation coefficient between post time and votes, versus that between ranking and votes
 
1:29 AM
My answer is winning the four-man standoff, and it has 4 votes.
 
You could also use popularity contest, but I think the KoTH examples make a stronger case
Fire up R
 
Opens Mathematica
 
It seems to apply to any question type. I think it's more pronounced for questions that attract long answers, because then people who start viewing with the highest voted answer look at fewer answers before moving on
 
I agree that the longer the answers, the fewer people will read.
 
It does, but the correlation coefficients are objective, measurable things
 
1:32 AM
We would have to define some sort of model for answer popularity.
 
In that case maybe it's worth calculating such coefficients for questions in general, and also for various subtypes.
 
In statistics, a rank correlation is any of several statistics that measure the relationship between rankings of different ordinal variables or different rankings of the same variable, where a "ranking" is the assignment of the labels "first", "second", "third", etc. to different observations of a particular variable. A rank correlation coefficient measures the degree of similarity between two rankings, and can be used to assess the significance of the relation between them. == Context == If, for example, one variable is the identity of a college basketball program and another variable is...
 
That sounds like something that would work.
 
Would you need two different orderings of the same answer set to do the comparison?
 
It does, easily. There are 3 rankings: post time, # votes, KoTH rank
 
1:34 AM
Ah of course
 
no, you just take a post, rank all the answers those 3 ways, and see how the coefficient of correlation between (post time, # votes) compares to (KoTH rank, # votes)
 
So post time to votes is the correlation we want to measure, and KotH rank or code golf length would be interesting background
 
yeah
code golf is a little more problematic, because people would probably favor a 30-byte C program over a 20-byte golfscript one
 
Actually KotH rank or code golf length would be more than just background interest - it would say something more if you could establish that time of post has more of an effect on votes than answer quality does.
 
So the actual byte number is not a great measure of post quality. In king of the hill, there's really no other consideration
 
1:37 AM
KotH rank does seem more reliable than code golf, yes. There are still some frowned upon submissions that happen to do well but aren't disqualified, but mostly KotH rank seems to be a good indicator
 
This form of rank correlation coefficient seems easy to calculate: en.wikipedia.org/wiki/…
 
also seems a bit crude
 
Although I think this seems a lot more accurate: en.wikipedia.org/wiki/Spearman%27s_rank_correlation_coefficient
 
the rankings could be identical except that you move #1 to last place, and then it gives you a -1
Yes, that one looks fine
 
Since we aren't exactly limited in our computing power, we should use the second one.
 
1:42 AM
Okay, I'll write the input as:
<name> <KoTH rank> <post date rank (initial post, not edits)> <vote rank>
and you write the code to calculate the two rho values
 
Me, write code? What do you take me for, a programmer?
 
8| I thought the task I set before myself was maybe more annoying
 
I'll do my part.
 
i'm starting to make the dataset for rock paper scissors lizard spock
 
There's a Mathematica function called SpearmanRho which does exactly this.
 
1:48 AM
I'm just going to put the values, not the ranking
 
yes, just the values
 
it's easier for you or mathematica to do the sorting and the adjustment for ties
 
Mathematica takes two lists of numbers and gives the rank correlation.
 
How should I include people with multiple entrants in one post?
just include their best one, or include them all?
 
I don't know if it matters.
Include them all, I guess.
 
1:53 AM
Can you use answer number rather than username? Each answer has a unique identifying number (unique across the whole site, not just the question)
 
sure
 
We don't need to use either
but it will help me for the moment, if I'm deleting multiple entries from the same answer
Which way should I do this? I need to make a decision
 
If you're doing this for the whole site, will data.se give you the results by answer number?
 
I think we should do one data point per post.
 
okay, I'll just use their best entry for a given post
 
1:58 AM
okay
 
and delete the others from the rankings
 
I'm pretty much ready.
(Thank you, Mathematica's libraries)
 
I'm going to be a while
 
@githubphagocyte In fact, questions and answers share the same primary key. I understand it's called postid, but maybe the data explorer will say something different.
 
In the meantime, I'll try to see how a significance test would work.
The wiki has a formula for the standard error.
 
2:17 AM
Halfway done with RPS
LS
 
I assume it would be (Rho / StdErr) = z-score -> p-value
 
2:32 AM
// Score, date posted rank, #votes for Rock Paper Scissors Lizard Spock
64  28 18
62  55 1
62  65 2
61  64 1
60  29 5
58  27 3
57  32 2
53  61 1
51  42 1
51  16 20
49  31 2
49  49 1
47  34 3
46  56 1
45  24 4
44  48 1
43  40 4
42  20 4
42  53 1
41  12 2
41  13 3
40  23 3
40  46 2
39  62 1
38  38 4
38  33 2
38  43 2
37  50 1
35  15 8
35  58 1
35  54 1
34  3  4
33  10 3
32  4  8
32  44 1
32  14 5
32  17 2
31  19 8
31  45 2
31  51 1
30  1  3
30  60 1
29  22 4
28  25 3
27  57 1
25  30 2
24  2  7
24  39 3
 
Thanks
 
Same for 3D Dogfighting (evidence against fastest gun):
14 7 11
11 8 2
 9 4 4
 8 6 4
 8 3 4
 4 5 3
 2 1 3
 0 2 2
I'll do the same for 4-man standoff later
 
I expect the FGITW effect to have a larger effect on more popular questions.
Not just a linear scaling of more views -> larger effect, but something bigger.
 
Yeah. 4-man standoff is large enough to have it. I'm curious to see what you get from the RPSLS data
 
There are two types of viewers - "regular" and "visitor"
 
2:40 AM
right, and big questions get visitors who are more likely not to be as thoughtful with their voting
 
Basically, the regular people read everything and don't really create the FGITW effect, but the visitors just look at a particular answer/thread.
Yeah, that's what I'm saying.
 
3:01 AM
@EricTressler Are you still here?
 
Okay, SpearmanRankCorrelation[dates, votes] // N gives -0.766193
 
well, that's extremely strong
 
The negative sign probably needs to be ignored, based on which way you ranked it.
 
no, it's a negative correlation, it's meaningful.
 
3:04 AM
SpearmanRankCorrelation[scores, votes] // N gives -0.250174
 
you're right, though, if you change the date to most recent -> least recent, it would be 0.766
That's just weird
it's probably because a lot of the really good entries came at the end
With knowledge of other people's strategies. I guess this isn't a perfect experiment; the most important number is the first one
 
No, I think it supports the earlier-> most votes.
 
there should ideally be 0 correlation between dates and votes
 
Because earlier posts have lower dates but higher votes.
So a negative correlation supports the FGITW.
 
I'm talking about (scores, votes)
 
3:08 AM
SpearmanRankCorrelation[dates, scores] // N gives 0.243443
The (dates,scores) means that the better posts came later in the competition.
(score,votes) is negative because the highest-votes posts are the earlier ones, which have lower scores.
(dates,votes) strongly supports FGITW
 
right. Yea, this all makes sense. Well, now we have some data.
 
The real problem is (votes,scores) being negative (!)
By the way, standard error is 0.0790625.
This means, in terms of proving the FGITW, we have a nine-sigma level of evidence.
In particle physics, major discoveries are announced with 5-sigma evidence.
 
What was this for, petitioning SE to give us a random default ordering?
 
I guess that is what this could be used for.
 
Well, if one of the moderators takes an interest, I'll make some more datasets; it took long enough that I don't want to do it just out of curiosity :)
 
3:18 AM
Regarding the fact that posts with more votes actually perform worse then posts with few votes, that was at a 3-sigma level.
3-sigma = 0.16% chance that a purely-random sample would produce a result this strong.
 
That's not causal at all, though, that's a side effect of the best entries coming late
 
The real cause is the FGITW.
Early posts -> more votes. Later posts -> better performance.
I think the default should be random sorting.
I'm creating graphs of the data right now.
This will be the last thing I do tonight.
Bye for tonight.
Thanks for your help.
 
Yours, too. 'night
 
 
1 hour later…
4:40 AM
Calvin's hobbies posted another question
2
Q: Take every Nth character to output N (christening source-layout)

Calvin's HobbiesWrite the shortest program possible such that when you take every Nth character (always starting with the first character) and run the result as a new program, the output is N. This must work for N = 1, 2, ..., 16. Another way to say it is, if you remove all the characters except for every Nth c...

 
 
2 hours later…
6:51 AM
@EricTressler I've seen a few comments over the years where people have upvoted a GS program because it contained a smiley rather than because of its shortness or cleverness.
 
 
3 hours later…
9:46 AM
@EricTressler Moderators can't change the default sort order on the site. I'm not even sure if the server-side code supports it; it may have to involve developer effort.
 
 
3 hours later…
12:51 PM
now, how awful does my new picture look scaled down
 
1:02 PM
@VisualMelon Uh, what?
 
@ChrisJester-Young I think he means his new profile picture.
 
I've used my expert photo editing skills and super high quality phone camera to make a new profile picture, and I can't work out how to make it appear in Chat
what ProgramFOX said
 
@VisualMelon It will appear here after some time. Your previous one is still cached.
 
ok, thanks
 
1:08 PM
@VisualMelon With any luck, bad enough that it's not obvious that it isn't a melon ;)
 
hehe, that's the joke ;)
I actually really like lemons, and I really don't like watermelons, which is what people usually end up calling me
(but those nice yellow melons are really nice)
 
Galia?
 
possibly, I forget what they are called, I just buy/eat yellow melons when they are available, hasn't gone wrong yet
 
@EricTressler @PhiNotPi note that even with completely random ordering of answers, early answers will still have more than their quality based expected number of votes, due to being seen by more people.
 
1:23 PM
I think pagination is a bigger issue than ordering--it'd be nice if the answers/page limit could be increased or removed entirely
 
@ChrisJester-Young thanks for clarifying that answers and questions share a primary key list. As long as we can filter questions and answers separately that post-id should still give a convenient unique label if we were to generate a list of all answers and their associated question.
@FireFly that would indeed help with the ordering, but there are at least some questions where that would prevent the page from loading in a practical length of time. Mainly the image based popularity contests and code challenges.
 
@MartinBüttner colour me interested (google isn't being helpful here)
@githubphagocyte hmm, on that point I wish there was a way to create an expandable box that starts out collapsed (like "spoiler" tags in some forums), to put things like images in to not take up such large amounts of vertical space. It'd be very useful for image-based questions, I think
 
1:38 PM
@FireFly I'd love to see that - and I'd be inclined to include just one small preview image with all the other images and larger sizes behind the expandable box, but in a popularity contest it might be hard to enforce that as everyone wants their image seen first...
 
1:58 PM
Visual Studio's file diff doesn't differentiate between Linux line endings and Windows line endings :-/
It looks like I changed nothing
 
@Rainbolt does it have a setting for that or is it just fixed as it is?
 
I'm not sure. I'm checking in my change now to see if it will actually recognize that something did indeed change, even if it is invisible
I have a GO statement in my SQL that won't run without being following by a proper newline
You know what would be cool? A browser that normalized line endings.
If Chrome said the world was going to work with Linux newlines (and the metric system), then perhaps the rest of the world (Windows and the US) would just give in.
 
A web browser? When, during file upload?
 
Just for display
So that after enough people complain that copying and pasting is buggy, Windows will cave
 
Oh, like that
People would just blame the browser
 
2:02 PM
Only Windows people would blame the browser lol
I see your point. The only person that loses is Google
 
Yeah, they'd only move to IE :p
I'd expect copying from a browser window to use the system's newline convention regardless of the HTML file's actual content
 
Someone needs to write a worm then that goes through every hard drive everywhere and replaceAll("\r\n", "\n");
It has to be the most viral worm ever created
We'll call it "Philanthropist"
 
Poor binary files
 
my only issue with \r\n is that it means that newlines in my texts are a char longer
 
2:21 PM
I kind-of like \r\n from a terminal point-of-view, but having a single character delimit lines makes more sense from a file-content point-of-view
 
@FireFly aye, but for real documents you should be using paragraphs and breaks, not line feeds or carriage returns (Unicode has it's own new-line char I believe, not sure why)
 
@Geobits added a scathing comment so you don't have to
Hmmm... maybe too scathing.
 
@VisualMelon The "Record separator" from C0 could certainly be abused for that.
 
it's a shame that CSV and TSV are things, when ASCII has such good support for separators
 
2:37 PM
@githubphagocyte It's too bad he just deleted it. I think it really would have been a great answer if he could close the loop somehow and connect the clipping edges.
 
@Geobits and cap the frequency to prevent the line touching to give areas of dense black that give it an unfair advantage over other answers...
@Geobits yes I'm still hoping they will go back and make it fit the question - I'd love to see a version that fits. There's a lot wrong with it but clearly this is someone capable of fixing it all.
 
Yea, it definitely had some problems, but it was a nicely "different" take on it. I'd like to see a properly done spiral.
Of course, that would mean a double spiral probably, to fit the closed loop.
 
I was trying to go for complimenting the answerer while making clear the laziness in the answer is not acceptable. I think I misjudged the person though, and they took note only of the criticism...
Yes a closed double spiral would do it - even if it means putting the image on a somewhat larger canvas to contain the unused outer loops.
 
Put it on a circular canvas and it would look good on my wall :D
 
@Geobits brilliant. Write it tailored to a circular canvas and then we can all use the code on our favourite images...
I might also have been biased in my comments by the fact that I'm working on an answer to that question myself, and taking pains to ensure it fits the rules.
 
3:18 PM
@githubphagocyte I don't think you overreacted. The author of that post got -3 +1 which totals to +2 reputation (I think) for an answer that missed the only defining rule of a six sentence spec.
 
+10 for UV, -6 for DVs, so +4 total (before deletion, that is).
 
@Rainbolt thank you :) They've deleted the question now so the rep has gone back to the repvat
 
@Geobits I can't see up and downvotes (only total) and I wasn't sure how much rep a downvote cost you, so I was guessing on two fronts lol
 
You should be able to see vote counts with 750 rep. Are you using the app, where you can't see them?
 
I'm using Chrome Version 36.0.1985.143 m
 
3:27 PM
Me too :p
 
Hmmm
 
What happens when you click the vote count? Nothing?
 
It separates lol
Wow
 
New superpower!
 
Can I just click myself and become a mod?
Can I click users on Board and Card Games to make them disappear?
 
3:29 PM
I've heard that if you click yourself too much you go blind.
 
If I click on a cookie will I make a cookie?
 
@Rainbolt trying clicking on the middle of the mousepointer
 
I immediately Googled that and arrived at this incredibly helpful tutorial
 
@Rainbolt do you have to post a lone link with no introductory text?
 
@githubphagocyte As far as I know, yes.
Perhaps Wikihow is missing some of the information that is scraped when building a mini preview window.
 
3:34 PM
I'm amazed that the link exists (not because it would never be helpful, but how would someone needing that knowledge ever get to the page??)
I'm also amazed at the amount of advertising on that page. Who did they think they would reach?
 
That is the work of a person being paid by the word.
 
Only certain sites are special enough for oneboxing.
 
@githubphagocyte Jeff Atwood spoke about that. He's for minimal advertising, and refuses to allow moving ads. That could change I guess, since he's not working for SE any longer.
Oh, that explains a lot
Between that list and pictures, nearly every link gets oneboxed.
 
Special allowance for xkcd comics...
 
If only PBF had been more nerdy
 
3:39 PM
PBF?
 
Interesting. The top voter on questions is someone who has never asked or answered one.
5
 
@githubphagocyte Perry Bible Fellowship
 
Pretend Boyfriend?
 
Yes, Perry Bible Fellowship
 
3:41 PM
I'm unfamiliar
 
Peanut Butter Fudge!
 
Oh wow. I was actually joking about Perry Bible Fellowship
 
I, uh. That's not possible
 
It was almost last on the list of potential matches for PBF. The highest rated match was Play-By-Forum (game)
 
3:44 PM
Oh, I see
Well, I like xkcd too, but there are other good webcomics. Xkcd gets a little too much love sometimes
Here, like this: pbfcomics.com/259
 
@PeterTaylor I've noticed that user before. Even stranger, he seems to like voting around 1000 times, give or take a hundred or so, independent of how much rep he's got on the particular site.
 
Maybe to get the gold badge
 
Yes, I think he gets the Electorate badge and moves on. Not seen for over a month.
 
Electorate only needs 600, though. He's not in danger of having less than 25% on questions anywhere, so it's just weird either way.
I mean, yea, but why not go for other badges, too? Even with a few minutes of time he could farm some "easier" badges he doesn't have.
 
Spite?
 
3:50 PM
Maybe it's a bot?
 
He knew you would find out, Geobits, and he's doing it to annoy you
 
@EricTressler That's been my assumption all along. It's clearly targeted.
 
That could be a competition - write a bot that goes the longest undetected as a user. It would need to vote in an apparently human way (presumably based on following voting patterns rather than really understanding what it's voting on).
I guess it would be too easy though - just vote on the first answer to every question and you'll come across as human
 
Even better, the first bot to become a site mod wins
5
 
@EricTressler 3 stars in 17s - is that a record?
 
3:54 PM
You can't upvote unless you have a bit of rep (association bonus is more than enough), and you can't downvote without 125 rep (association bonus is insufficient). So the bot would probably want a bit of bootstrapping and then upvote.
 
@EricTressler I'm going to need the results of this competition for SkyNet integration.
 
I can't think of another forum that would run something like that, but it would be kind of fun. Well, another stackexchange forum anyway. Somethingawful etc. might do it
 
@FireFly Whenever you you place a stone, you look along along the four horizontal/vertical directions for the nearest neighbour. If there's a a stone of the opposite colour it gets pulled in until it touches the new stone and if there's a stone of the same colour, it's moved away as far as possible. basically like attraction and repulsion of charges.
ugh, I haven't been in this chat all day, and I still wasn't any more productive :/
 
Somehow I know exactly what you mean with that latter line
Hm
As if Go wasn't complicated enough already to read
 
Haha, that's true. Although I find the end game a lot easier than reading anything in chess. :D ... but that Go variant is quite fun on 9x9
 
4:04 PM
Oh, yeah, I can imagine
 
@MartinBüttner maybe you should conduct a series of experiments, where you allow yourself a set percentage of time in here, and see if there is a percentage which maximises productivity.
 
@githubphagocyte you mean something like a Ballmer peak?
 
@MartinBüttner lol I didn't know that term but yes, exactly
 
^_^
/me notices the title text works with inlined xkcd's.. interesting
 
4:08 PM
I think there's actually a little bit of truth in it. In that maximum programming productivity is not achieved in a sober state. But I think if you drink a little, there could be a regime where your productivity is increased due to higher enthusiasm and less attention of peripheral stimuli, but you don't yet make so many more mistakes that it would be detrimental.
@FireFly Only when you actually paste the URL to the xkcd page (and not the picture). Then the chat oneboxes it.
24
A: Which links and sites are handled specially in chat?

Juha SyrjäläThe current list of integrated (we call this onebox, or oneboxing, ala search engines) sites is: Stack Exchange sites: Questions / Answers / Users / Comments Stack Exchange Chat: Messages / Rooms / Bookmarked conversations Area 51 proposals Posts from the Stack Exchange blog, the Server Fault b...

 
Aha
 
@MartinBüttner Anecdotal evidence is that although programming productivity may be improved by the use of certain chemicals, trying to maintain the code while sober is rather painful.
 
The obvious solution is to never be sober while writing or maintaining.
 
I think the point is that either you commit to never coding sober or you commit to always coding sober, but you shouldn't try to find a compromise solution.
 
4:12 PM
Or shunt the maintenance work off on sober grunts :P
 
/me notices @FireFly is an IRC user
 
4:29 PM
0
A: Proposed Question Sandbox - Mark XIV

VisualMelonProgrammer's Garden Sandbox note: This nonsense at the start might be too long... The new office unit for that new startup you are working for because the boss knows someone who worked with your uncle back in the 90s has, as part of it's grounds, a small rectangular flowerbed! However, being pr...

 
@PeterTaylor re the trig challenge. You can only plot a finite number of frames. Just shift the angle slightly off 0, because you're not gonna hit the multiple of Pi/2 anyway.
 
@TheDoctor there's somebody with your name on multiplayerpiano o;
 
@cjfaure No offence to TheDoctor but it's not the most unique of names, I guess. ;)
 
@MartinBüttner xD
indeed, but I like hoping.
"Wait, those are the same words that person A said on site B! It must be them." Then I proceed making references and seeing if they get it.
 
@MartinBüttner You're right: the best solution is to only plot odd multiples of pi/4. Then it's only necessary to draw half a frame and reuse it 7 times.
 
4:37 PM
:D
 
if math has pi in it, then i don't understand it
 
@cjf it's not me
 
@TheDoctor aaaand today's wild dream is gone.
 
i'm at work!
and on irc
 
does your job have to do with programming?
 
4:41 PM
yes
not "fun" programming tho :\
 
hm, uh, your boss might not like this site showing up on his router. lol
unless, of course, my space-age super hard-drive code-golfing entrepreneurial plan was implemented.
 
its been a while since i've been here
 
yeah :P
 
i barely ever answer questions
 
;-;
 
4:48 PM
but i have Dogecoins!
 
@TheDoctor wow
 
@TheDoctor Indeed
 
I'm working on the full-blown coloured animation now. That seems more fun. :D
 
@MartinBüttner Related: I bought Pyxel Edit today.
 
You'll probably enlighten me in a second about what that is.
 
4:54 PM
Sounds like a pixel editor to me :/
 
I've used pixel art programs before, but to be completely honest, all you really need is paint or gimp. Drawing pixel art is really basic (don't read that as "easy").
 
Or Minecraft!
 
@Geobits yeah, but the tiling and animation features!
 
I guess an editor is handy if you want to draw tiles that wrap
 
4:56 PM
An animation editor is a nice addition, but I believe I found a nice gimp plugin for that at one point.
 
user image
2
 
@Rainbolt where'd you find that? :P
 
@Rainbolt To be completely honest: That's ridiculous :D
 
one of my friends is making a pixel art exhibition in minecraft
with every single pokemon ever
THAT'S ridiculous
 
@cjfaure Please send a link when done. My son's brain will explode.
 
4:59 PM
@Geobits ;3
 
Are there free and free IDEs available tailored to golfing?
 

« first day (1298 days earlier)      last day (3531 days later) »