12:18 AM
@Doorknob Eric Tessler and I did some measurement of the FGITW effect.
It was interesting.
Most people know that FGITW exists, but nobody really knew how prevalent it was.

1:06 AM
0

Measuring the Strength of the FGITW Effect Although this answer isn't a suggestion for a solution, it should provide a better understanding of the problem. A proper understanding is necessary to create a solution. The FGITW effect ("fastest gun in the west" effect) is the phenomenon in which t...

@PhiNotPi check the sentence after the second graph

Tautology is tautological.

Well, I wasn't wrong.

@PhiNotPi +1 by the way

We only collected data on the RPSLS tournament because it took forever to collect the data.

1:15 AM
I have a question about the submission with the most votes: if there's a bounty on the challenge is it automatically the highest upvoted submission that receives it?

@PhiNotPi did you look at data.se?

No, the user who created the bounty can give it to anyone.
I don't really know how to use data.se

@AndoDaan it's only automatically assigned if the user who posted the bounty doesn't assign it until it expires
@PhiNotPi nor me but I'm thinking it could save you a lot of time and amplify your statistical powers

Ah, I saw the bounty on the dogfight expire I think.

@AndoDaan if it was auto-assigned I think it's only at half the amount

1:17 AM
It surprised me that it was assignined right on the due time seemingly without input from the poster.
Okay, thanks.

@AndoDaan there is a 24 hour grace period after the advertised expiry time, then it is autoassigned
@PhiNotPi it lets you write queries that will give you all the results that match from all the questions/answers on PPCG

It is something I should look into.

@PhiNotPi I keep meaning to look into it too...

The negative correlation between votes and scores surprised me.

@PhiNotPi what I'd really like to see is the same statistics but based purely on votes that came in after the final answer

1:25 AM
What could also be used is the number of question views over time.
That way, the effect can be separated from the fact that earlier posts will have more votes anyways.

@PhiNotPi yes it would be great to have some perspective on it. I don't know how much fine grain control data.se will give you but there are a lot of example queries for you to look at to find out.

I think random sorting is a good default, even though I will change my sorting back to votes.

@PhiNotPi I think that would reduce the effect of FGITW but there would still be a certain amount of bias. I can't think of a way of removing that though.

There will always be a baseline amount of bias, but I don't think that can be removed.

2 hours later…
3:10 AM
@PhiNotPi Thanks for typing all of that up

3:27 AM
I agree with your assessment, after reading it carefully. Note that I did have to manually throw out a fraction of the data, and that some of the posts with multiple entries confounded it a bit
They may have gotten extra votes based on one of their 2-3 entries (I think 3 was max per post), which we aren't counting

3:47 AM
marbelous room has been dead for three days. anyone still interested in that project or should I start working more solo and less aim for consensus? @MartinBüttner (can't find @overactor or @nathanmerrill or @es1024 or @cjfaure in either room lately)

0

Description As it appears the sandbox on meta is here to stay, I've written a very simple userscript that will allow you to collapse sandbox answers to reduce clutter. (This may be able to be replaced when the fancy sandbox viewer thingy is finished, but for now it'll make navigating the sandbox...

@Sparr , @cjfaure and @nathanmerrill and @overactor have been around
When I have, I mean
Approximately 6-12 hours ago

4:27 AM
@Sparr I'm pretty swamped right now. I'm also just about to start school with 18 credits and 20 hours of work, so I don't think I'm going to be any less swamped (unless I pop in here during class), but we'll see
I am still interested in the project

I found him, @Sparr. Do with him what you will

4:51 AM
@NathanMerrill I think I'm going to choose a path down the potential spec road and let my implementation diverge if it goes that way, then re-converge once we find some common discussion time in a week or a month or three, like I did last time we had a good meeting of the minds.

I've been around but a bit more busy lately. If you want input, you can always ping me.

5:09 AM
@overactor can't ping you if you don't visit the room every day or two

6:06 AM
@Sparr I'm usually watching it though

6:19 AM
weird. I couldn't tag anyone there today or yesterday, and my client says I'd been the only one there for a couple of days

It could be that I was never there yesterday
Can you ping me there now?

6:48 AM
@Sparr still interested, though i've been busy recently

7:21 AM
@MartinBüttner Give it input of the maximum number in the range (in unary, obviously), and add a leading or trailing .{0,D} where D is the difference between the max and the min.

7:39 AM
@PeterTaylor I thought about something like that last night after I signed off. Alternatively, you could use the minimum and maximum number, separated by a comma.
In any case, it sounds doable, but I think I'll switch to .NET again

@MartinBüttner regular-expressions.info/subroutine.html might be the simpler way

hm yeah, thought about those as well, but I think .NET would make it conceptually simpler
because then I could just do the prime check right after the initial (11+) by using an RTL lookbehind
in particular, by "simpler" I meant something less exclusive to PCRE (because the current solution needs branch reset and subroutines are also very rare). But of course, switching to RTL lookbehinds isn't any better, just a bit more concise.

2 hours later…
9:24 AM
@Sparr apparently, Manu improved on his Crossfire planes, I'll run some trials soon

1 hour later…
10:26 AM
@Sparr I'm definitely still interested, but I don't currently want to spend hours on design discussions

@MartinBüttner do you use IronRuby?

nope

is there a .NET style regex lib for proper Ruby then?

no, the .NET flavour is pretty unique, but Ruby still has one of the more interesting flavours
why?

you said above chat.stackexchange.com/transcript/message/17239625#17239625 I was wondering if you were able to use it in a Ruby environment

10:29 AM
oh. I don't know if IronRuby uses the .NET flavour to be honest. But I meant I'd switch to using C# or something. but I didn't really think about the actual language, because I was looking at a pure regex solution

fyi, it looks like the regex code for the IronRuby implementation is still in a directory called "Experimental" :D

hehe, presumably you can just use System.Text.RegularExpressions though
I think I have IronRuby installed... I don't know why I have all these things

@VisualMelon yeah right, so that would give you .NET regex then
I wonder why you came up with using .NET regex in ruby though :D

@VisualMelon I like that your avatar is now a visual lemon

10:36 AM
@overactor lol we discussed that at length yesterday
somewhere around this message:
16 hours ago, by Eric Tressler

I do hope that's intentional

yes it is

Your avatar is now almost as unnerving as @githubphagocyte's

21 hours ago, by VisualMelon
hehe, that's the joke ;)

@MartinBüttner btw, I'm stuck on teh dark room of the temple of quetzalcoatl
did you manage to get passed it?

10:51 AM
but @overactor, I thought my new more central avatar was an improvement...??

@githubphagocyte Oh, that's a lot less unnerving indeed
I don't like it

codegolf.stackexchange.com/q/36407/194 : OP says in a comment words to the effect of "I'm not sure what the correct result for that input would be because my spec is too ambiguous", but somehow it can't get a 5th vote to close as unclear.

@overactor no that's the one I stopped on

@VisualMelon In order to run GolfScript under Windows?

@MartinBüttner Did you have an idea for it?

10:54 AM
Any ideas for what my avatar should be?

@PhiNotPi, a $\varphi$ crushing a $\pi$

@PhiNotPi a butt?

A stylized φ seems like the obvious choice

@overactor I think they were asteroids? Maybe replace the names by their asteroid number.
@PhiNotPi I was going to suggest a pentagram covering a circle, but that might be misunderstood.

@MartinBüttner That's what I was doing, but I'm not recognising a clear pattern in the numbers

10:57 AM
Highlighting the digits of phi in order in the digits of pi wouldn't fit well: the first 1 after the first 6 is the 38th digit.
You have to go to 101 digits of pi to get the first 8 digits of phi in order

@Sparr, Crossfire just kicked your butt in a dogfighting tournament, check the comment I left on your entry

So I tried following people's suggestions on the "how do we get more questions" meta and have beaten an old code golf. Let's see if that gets any attention. :D

@overactor Worse.

@PeterTaylor No, what I meant was a picture of the symbol pi, composed of the digits of phi
but that's interesting
@PeterTaylor How does it go on for 100 first digits?
Should be roughly 1000, no?
although, good streaks have less of an effect than bad ones do I suppose

3.141592653589793238462643383279502884197169399375105820974944592307816406286208998628034825342117067982148
^     ^                             ^              ^ ^         ^                     ^             ^^   ^

11:08 AM
@PeterTaylor sorry, I missed your ping, and I don't understand the question

1.61803398874989484820458683436563811772030917980576286213544862270526046281890244970720720418939113748475
^                            ^                       ^               ^                             ^

Are the digits of phi normally distributed?

@VisualMelon I was guessing why you might install IronRuby.

no, I've only be aware of golf script for about 2 months, I had Iron Ruby on my previous machine, and that was 3 or so years ago
maybe I thought I could find a better scripting language for .NET, but frankly I know C# so well I don't bother with anything else these days

I think the normality of pi is unproven.

11:15 AM
@PhiNotPi oh, right

I think that's probably also unproven.

phi is doing better than pi in that there's a purported proof which would show that it's normal relative to base 2. See the mention of Isaac (2005) in mathworld.wolfram.com/NormalNumber.html

@PeterTaylor hah, you got an electorate badge, congrats! I didn't know they were that rare yet.

Thank you.

11:31 AM

sorry that was a very ambiguous sentence

That was an impressive level of ambiguity to fit into two words... :)

Probably gets the blame for any votes cast by users who are subsequently deleted.

@PeterTaylor ah, that might be

11:32 AM
You can see who has voted??

You can see how many votes people have cast

The top upvoter on there has never downvoted...

1

Stripes function RD(x,y){ return x<0 ? 1+pow(round(((x+1)%DIM+DIM/2+1.5*BL(x,y))/(1+10*y/DIM)/5),3) : GR(sin(RD(x-DIM,y)),-1) } function GR(x,y){ return y<0 ? floor(DIM*(1.5+x)/2.5) : GR(sin(1+RD(x-2*DIM,y)),-1) } function BL(x,y) { return x<0 ? 1.5*cos(x*x+y*y) : GR(sin(2+RD(x-3*DIM,y))...

Beautiful stripes

The top question voter has never had enough rep to downvote.

Those are some pretty awesome colour schemes

11:34 AM
We discussed that yesterday.

@PeterTaylor yeah, I noticed it in the transcript/sidebar
that's why I started looking around the top voters
@githubphagocyte I like ambiguity more than most people.

@MartinBüttner that sentence is somewhat ambiguous

oh?
:P

lol
Are both interpretations true?
I can see votes on that page but I can't distinguish up and down votes

depends on whether I "like" people I don't even know

11:39 AM
it took me far to long to work out the second meaning there

@githubphagocyte you can see that on the profiles

@MartinBüttner if I have to look at the profile, how did you know who was the highest upvoter?

well the guy with the most overall votes has only upvotes... so...
(also initially I didn't look at that user page, but used data.SE ... but you can find the same information on that user page, just in a more convenient way :D)
well, I'll go grab some lunch, and then tackle that smooth number regex

12:16 PM
ugh... of course I can't work on the smooth number regex here... I'm on a unix system -.-

well I want to use the .NET regex flavour

I'm not certain, but I think that there is a portable slice of F# which may have the regexes

yeah I can't install stuff on the machine here, and even if I could, I wouldn't just for my lunchtime golfing :D ... furthermore, I was gonna use regexhero.net, for which I'd even need silverlight

no, looks like it would require mono, no idea where I got that idea from
I can't prove the binomial expansion of (1+x)^n :( I am sad
I forgotten all of my A-level maths already

12:30 PM
@PeterTaylor I was thinking about the projection method... how certain can you be that floating point errors don't botch the tiling?

12:41 PM
I'm using doubles. They have sufficient precision that the source of errors would be the cast to int. As long as you use the public API, you should run out of memory before you run into numerical problems.

1:16 PM
@PeterTaylor right, I was just wondering, because any point of the irrational hyperplane your projecting on could theoretically lie arbitrarily close a point of the 7D grid, right?

I think so, yes. Maybe it does need a bit more detail in the numerical analysis.

1:33 PM
Am I the only one who finds this unclear? codegolf.stackexchange.com/questions/36507/…
Don't wanna close vote just because I don't have the patience right now to get my head around it

I don't think anyone will understand it except OP
I was actually hoping from the fact that your only comment was about scoring that you had understood it.

no I was still reading it, but I checked the winning criterion first
I have a rough idea what he's asking for, but apart from lacking a few definitions it seems very underspecified
for instance, it doesn't say when to terminate each "fork"
also what are "intersections"

One of his images seems to be maze generation, the other seems to be artificial coral construction or something.

They also mention 6 dimensions. I assume that they mean 3 in two directions otherwise it's even more underspecified.

lol yeah

1:39 PM
I think that's enough to close as unclear - the exact interpretation of what is required can be left to the OP

did he do that 3d animation thing himself?

sounds like he did

I'm sure there's an interesting challenge in there somewhere
He can't post in sandbox yet, can he?

he can
he's got an association bonus

Probably, although it depends on how much he ties implementers' hands into doing it one way.

1:42 PM
the only thing it doesn't count for is protected questions AFAIK

does that count for posting on meta?
oh, right
Put it on hold and tell him to put it in the sandbox?

In progress.

There, posted lots of sandbox-related stuff on meta that I've been working on for a while. (Funny how even though I'm a moderator, the site nailed me with a "You may perform this action after 60 seconds" and then a CAPTCHA from posting 2 meta posts in succession.)

:D
is that because you used the same title twice?

With bots competing to become moderator you can't be too careful

1:46 PM
@MartinBüttner Ummm. Whoops. I dun mess up

@MartinBüttner It is
Oh, stop your silly oneboxing @NewMetaPosts

Hm.
At 100 answers, and 2 active sandboxes... is there still any benefit in retiring the sandbox?
2

1:52 PM
@MartinBüttner Good question. It is possible that we could raise the limit to something like 150 or 200 answers (although it's clearly not feasible to keep one forever, unless we're willing to delete posted challenges, which is something we don't seem to want to do). You could post an answer to the meta post if you want.

Why is it not feasible to keep one forever?

"It is possibly" - wow I'm good at grammar. I just woke up :P

In fact I'd have no objections against deleting answers in the sandbox. Who follows those links anyway?

@MartinBüttner Because eventually the sandbox will be cluttered with "challenge - POSTED" answers, and it'll take forever to actually find the "real" posts while looking through page by page.

I'm intending to post an answer advocating to not retire them at all, but I thought since we're "all" here, I;d check your arguments in the chat first.
@Doorknob they're on top.
(sort by activity)

1:54 PM
@Doorknob Can't a mod go through and hard-delete answers every so often? A monthly cleanup on one question doesn't sound so bad.
Or can only question be hard-deleted?

even if there are 100 (or 150 or 200) answers, you can't go through them all to find the active ones

@MartinBüttner Ah, right. Forgot about that.

The posted and deleted ones will bubble down to the later pages.
So will proposals that are abandoned

@Geobits Mods can't hard-delete stuff; that's an employee super-power as far as I know. But delete answers appear at the bottom by default anyway, so that wouldn't be a problem.

And if people keep their stuff in there for a while, but still intend to post it (like my Solar System KotH), they can always bump it with an edit.
@Doorknob they don't appear at the bottom for sorting by activity

1:56 PM
@MartinBüttner What if I want to sort by votes?

@MartinBüttner You know, that might actually work. That doesn't sound like a bad idea, when you put it that way.

Huh, I thought mods could do that. Yay learning!

The only problem I can see is that <10k users can't see deleted posts, but we don't have to delete them; nobody's going to edit a posted sandbox challenge anyway.

@MartinBüttner you can sort by activity, no?

@overactor then deleted answers are at the bottom. hence, delete your posted submissions.
@FireFly see my message just above that

1:57 PM
oh
Missed that

@Doorknob why do people even need to see the deleted posts?

@MartinBüttner oh

@MartinBüttner Uhh, I'm.... not entirely sure. I'm pretty sure I remember a reference somewhere; let me try to dig it up.

let me know when you find it. in the meantime I'll start writing up an answer (will wait with posting until you get back to me)

Until we get out of beta (if/when), it's only 4k to see deleted, not 10k. I can, for instance.

1:59 PM
@Geobits 2k, not 4k.

2

Keep them. I think that being able to view the history of answers in the sandbox and the corresponding comment threads provides a useful resource to people who want to learn how to write a good question. (How many such people exist is another matter).

lol, the downvote on that is mine

@ProgramFOX Yea, that. 4k does other deletion related stuff :)

@MartinBüttner It's a meta-post, so I won't get upset. ;)

phew ;)

2:07 PM
Someone's added another downvote. I assume the same person upvoted githubphagocyte's answer.
So looking at the vote totals, we seem to have the situation that there's little agreement on whether to delete the answers or not, but definite agreement that either way they should be edited down to the title and a link. I presume that the people who want it edited and then deleted have the rep to see deleted answers.

1

Proposal for the handling of retired sandboxes - Don't! I suggest to stop retiring sandboxes and keep using a single one. Sandboxes should be sorted by activity, because even at 50 or 100 answers, there are too big to find the active ones when sorting by votes. Assuming everyone does that, we ca...

how on earth did that get upvote before I managed to post the link here o.O

@MartinBüttner My phone vibrated :D

I think edited down and deleted is good simply because then they don't add much clutter no matter which sort you use, and the "deleted red" causes my eyes to skip them quicker.

@PeterTaylor I haven't voted on any of the answers, because I'm not really sure what the best option would be. And it really depends on what the result of the discussion on meta I just posted is.
> That seems like a very romantic idea
Lol, I love posted sandbox challenges so much <3

2:12 PM
@Doorknob btw, I'm tempted to close your question as too broad because it asks 2-3 question in one post :P

@MartinBüttner Would you prefer three meta posts at once? :P

The "sort by activity" is something that is easy to overlook/not bother with since one needs to remember to switch when switching between questions

@Doorknob Yes. It's clearer.

It'd be nice if sorting order was remembered per-question

@Doorknob Sure why not. They seem to be distinct questions. Although I think you can fit the "how do we format sandboxes" into the FAQ post instead of this one.

2:15 PM
@MartinBüttner Okay, sure, that makes sense. I'll remove the formatting question from the "proposed overhaul" post and add it to the FAQ (since nobody seems to nor would it make much sense to disagree with it, and if they do, they can always comment on the FAQ post). /cc @PeterTaylor
Actually, this might not even be necessary if we decide to retire the sandbox.

Okay something else... it seems that it's getting harder to post my planned GEB-based 18-hole golf course. But I was thinking, what might be (more?) interesting (anyway) would be a regex golf course (9 or 18 holes).
I could start with rather simple problems (actual regular languages and those cases where you have a hit and a miss list, as in the metagolf xkcd)... and then work my way up to context-free/sensitive languages, mathematical problems and weird stuff (something like vertical matching).
I'd consider giving bonus points or penalties to using different flavours for different problems (or single answers being as flavour agnostic as possible).
The main question would be whether to put it in a single post (like the old 9-hole courses) or whether to make it a series of posts, like once or twice a week, with a leaderboard in the most recent challenge.
Is there even anyone in this chat room who likes regex enough to tell me if that would be fun or not? :D

I dislike it enough to tell you... :P

...that it's not? :P

They're a good tool for some jobs, but I have no fun golfing them.

Well I'd try to make the latter problems less about golfing and more about just even solving them.
That is, problems you wouldn't normally use regular expressions for :D

2:28 PM
I like regexes enough that I'd attempt it, and probably enjoy doing so :p

@MartinBüttner So, you mean, fancy Perl-style regex magic? Also, closed as not having a primary winning criterion :P

Like checking for prime numbers (except there is a wellknown and probably unbeatable solution for that particular case)

Personally I'd skip, but some people seem to like playing with them. They seem at least semi-popular here.

@Doorknob How does regex golf not have a primary winning criterion?

I like how the composite-checking regex works

2:29 PM
1 min ago, by Martin Büttner
Well I'd try to make the latter problems less about golfing and more about just even solving them.

@Doorknob Well yes it'd still be a code golf.

You can have a problem that is hard to solve to begin with, and still make it a golf question
Anyway, @MartinBüttner, you'd need to specify well what makes a valid "regex" flavour..

I'm just saying the problem itself would be such that just finding an answer would be very interesting, even you're not that interested in the normal kind of regex golf.

Ah, okay. My brain already hurts way too much though when figuring out mildly complex regexes; I don't know how well I would fare golfing them :P

@FireFly I could whitelist them.
Or just point to the reference at regular-expressions.info

2:30 PM
Pointing to regular-expressions.info would probably invalidate vim-flavour regexps, since they're pretty...weird
But vim-flavour has some neat features that I haven't seen elsewhere

You're right, their list is by far not exhaustive.
But I'd definitely exclude anything which pulls in things from the host language/tool.
Like using eval in Perl.

That makes sense

@Doorknob Why is our regex tag named "regular-expression"? :/
that's not very golfy :D

That's a pretty good list

@MartinBüttner I thought there were separate regular-expression and regex tags with slightly different meanings?

2:34 PM
@PeterTaylor there used to be regex-golf
or maybe there still is
no there's not

Oh. Only one tag matches the substring reg

that was deleted in favour of using both [code-golf] and [regular-expression]
I'd just prefer if it was called [regex] as on SO ^^

That's a more accurate name.
That reminds me: I've been neglecting the puzzle-retagging task.

@Doorknob actual renaming of tags isn't within your powers, is it?
btw, for people interested in dark regex magic, check out this series on SO if you haven't seen it: meta.stackexchange.com/questions/62695/…
in particular, it taught me that Java has a bug that allows you to get arbitrary-length lookbehinds
Btw, alternatively I could include flavour-specific challenges in the golf course. E.g. a challenge that requires you to mimick lookbehinds in the ECMAScript flavour

@MartinBüttner ... because what else would it be named?

2:47 PM
Which, as Peter points out would even be more accurate since modern regexes aren't technically regular expressions.
So we could have a regex tag for real-world (non-regular) regexes, and regular-expression for regular language related problems.

If there are enough regular language related problems to need a tag.

@PeterTaylor Yeah, good point. It would still be nice to make it unambiguously clear that regex challenges are about non-regular regexes.

@MartinBüttner Wouldn't that be [regular-language]? Also, [regex-golf] is already a synonym of [regular-expression] IIRC

That tag would probably be better off called 'regular-language' in that case

@Doorknob no it's not, we ditched that
yeah [regular-language] sounds good as well
anyway, doesn't really matter...
@Doorknob I still can't solve your buildings-throwing-shadows challenge because of a bug in the .NET framework :D
maybe I should try if it works under Mono :D

3:01 PM
@MartinBüttner Lol, which one is it? I can't imagine a bug that would break something like this
I'm poking more people with sharp sticks to figure out if there would be something like a hard limit on answer count, if we do go with the "no retiring sandboxes" thing.
In any case, it might be practical to stop at 500 or so :D

549? It just shows 518 for me. Is 549 including deleted answers?

Yes, the counter includes all the answers that you can see. (So if you had your own deleted answer, it would show up as 519 for you but 518 for other non-10kers. Pretty fancy. :P)

@Doorknob If we can deal with 500 answers in the sandbox I see no reason why we can't deal with 1000 answers in it.

@MartinBüttner Or 5000, or 10000, or 10000000? There's got to be a stopping point somewhere.

Also if it becomes problem data-wise (which I doubt), we could still ask for some help from SE like twice a year to perma-delete all the deleted proposals.

3:08 PM
On a related note, it might be a good idea to merge all the current sandboxes so we don't have all these old artifacts lying around.
@MartinBüttner Meh, one sandbox per year or something wouldn't be too bad.

At any point where you retire the sandbox, you'll punish some fresh answers in the old one.
How do you merge sandboxes if you can't even migrate answers?

We can merge things completely, but we can't move specific answers.

oh okay

@MartinBüttner There could be some kind of preliminary announcement or something. (Or the rolling window thing that we have now, with 2 sandboxes.)

@Doorknob Let me see if I can find the bug report

3:13 PM
@MartinBüttner Wow that thing is ugly.
A dev that I poked with a stick is getting some information about what a good guideline for "way too many answers" would be.
@MartinBüttner What are you talking about that's exactly what I said

Anyway...
Hopefully we won't even get there until the SE-wide solution mentioned by Grace Note will be implemented

3:34 PM
Stick poking success - response is:

(@MartinBüttner)

or you do

done
did you guys see this?
0

Background You are all traders for a slightly less than reputable stock firm. You all are part of a group of traders who focus only on one specific stock. Each hour each trader has a chance to Buy X stocks or Sell X stocks. There are 50 hours per round and 3 rounds per competition. At the end o...

(especially @PhiNotPi might be interested)
@Doorknob btw, while we're redoing the sandbox... this might be because I'm not a native speaker, but I think "Proposed Question Sandbox" can be read as "Proposed 'Question Sandbox'" instead of "'Proposed Question' Sandbox". That's why I never cared to look at it when I was new here. It sounded like some decision-making going on in the community about whether a sandbox should be made. As a new user that obviously didn't bother me.
That might be a reason why so few new people check it out.

3:49 PM
Hmm, possibly. "Sandbox for Proposed Challenges" might be a better title.
Also, while we're on the sandbox topic, we might want to submit a few feature-requests to SE:
- Ability to disable 5-rep requirement for specific meta posts
7
- Ability to change answer sort for one specific post

Might want to clarify the second one: is that "Each user's answer sort is per-post" or "Mods can set an answer sort for one specific post which overrides the user's one"?

@PeterTaylor I'd prefer the former version as "Each user can lock the sorting mode on individual posts."
that is, my sorting mode would still be remembered between posts, except when I visit one where I ticked a "remember sorting" checkbox.

@PeterTaylor Either would work, I suppose