3
$\begingroup$

Sometime we have questions about a home-grown scheme, typically a weak cipher, sometime ill-posed; e.g. recent example 1 and example 2.

This is related to Should we allow questions about amateur ciphers?, but I'd like to focus on what should be concretely done when one such question surfaces. Should it be:

  1. Closed only when too unclear (my initial reaction to example 1)?
  2. Closed on the grounds of the FAQ entry Do we accept questions asking for cryptanalysis of your cipher-hash function design ?
  3. Closed unless clearly worded and showing some potential?
  4. Same, but commented with a short refutation (rather than a full answer), and closed?
  5. Answered by a simple argument, perhaps after re-phrasing (my second move, after example 1 improved to the point I understood it)?

I'm not too keen of (2.), being myself a culprit of something similar. But (5.) and even (4.) come with the risk of attracting more similar questions, and making the site a mix of academic-level and overly trivial stuff.

Edit: some excellent contributors did (5.) for example 2, and managed to inject some wisdom in answers to a rather uninspired proposal. We seem to seldom apply (2.)

Also, perhaps there should be a tag for this kind of question? I have used the existing "classical-cipher" but it is a poor fit. "Home-grown"? "Trivial"?

$\endgroup$
3
  • $\begingroup$ I wouldn't tag your question with "Trivial" :) $\endgroup$
    – mikeazo
    Oct 22 '12 at 12:17
  • 1
    $\begingroup$ What has changed since Should we allow questions about amateur ciphers? If nothing has changed, seems like a natural answer is to lean towards 2., though this is an individual decision to be made based upon the specific question, and some questions might well be handled better in a different way. In other words, "do something reasonable". $\endgroup$
    – D.W.
    Oct 26 '12 at 8:11
  • 1
    $\begingroup$ @D.W.: Indeed, for something with no particular interest, closed with a comment citing the FAQ entry and perhaps refutation hint (rather than answer with full refutation, as I did recently) seems the Reasonable Thing. $\endgroup$
    – fgrieu Mod
    Oct 26 '12 at 15:43

You must log in to answer this question.

Browse other questions tagged .