0

I would like to experiment with trapdoor functions as trottle tools for stuff like browser side API, blog comments, etc.

The idea would be to generate an x on the server, and send f(x) on the client. If the client want to send content anonymously, he must find x and send it back.

I don't know if it's a good idea or not, but I'm trying concepts to expose services that currently allow registration content production only, and would like to expose it anonymous without being hamered or maintain a record of past activity.

Obviously I don't want the client to make huge calculations, so I'l like to be able to tune it so finding back x can take only a few ms/tenth a sec/sec depending of the need. I do understand it depends very much of the client hardwayre and current load. This is an experiment, I'm not planning to put it in production any time soon.

I really wish for an explained algo that I can port to JS, Python, etc. But if you just can point me to good learning materials or existing implementation, I'll take it.

Bite code
  • 578,959
  • 113
  • 301
  • 329
  • Why don't you just limit requests at the server end? (The suggested mechanism won't prevent a malicious client from bombarding you with requests.) – Oliver Charlesworth Jun 20 '14 at 06:52
  • I already have nginx throttling setup. I'm just curious and want to try this out. – Bite code Jun 20 '14 at 06:57
  • On of the thing I'm wondering is : if I got a 100 000 blogs with this mecanisme, a bot trying to spam all of them will take a lot of time, and maybe it will make their life harder. I don't know if it's valid. Just an idea to try. – Bite code Jun 20 '14 at 07:06

0 Answers0