3
$\begingroup$

In the past, I (and presumably others) used to tag questions about ciphertext indistinguishability under various attack scenarios with the tags for the relevant attacks (e.g. for IND-CPA, for IND-CCA or IND-CCA2, etc.).

However, recently, the more specific tags and have been created. At the moment, these are only used on a handful of questions, and are mostly redundant to the older tags, but presumably their existence will lead to people using them. I do feel that these are sort of "natural" tags, but the redundancy with the older, more general tags concerns me a bit, so I'd like to ask for the community's opinion on what to do with these tags.

Specifically, here are a few options I can think of:

  1. Do nothing, just let these tags live and grow as people use them. This is, of course, the default option. Alas, this choice would likely mean that most future questions on, say, IND-CPA security would be redundantly tagged with both and , except for a random few where the asker either missed one of the tags or bumped against the five tag limit.

  2. Merge the new tags into the older ones, making a synonym for and a synonym for . (I already suggested the former synonym, but I don't have enough answer score to do the same for the latter.) This would eliminate the redundancy, but only at the cost of some expressive power.

  3. Keep both sets of tags and try to make them less redundant, e.g. by using only the "ind-*" tags for questions about formal ciphertext indistinguishability, and reserving the "*-attack" tags for other questions about such attacks (e.g. "how to decrypt this message using a chosen plain/ciphertext attack?"). This would mean going through a bunch of old questions (12 + 7, at the moment) and retagging them where necessary. If we go this route, we should also add clear tag wiki summaries for all those tags describing how and when they're to be used, and keep an eye open for mistagged questions in the future.

Of course, there might also be other options I couldn't think of.

Personally, I'd prefer either option 2 or 3 over the default of just leaving the tags like they are. I have no strong preference between them, though, so I'd like to ask what others think:

Should we have tags for and separately from and ?

$\endgroup$
2
$\begingroup$

I don't have a strong opinion, but here's my sense:

Use where the question asks about IND-CPA, or about semantic security, or other notions of provable security that are equivalent to IND-CPA.

Use where the question asks about a particular chosen-plaintext attack, e.g., differential cryptanalysis or something like that.

Examples:

Do something similar for vs .

I do think we should make a synonym for .

$\endgroup$
  • $\begingroup$ Sounds pretty reasonable to me. $\endgroup$ – Ilmari Karonen Nov 3 '12 at 21:36

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .