Kermit Alexander II

If I understand you correctly, you're trying to modify the way the imageboard posts themselves are parsed. [CONTRIBUTING.md](https://github.com/DangerOnTheRanger/maniwani/blob/master/CONTRIBUTING.md#modifying-the-markdown-parser) has a high-level overview of what you'll need to modify and/or add....

reCAPTCHA and captchouli (at least for now) have no widespread automated solvers out there, and writing/maintaining a solid, readable CAPTCHA from scratch just honestly wasn't something I was/am personally interested...

Fair enough - for the record, captchouli lets you configure [what characters to use](https://github.com/bakape/captchouli/blob/master/cmd/captchouli/main.go#L22), in case that helps; docs on how to pass flags to Dockerized captchouli are available [here](https://github.com/DangerOnTheRanger/maniwani/tree/master/build-helpers/captchouli).

@samip5 I'm still interested in merging this if you're still interested in working on it - if you have any questions about Maniwani's codebase/architecture I'll do my best to help...

@samip5 Using C++ would be fine with me, though I am slightly concerned about how many more steps it would add to the build process (installing a C/C++ compiler, installing...

I'm in it now, same username as here. I'm also on #maniwani on rizon.net at most US times of the day if IRC is more convenient.

No worries, we're not operating under deadlines - it's a hobby project after all ;)

Post hashing is simple enough that there have already been attacks on Futatsu that have circumvented that sort of strategy, so I don't recommend it. Something like integration with [SpamAssassin](https://spamassassin.apache.org/)...

The issue is knowing what is and isn't spam from just text analysis requires a pretty large dataset if you're not just doing things like checking how many links are...

Anything not using NLP is too easy to circumvent (hash-based methods get beat by simply changing single characters, URL counts don't catch people simply trying to flood a board and...