Created attachment 67558 [details] patch for 276 Version: Git (using KDE 4.7.4) OS: All 276 is used by Rizons modded hybrid IRCd and Freenodes ircd-seven (and probably others) to indicate the user is using a client certificate [0]. 378 is used by Freenode to indicate where a user is connecting from [2]. [0] http://code.rizon.net/ircd/src/1faa4a4b102a/include/numeric.h#cl-167 [1] https://dev.freenode.net/redmine/projects/ircd-seven/repository/revisions/master/entry/include/numeric.h#L123 [2] https://dev.freenode.net/redmine/projects/ircd-seven/repository/revisions/master/entry/include/numeric.h#L204 Reproducible: Didn't try Steps to Reproduce: Run a whois on either Rizon or Freenode on a user Actual Results: 276: [276] Mineo Mineo is using a client certificate 378: [378] Mineo Mineo is connecting from $host Expected Results: 276: [Whois] Mineo is using a secure connection. 378: [Whois] Mineo is connecting from: is connecting from $host Note: The result for 378 looks very weird, but 338 behaves the same way.
Created attachment 67559 [details] patch for 378
http://www.alien.net.au/irc/irc2numerics.html lists dupes for both. The dupe for 276 appears to be related to this: http://www.wohmart.com/ircd/pub/hybrid/2-Patchsets/blitzed/ircd-hybrid-blitzed/doc/vchans.txt IIRC basically nothing implements VCHANs in the real world, so we can probably ignore this. 378 is more complicated though. While the dupe again seems irrelevant, I remember investigating 378 before, and IIRC other ircds seem to implement it with varying formats. Some post the host info twice, once before the colon and repeated in the message after it. Some include the IP, others don't. This makes it pretty hard to parse in a generic way. Bottom line, as far as I can see the 276 patch can go in, but 378 is troublesome. Eli, a second opinion please?
Ping? It seems like this has been stuck on the request for a second opinion for more than a year now :/