Talk:Q (cipher)

Notability

Rather than just looking at this cipher alone we should discuss whether generally ciphers submitted to important standardisation processes should be kept or not. E.g., if we delete this page, should we also delete the pages for ciphers such as Nimbus (cipher) or MAGENTA which were broken almost immediately during the first AES conference. I.e., while these ciphers are not very interesting alone, having pointers to the papers that cryptanalyze them are in my opinion a reason to keep the pages. And if we keep some of these pages it makes sense to just keep all of them for completeness. 62.203.19.16 (talk) 18:26, 16 January 2009 (UTC)

Don't the References show "significant coverage in reliable sources"? Ntsimp (talk) 19:08, 16 January 2009 (UTC)

My thoughts: I went through a similar discussion recently with regard to Wikipedia:Articles for deletion/Maraca (hash function), and I tend to stand by the basic principle there - if the standardization process accepts any submissions, without minimal requirements for accepting those submissions, then such a submission is not enough to establish notability. However, passing the first cut in such a process probably is sufficient to claim notability. I also don't think that publicly available cryptanalysis (that's what our references are here) is sufficient to establish notability - essentially, such publications just indicate that the algorithm has no future, which isn't an indication of importance. My brief web search found that essentially all online mentions are either a pointer to the cryptanalytic paper, or a link to this article. Of course, it's possible for a broken algorithm to be notable; FEAL is notable precisely as a sort of canonically broken cipher. I don't see anything comparable here, however. — Gavia immer (talk) 20:17, 16 January 2009 (UTC)

Ntsimp, thanks for drawing my attention to the prod removal!

Gavia immer, I agree entirely. Note that neither of the cryptanalysis papers themselves have significant cites; the attacks are fairly straightforward stuff (though clever) and no new techniques are presented that can be applied elsewhere. If you submit a weak cipher to a public contest and it's broken, it's perverse to say that this is enough to make the cipher notable. ciphergoth (talk) 20:37, 16 January 2009 (UTC)

That being the case, the IP above is right that Nimbus (cipher) is in the same boat. I would also draw your attention to Grand Cru (cipher) and NUSH. I created all 4 articles, but if consensus is to merge them with NESSIE, then I won't take it personally. Ntsimp (talk) 22:17, 16 January 2009 (UTC)

Yes, I think they are in the same boat. It's a shame to lose the article on MAGENTA - unlike all the other AES candidates, the inventors proudly kept it under wraps until it was unveiled at the first AES conference. As Schneier tells it, when the presentation finished, a hand went up, and John Kelsey said "Could you go back to slide 5 please?", and proceeded to break the cipher then and there. Now that's quality - especially when it was also the slowest entrant. Hopefully Deutsche Telekom have got better at crypto since then :) But I don't have a cite for that story, so it can't go in, and really it's only a footnote in crypto history.

If you agree, what's the best way forward? Should we restore the prod notice, or go thorugh AfD? ciphergoth (talk) 12:02, 17 January 2009 (UTC)

According to WP:PROD#Conflicts, you can't put back the prod notice. I think the right thing per WP:MERGE is a merge proposal at Talk:NESSIE, probably with notification at WT:CRYPT. Ntsimp (talk) 00:50, 18 January 2009 (UTC)

Done, please let me know if I've got it right - thanks! ciphergoth (talk) 12:57, 18 January 2009 (UTC)