« first day (186 days earlier)      last day (3537 days later) » 

9:00 PM
it's hard to invent a clever mapping though
"Guess the card I'm thinking of! Hint: it's red. For every wrong answer, I'll post another hint. Remember to post your answer in spoilers!"
@Emrakul oh boy
that sounds like a chat room
not a Q/A
@AJ In fairness, most of our questions aren't strictly 'questions,' either.
@Emrakul fair
i feel like the spaghetti questions were mass produced, but not neccesarry mass-producible, ironically enough
if not for the flood of derivatives, I would not have though the first couple to be a template
9:01 PM
ok, what are the "spaghetti questions"
guess-the-password questions
what's the tag?
[tag: password]
security to the party
thirty-six of them
I have to admit, when I first saw "[Part 26]", I thought it was a joke or something.
i had planned to satirize them that way until i saw my number was actually there
is there a more robust way to say that a question must be unique?
i don't want to ban new twists on known puzzle, but do want to ban pointless copycats
maybe one could argue that these questions are a chestnut at this point
9:04 PM
yeah, I'd argue those are a class of puzzle that is mass producible. Starting from a basic rule (the solution) and obfuscating
the solution still doesn't require particular care
i'm just having trouble expressing why these questions are low-quality in a concrete way that doesn't rule out real puzzles by being vague
puzzles where you're supposed to find a rule mapping an input to an output is a quite general class
the means of creating the problem is the key I think. If you are starting from a solution and working backwards, it likely isn't good
at best, it could have a question discussing how you solve the category of problems
but it's really just figure out the missing data
do you mean, here, arbitrarily choosing input/output pairs, then fitting a rule to them?
rather than coming up with a rule
or in the case of passwords, it is start with a rule and arbitrarily make data for it
"Make rules to fit this data"
9:07 PM
like with the animals password, the solution is how the numbers map to letters and you are then simply giving several examples of input and output
why doesn't this also rule out this though? puzzling.stackexchange.com/questions/4784/a-fragmented-puzzle
i guess you're only given inputs and not outputs, but the idea of finding a linking rule is the same
because it requires more than simply application of the rules
Uniqueness, I guess. Different logic is involved.
actually, i have a better example, one minute
if my logic is a = 1, b=2, c=3, I can choose any input for my pairs
do you have to solve a puzzle to invent the puzzle could be another way to put it
for the favorites one, you need to be able to form a word from words that fit a particular pattern
it would be even better if the words had additional relationship between them, but forming a word from words of that class is atleast reasonably challenging, even if it can be solved with a dictionary list of those types of words
it's still a puzzle (even if an easy one with proper resources)
9:10 PM
it apparently was extremely constraining, the writer has to find the few examples with a computer search
ok, i see your point about constraint
does it apply to riddles also though?
would you say finding clues and fitting it to meter/rhyme is a constraint?
I'd say that in general, if you can trivially work backward from the solution to the question, then it is mass producible
I'd have to see an example, but possibly
ok, so enforcing a ban on mass-producible puzzles sounds good
what else can be done on the quality front?
Encouraging downvotes
so that's hard
Our downvote to upvote ratios right now are... poor.
9:15 PM
Poor in a way that's shocking.
i see bad answers with many downvotes but even more upvotes
i think the issue is that site regulars are outnumbered by the masses coming in from HNQ on popular questions
@Emrakul well, not so shocking after looking at that one question you linked
but certainly problematic
@AJHenderson Fair. But, well, I'm curious, do you know what the up/down ratios are on your sites?
(Maybe post in TL?)
(I'm curious to obtain at least one point of comparison.)
9:18 PM
@joez hi
@xnor: I'm here
we've been having a long discussion about what's on topic
about your post, i want to talk about the middle ground
I saw the chatroom for that one question
although I didn't realize you guys were talking about it in here too
of course, the spaghetti questions are bad
but i think mathematical rigor is too high a standard
those are not language-based, they are dictionary-based
9:20 PM
so the fact that a regexp can verify it makes it OK?
hold on a sec
you're right that I'm trying to rethink the "mathematical abstraction" standard
because you're also right that it feels too tight
ok, cool
i think i agree with what you're trying to express, it's just hard to pin down the boundary
but in that specific case, yes, I would probably make that argument
in the absence of a better criterion, anyway
let's say that there is another solution for that mechanism
but it uses words that exist in SOWPODS or whatever and are very obscure
that a computer would find but a human would not choose over the intended solution
would that be OK?
9:23 PM
alternatively, one could say that the mechanism, while verifiable, is a very human one
maybe we should post this as a puzzle and see what answers people get
post what?
it can be kind of like my non-meta meta question on moderators ;)
yeah, I think the mathematical abstraction thing is more like a rule of thumb
that's probably what I'm thinking of using it as
9:24 PM
if your problem can be mathematically abstracted (including converting the problem to a computer doing something to a SOWPODS database)
then it's probably okay
definitely more likely to be okay than something that isn't
Q: How can I encourage more participation in a community that has been active in the past but is going through a lull in activity?

AJ HendersonSometimes active online communities can have a dry spell of activity. They've been active in the past, but participation has dropped off and it can become a self fulfilling prophecy that nobody participates because participation is low. How can I help boost activity in such a community and get ...

that's me showing my sense of humor
for me, the condition I want to express is that once you know the solution, it's clearly the right one, not alternative solution is nearly as good, and you can either solve it or feel like you could have solved it
mathematical problems are ideal for this, as the correctness condition can be exactly specified
for riddley ones, it's very hard, but i think possible, to ensure that the intended solution stands far above the rest
I think the reason that I exclude riddly ones specifically is because there's no way to tell just from looking at the question that it's supposed to have one correct answer, or that there's an objective criterion that the answers are valid
ref. multiple subjectively-correct close reasons
^ exactly
9:27 PM
@Joez earlier in the discussion, i proposed a requirement that questions post an answer in spoiler
so that those who peek can see if it's bad
for riddles?
That makes me a little uncomfortable, because it raises the barrier-of-entry for new users.
but then why would anyone bother to actually answer it?
because riddles are fun?
that's the problem I have
no, like
posting an answer in the question itself means that nobody can actually post the answer as an answer
9:28 PM
this is coming back to the question of the site goal
which is further from the intention of the SE format than I'm willing to deviate
i think people will still solve questions for fun even if they could peek at the answer
is your concern that there's no way to decide what to upvote then?
or that it takes too much self-control not to look?
no way to decide what to upvote, and that the correct answer as an answer would just be parroting content already in the question
ok, i see
even on PCG.SE, nobody ever puts the answer in the question itself
9:31 PM
If someone posted an answer in the question, I'd edit it out and post the answer as Community Wiki.
@JoeZ. what do you think of that idea?
oh hey, emrakul, looks like up/down vote ratios is queryable from data.stackexchange.com
Does it include deleted votes? o.O
Actually that probably makes little to no difference.
9:33 PM
that's interesting
but somehow I feel like it would encourage people to post low-quality riddles and their answers (even if the answers are community wiki), and that we'd get the same glut of them that we've been getting
like, I understand that posting your own questions and answers is acceptable on SE
i we're willing to close riddles whose answers are ambiguous, that could handle it
but riddles? I'm not sure
I would say the riddles with ambiguous answers should probably be deleted instead
if we're going to mandate that riddle-askers have to post their own answers as community wiki
Creating a riddle, posting it, and then self-answering it may not be off-topic or against Stack Exchange policies, but it's still probably a waste of bandwidth.
(i.e. that's what I'm getting at)
and you're saying it's the posting of the answer that makes that the case?
if someone else answers it, it's not a waste of bandwidth?
9:38 PM
no, it's not the asking/answering itself that's a waste of bandwidth
what about PPCG then?
but more the fact that asking/self-answering is supposed to be for useful information
on PCG, there are generally multiple valid answers and the accepted one is whichever one is best
I've self-answered many of my own problems on PCG
Whichever one is best on PPCG must be selected based on objective criterion.
9:39 PM
adjust the offset as desired
Those objective criterion are public knowledge
@AJ Coooool, thanks.
i see that, maybe i'm just misunderstanding "waste of bandwidth"
PPCG poses challenges that are useful to nobody
eh, 'waste of bandwidth' is a little harsh of me
nobody will use that code
9:39 PM
data.stackexchange.com is one of those fun playgrounds that probably mods make more use of than most other people, but there is still a lot you can do in there
nobody will google later and find something useful on PPCG
but people do it for fun
yes, someone asks a "question" and someone "answers" it, but there's no real use that comes from it
is that OK?
PPCG is a huge exception to the way Stack Exchange operates
We should avoid basing ourselves on the way PPCG operates under all circumstances.
right, but I think that this is a good model for "challenge-answer"-type questions
9:41 PM
actually, that's exactly what I'm trying to do (base us on PCG, but for regular puzzles)
and people seem to believe that posting specific challenges is on-topic
but part of that is having the same quality standards that they do
given that PPCG seems to be high-quality and successful, I'm not convinced by argument that would also apply to disallowing PPCG
if we're going to go for being a regular Stack Exchange website, then the Puzzle Theory advocates are right
we can't allow puzzles as challenges
the only reason I can see us allowing puzzles as challenges is specifically because we're a non-programming analogue to PCG
while i'm open to considering the option of banning challenges (as much as it pains me), PPCG is the alternative model for challenges
9:43 PM
If we look way back when, the site definition phase didn't really include many questions that were strictly puzzles.
umm, the second question
with 14 upvotes
"Given two hourglasses, one lasting 4 minutes, and one lasting 7 minutes, how do you measure 9 minutes precisely?"
Position doesn't matter too much. Look at the rest of the questions.
yes, and the next one was the 10th one
not very many
I'd be okay with banning challenges as well if it'll help us regain our sense of purpose, but in that case it wouldn't be the site I signed up for
Like I say, there were a couple, but they were few and far between.
and I'd probably leave
9:44 PM
@emrakul the other one (camels and bananas) was closed a duplicate for covering the same ground
(well, leave being active, I mean)
Iunno. Looking back to Area 51 definition isn't necessarily the best approach, but it's important to look at what the site was intended to be.
the idea of the definition question is to intentionally make a variety of questions
(I'd probably drop in once in a while to ask an offhand question like I do many other SEs)
"Because both questions are equivalent from the point of view of "which questions should be asked on the proposed website": both of them are puzzles. I think that just one question per type is enough, otherwise we could just open hundreds of questions, each of them proposing a different puzzle."
now, i'm with you that challenge questions were not intended to be such a majority
9:46 PM
yes, questions about puzzles are supposed to be an important part of the site too
Questions-as-puzzles are different than asking a question about a puzzle.
"I've gotten this far, and these are my thoughts - where do I go from here?"
they are
and I specifically said in that new meta question I posted that questions about puzzles aren't what I was getting at when declaring certain things to be off-topic
@emrakul: this one
but in order to cut down on the number of challenge problems being asked, there was my suggestion for what to exclude
okay, uh, brb, I'm going back home for a moment
Hmm. I think we may be manufacturing the people who think puzzles shouldn't be asked here.
9:50 PM
Also, in ref. to your answer, @JoeZ: I think it's possible to disallow questions as puzzles while still allowing questions about puzzles and working through them.
manufacturing means inventing them, or producing them?
@xnor As in, who on the site thinks specific puzzles should be totally disallowed?
(wow, this answer was prophetic: meta.puzzling.stackexchange.com/a/105/4551)
but, yeah, looking at that meta question, the top voted answers are "i guess, probably"
One of our most upvoted questions was simply, "why does this solution work?"
which one?
9:55 PM
Q: Why does this solution guarantee the puzzle is solved?

WendiKiddI found this puzzle online: On the top floor of a castle lives a princess. The floor has 17 bedrooms arranged in a row. Each bedroom has doors connecting to the adjoining bedrooms as well as to the outside corridor. The princess sleeps in a different bedroom each night by opening the door to ...

oh, yes, that was a great question
it was from May 28 though, a different era
looking through your top posts, this type of thing seems to have been more common in the early days
Yeah. That's the sort of site I imagined this growing into.
Of course, when it comes to the site's ultimate future, I'm just one person. The users drive the site.
But it was different then than it is now.
i really do like your questions and answers
i'd like the site to be like that
9:58 PM
Then we need to push to get it back that way.
It's a totally feasible goal, but we have to actually work towards it.
going back to private beta was shot down, but is there some other action that would signal a reset lever?
something that would tell puzzlers who already left, "this is different, give it another chance"?
of course, posting high-quality question would help, but i'm worried that they'd get lost
and not be seen by those who'd most appreciate them
A lot of the people who don't post anymore still watch the site.
Or come back and visit occasionally.
Also, there are always new people incoming.
10:01 PM
that's true
PPCG chat periodically does puzzles in chat
Q: Should we have a "weekly puzzle"?

EmrakulI was thinking of ways we could further engage community members, existing and new, and I had a thought: maybe we should run a 'weekly puzzle' chat event? This could be a good forum for discussion of questions that may not necessarily be on-topic on the site for being too broad, or could be a fo...

actually, PPCG has, funnily enough, had a lot of discussion that should be in this chat room
i hadn't thought about that, but the idea of community puzzle solving is a great one and could be a big draw to the site
The problem we ran into was generating content.
people wouldn't make puzzles?
oh, shoot, i gotta run, TTYL
No worries! Seeya!
10:23 PM
@Emrakul I know that; I specifically said in the meta question that questions about the disallowed puzzle types would be fine
[@ a really old message, so...]
Yeah, I just reread that. Makes sense!
11:02 PM
i have a puzzle-hunt style puzzle i've previously written I could post to test the waters on puzzles that are more involved and are targeted more at puzzle enthusiasts
is it worth a try?
@Emrakul I'd have no problem volunteering a puzzle for that. If you do too and xnor does as well and we find one more person to put one in, then we could have weekly events for a month. That would be long enough to test out the concept and see how well it works.

« first day (186 days earlier)      last day (3537 days later) »