@PeterTaylor: I see your point with respect to the three posts. The common thread is that they're trivial.
Even so, the three collectively generated well over 30 answers. The operative question is whether they clutter up and therefore detract from the overall quality of the site. Ultimately that depends on how high one wants to set the minimum bar for participation. Golfing a program to output multiples of 3 is still a challenge for novice programmers, for example.
If trivial questions aren't what the site wants or needs, adding a "too trivial" reason for close and then closing such questions on that basis seems prudent. I'm always and advocate for formal, explicit rules rather than soft "community pressure" -type rules governing minimum content standards.
The other way we might deal with this is by adding "beginner" or "5-minute challenge" tags to the most trivial questions. Something to acknowledge the fact that the task is trivial and can be tackled by most any programmer in a matter of minutes. This would also give any newcomers to the site a cle…
Ultimately I see the problem of trivial questions as self-limiting, as there are only so many reasonably distinct simple programming challenges. And like it or not, they seem to be more popular than 90% of the questions asked, almost certainly because they don't involve a significant commitment of time to solve.
@MartinBüttner: It's not a perfect solution, but it still strikes me as more elegant than a site littered with highly downvoted, untagged questions with no immediately obvious reason for the downvotes.
I have no problem with the "trivial questions need not apply" approach either, but my personal practice is to only burn content using hard tools (e.g. vote to close), and to invoke those only if content violates some plainly codified set of rules. I'm not at all a fan of half-measures like downvoting.
@COTO I don't know if you're aware, but there are some relevant posts on meta about the difficulty level of questions (and tags concerning them). There hasn't ever seemed much of a consensus on it, though, so people will vote as they see fit: http://meta.codegolf.stackexchange.com/q/255/14215 http://meta.codegolf.stackexchange.com/q/1249/14215 http://meta.codegolf.stackexchange.com/q/64/14215
In my opinion, there's nothing wrong with downvoting. It's the primary tool to grade content on the site, and nerfing that would be bad.
Of course, as someone who downvotes a lot, I may be biased. But I do it because I believe it's good for the site as a whole.
It's really interesting to read old meta post... save for a couple of users it seems like there was a completely disjoint set of active users from today. it's like PPCG passed entirely into the hands of new generation of users ^^
The recently added list for golfing tips in ECMAScript 6 sparked a bit of a discussion in the comments (and the close/reopen votes) whether tips for different versions of one language should go in the same list or whether we should create a new list. This is mainly relevant if a new version of a ...
It's hard to tell based on the question/comments if I'm voting for splitting ECMA, or voting for splitting all languages by version. I voted against splitting "all", but I don't know enough about ECMA version differences to give an opinion on that one. For instance, Java 8 brought a few new things to the table (some of which change golfing techniques), but I wouldn't want to see it split.
but I guess you've always got the problem: no matter how many new things a new version brings to the table, a lot of old tricks will still be applicable... it's just there a handful will be superseded and there'll be a few entirely new ones. regardless of how much new stuff there is, splitting the versions either creates duplication or you have to check both posts.
@PeterTaylor Re your comment on meta, I'd really like to do a systematic cleanup of some of the list questions, but despite them being CW, I still feel people would object if we merged/split up/deleted some of their answers.
Task
Unfortunately, the Professor is not able to use his keyboard properly: whenever he is meant to use the Shift key, he presses Caps Lock once before typing, and doesn't bother to correct himself. If there are two or more keys in a row that require Shift, he only presses Caps Lock before the f...
Print all Country Codes
code-golf kolmogorov-complexity
ISO 3166-1 is the standard that defines all the country codes. The well-known two-letter codes (US, GB, JP, etc.) are called Alpha-2 codes.
With two letters, there are only 262 = 676 possible codes, which can be nicely arranged in a grid....
Also BTW, the Wikipedia page makes me laugh. "If everyone reading this right now gave €2, our fundraiser would be done within an hour. Yep, that’s about the price of buying a programmer a coffee."
Must be written by someone who buys very expensive coffee.
In the spirit of this xkcd
Write a program that plays regex golf with arbitrary pairs of lists. The program should at least attempt to make the regex short, a program that just outputs /^(item1|item2|item3|item4)$/ or similar is not allowed.
Scoring is based on the ability to generate the sho...
hm, that got me thinking... regex golf might be suitable for cops and robbers... write the shortest regex, where it's not obvious what it does (or doesn't match).
The more brutey(?) ones so far have been based on "irreversible" math, etc. With a "short" regex it'll probably be harder to obfuscate. Even if I can only figure out what half of it does, that narrows the search considerably.
Note: I don't regex more than I have to, so my opinion of this may be flawed.
@PeterTaylor Huh? a) You don't have a choice with some flavours, b) There are some features in modern regex where you can't avoid catastrophic backtracking.
@Rainbolt I never said I disagreed. I even upvoted your answer. And I explicitly stated that my comment was just an observation and not meant to be judging at all.
It seems to me that you could actually fit tips into a regular stack exchange format.
But at the same time, they could remain a "mini-challenge"
Everyone is happy now, because we have challenges and we fit the format.
@MartinBüttner Fixed so that it's clear what you said and what I meant to say. Hopefully this can actually progress the discussion instead of starting a popularity contest in the comments.
@Rainbolt yes, thank you. I don't really have anything to say to that at the moment, but I upvoted the comment. you might want to read COTO's chat messages from earlier today
@PeterTaylor I think I might give the country code table a go tomorrow anyway... just because a <100 byte approach exists, doesn't mean it's a boring golf, I'd say. and we haven't had that many simple-but-good challenges lately.
I agree with COTO that trivial is not necessarily bad, but completely disagree that having a large number of answers has much relevance on the quality of the question, or its "fit" on the site. [code-trolling] being the obvious example.
I'm sure there's a way to write a good, trivial question/challenge. The problem I see is that normally when I see something I consider trivial, it's also badly formatted, lacking a winning criteria, a whole list of trivials jammed into one question, or has some other problem that makes me want to DV or CV it.
@COTO Also, as one of the downvoters on Yo Quiero jQuery, I'm just not a fan of mashing six challenges into one. It's not breaking any rules, so I didn't close vote. However, I don't like it much (and would rather see things separated more), so I downvoted. You can call downvoting a "half-measure" if you want, but I believe that DVs and CVs are intended for two different things.
regarding the regex challenge... I'd like to limit the flavours to testable things... is it alright if I limit it to flavours that have an online interpreter available? or should things that are freely downloadable be accepted, too?
Cops and Robbers: Regex Golf
cops-and-robbers code-golf regular-expression obfuscation
The Cops' Challenge
You are to write a short, obfuscated regex, satisfying the following spec:
You may choose any flavour that is freely testable online. There's a good list of online testers over on Stack...
I asked "What is a key?" and then saw your definition of key in italics. IT WAS EVEN EMPHASIZED FOR ME!
@MartinBüttner I do have a follow up question: If I choose my key to be foo/bar, but my regex also happens to accept baz and reject taz, is baz/taz considered a correct guess?
More generally, can you crack my regex with any successful key, or must it be the key I chose?
Ok good. I kind of figured it must be that way, but I couldn't help but wonder: why would anyone submit a failed guess? Surely there are easy to use regex matchers that will tell you if you guessed correctly.
I think I should remove the time constraint on matching if people can prove that a certain string is matched. Otherwise, catastrophic backtracking won't help conceal your key, because you can just abort after the time limit has run out (if you want to automate the search).
Another problem is that it will probably always be trivial to check if the empty string matches, so that you only need to find either a matching string or a failing string.
@Geobits you could still test more than by hand ;)
Maybe I should require the strings of the key to be at least X characters long?
btw, for all regexpal-users out there. decent alternatives:
Write a regex that matches a string s and fails a string t. s and t must be between 10 and 20 characters in length. The keys to your regex are any such s and t.
@Rainbolt Yeah, I don't think that's a problem. There's a time limit on the matching, and I'd prefer people to actually work out what the regex does, in which case the length of the key doesn't really matter.
Ugh. I tried to tell the Internet guy that I wasn't interested in upgrading from my "$40/month, 20 Mbps" speed to "$50/month, 20 Mbps" speed. He couldn't seem to comprehend that.
Oh.... Suddenlink. I had them in Arkansas... terrible people. Not that my current Comcast reps are any better, but...
@Beta re bounty: I think it's odd that you chose those languages on a [fastest-code] challenge and don't seem to care about their speed (only which is submitted first) but it's your rep; Bounty away :)
Visual Studio has this awesome copy/paste/select feature that just saved me some time that I immediately lost by sharing it with you guys: youtube.com/watch?v=dlAacV7H6sI#t=34
I think Powershell or cmd has block select, so you would think that all new text editors would have it too
@BetaDecay You should also consider "bounty not satisfied" when giving alternate criteria like that. If nobody submits an answer in one of those languages, what will you do? If you don't award it to anybody, the system auto-awards half (25) to the most upvoted answer that comes in since the bounty was posted.
If you select two blocks of text in Word 2010 using ctrl+click and drag, and then type something, it only replaces the last such selected block. That's such weird behavior.
You can do more with it. I have sql columns like Name Age Race Gender and suddenly I join another table that has all of those columns. Now I have to qualify them. Now I don't have to manually type table1 dot in front of all of them. I can block select the space in front and type and it appears on each line.
I can absolutely see where it would come in handy for things like that. For multi-file projects it would b easier to do an IDE-style rename, since it will catch them all from context.
@Rainbolt I use that if I'm golfing in Eclipse. Makes it simple to write it with "regular" naming and shrink.
and possibly because of its rosetta stone character
just because Optimizer has basically won with CJam, people aren't discouraged because the question kinda admits that solutions in different languages will be unique
Hmm. I was going to write up a spec to get a list of valid credit card numbers (using Luhn) given a partial card number (always including the checksum digit), but it would probably be at least close to a dupe of this one. Time to find something else, I guess :(
Is there any way cleverer or shorter than brute force to generate a given CRC-32 (with an input string at least x bytes long)? Trying to think of alternate checksum-related challenges...
@MartinBüttner: The sublimetext editor demo is neat, but my IDE has had all of those features for at least five years now. It also highlights all replacements and displays the replacement text a priori when using regex search/replace. I'm guessing most modern IDEs would have similar features. @Geobits: I don't mind the downvote if you genuinely didn't like the spec, but I prefer to know _why_ my content is being downvoted. You've clarified your vote now hence thank you.
@PeterTaylor: Regarding simple vs. trivial questions, I make the same distinction. Even so, I consider the number of answ…
@COTO The only truly relevant criteria for downvoting is your opinion. Voting lets people know what the users consider good and bad (and in between). If you like it, upvote. If you don't, downvote. Otherwise your opinion goes unheard. Hard rules are better suited for close/deletion voting.
@COTO Yes, number of answers is the best indicator for how popular something is, but what Peter and Geobits were saying was that popularity doesn't indicate whether something should be on-topic or not.
My previous comment was geared more to questions than answers IMO (especially here). For answers, I generally downvote only if it's broken or doesn't meet the spec for some reason. Of course, you can do whatever you want with yours ^^
The top four answers to that question I just linked to: 1. "How you spend your votes is entirely up to you." 2. "What you do with your votes is entirely up to you." 3. "That is entirely up to each reader to decide." 4. "As others have said the how you vote is up to you."
As for the CRC-32 problem, I don't know. But generating a particular CRC-32 hash sounds remarkably similar to the proof of work technique used by Bitcoin, which they use precisely because it does have to be brute forced.
@COTO As far as downvote comments go, there's been discussion about that on meta since the inception of the site (SO, not PPCG). Opinion tends to "don't make me comment".
1) Voting is supposed to be anonymous. Commenting takes that away (from those that want that). 2) Sometimes when I'm on mobile I don't feel like typing it out 3) Sometimes I'm just busy. A vote is a single click. A comment requires more concrete thought and time.
The best advice I can give to anyone when they ask why something was downvoted is "don't sweat it". Downvotes happen. You're just not going to be able to please everybody every time. It's only a 2-rep hit, even if you're only here for the rep game. I get that it's nice to know why, but does a vote or two really matter?
@Geobits: I've scanned the "when do I downvote?" links you sent, and the consensus does seem to be that voting is totally subjective. Like it for whatever reason, upvote it. Don't like it for whatever reason, downvote it. Personally I'd downvote the complete lack of structure or guidelines in that procedure, but I can live with it. Downvotes barely have any effect on rep anyway, so I shan't complain any longer. ;)
Yea, that. Don't get me wrong, when I first joined SE, downvotes mattered more to me, because I saw them as a form of 'personal attack'. Letting go of that was a step on the road to stress-free living ;)
It all depends on the commentor too. "You missed this." could be rewritten as "I think this should be added." The former is attacking the person, and the other is not really attacking anything.
ISO 3166-1 is the standard that defines all the country codes. The well-known two-letter codes (US, GB, JP, etc.) are called Alpha-2 codes.
With two letters, there are only 262 = 676 possible codes, which can be nicely arranged in a grid. This table can be useful as an overview, to see which cod...
@Geobits I'd have to check the details for whichever CRC you mean (IIRC there are about 7 "standard" CRC-32s), but generally they're just polynomial remainders, so you can generate all matches with polynomial multiplication and addition.
@COTO If you stick to one comment per comment it's easier to read because we don't have to click the "See full text" link
In the spirit of this xkcd
Write a program that plays regex golf with arbitrary pairs of lists. The program should at least attempt to make the regex short, a program that just outputs /^(item1|item2|item3|item4)$/ or similar is not allowed.
Scoring is based on the ability to generate the sho...
