/meta/ - Meta Questions

Every Mathchan board is going have custom-made captcha questions, so boards like /dev/ - Software Development and /os/ - Operating Systems will require no math but will require knowing programming.

>>590
neat, however these custom captchas are still liable to just being knowledge checks. If you get a guy who just does OOP to answer questions about functional programming, he's going to be filtered despite being compentent. Context and instruction on how to solve the problems is necessary to use them in any way as an assessment of capability. Being thrown obscure set-theory notation symbols is far from ideal.

Also, for the more general boards, maybe it'd be a good idea to have general questions. Maybe something like a ravien matricies quiz to ensure that someone is big brained enough to post would be good.

Sooner or later you're going to have to implement a proper way to generate the questions. I'd suggest liberating the source code to this captcha mechanism you have so users can contibute to it. If this isn't done, then brute force methods to "solving" problems will just be used.

>>591
Ensuring a certain minimum knowledge to post on a board is the best way to go about it in my opinion. Raven's Matrices would just be a waste of time. At least I don't see how it would have any substantial effect. Really, just having a good moderation team and only a few people reporting low-quality shitposts would alone already suffice. I doubt this site will ever become big enough where testing "general intelligence" would be of any use

>neat, however these custom captchas are still liable to just being knowledge checks
It's been agreed that the captcha questions for programming boards should resemble Leetcode/Project Euler/Advent of Code i.e. they will require programming, but be independent of any particular language. We'll have online compilers for 15 different languages, just like many other websites do, so getting a compiler at hand shouldn't be an issue. Of course, the problems are going to be much easier to do (they shouldn't take more than 10min to code for an experienced programmer) and won't be required for every single post to do (because that would be ridiculous).

I agree with the poster above that knowledge checks aren't bad, although I also agree with you that filtering functional programmers just because they don't know OOP (or vice versa) is a bad idea.

>Maybe something like a ravien matricies quiz to ensure that someone is big brained enough to post would be good.
Don't think an IQ test is a good captcha for any board except /iq/ and /ret/.

>I'd suggest liberating the source code to this captcha mechanism you have so users can contibute to it
I'll liberate some parts of the source code but not right now.

>>593
Is the plan with /ret/ that you'll have to score low enough to be able to post? Otherwise I don't see a point of a nonstandard captcha there.

>>592
>Ensuring a certain minimum knowledge to post on a board is the best way to go about it in my opinion
maybe on the topic-orientated boards that's a decent idea, but the more general boards should only require general knowledge. Best to not exclude the "jack of all trades" types.

>>593
Raivent matricies are the only competency check I know of that is agnostic to context. I only really condone using them on the generalist boards where knowledge checks aren't ideal.

i dont know how to do math and i bypassed the captcha using ai hahahaha

Why can't you answered on the /jp/-board without capitar?

I want more Anime- and Manga-Freaks here... Anime is soo beautyfull.

>>639
stop being an indian.

>>640
I'm European.

>>589
>programmers (such as myself) may get filtered
Fellow programmer with relatively weak maths skills here. (Although I'm working on it, because I want to learn about electromagnetism.)

Maybe it's changed since you posted this, but currently I think it's good as-is. My degree was remarkably light on maths but I still covered enough logic/graph theory stuff to answer most of the captchas I've seen. To be honest, programmers who can't answer any of the current question bank probably haven't studied enough theory to make good contributions or ask good questions.

Also it really helps that the questions are multiple-choice, because it removes the need to fully/properly solve the problem. You just need to get close enough to either rule out the wrong answers or do a brute force check.
e.g. for a question about subgraphs, I worked out the answer for a small value of n by hand/intuition and then plotted the functions given. Only one matched up. Easy peasy.
For another question about cartesian products, I just wrote a tiny bit of javascript.

tldr, it's fine, keep it.

It's alright!