summaryrefslogtreecommitdiff
path: root/kike.c
Commit message (Collapse)AuthorAgeFilesLines
* kike: fix up debug messagesPřemysl Eric Janouch2020-10-121-2/+2
|
* kike: use read/write rather than recv/sendPřemysl Eric Janouch2020-10-121-3/+3
| | | | | | read/write support non-sockets, otherwise they're the same here. This is in preparation for fuzzing.
* kike: silence an annoying build warningPřemysl Eric Janouch2020-09-201-13/+22
|
* Name changePřemysl Eric Janouch2020-09-021-1/+1
|
* Update NEWS, bump versionv0.9.7Přemysl Janouch2018-10-211-1/+1
|
* kike: fix wildcard handling in WHOISPřemysl Janouch2018-10-211-6/+6
|
* kike: allow STATS with no parametersPřemysl Janouch2018-10-211-4/+4
| | | | We were in plain conflict with RFC 2812 for no apparent reason.
* kike: explicit conversion from pointer to booleanPřemysl Janouch2018-08-011-1/+1
| | | | | In practice the values in the map may only be 1 or 0, so it doesn't matter, but in C it is better to be safe than sorry.
* kike: break out properly on errors in MODE processingPřemysl Janouch2018-08-011-1/+2
| | | | We used to only abort the inner loop, which was insufficient.
* kike: reset user modes while processing USERPřemysl Janouch2018-08-011-0/+1
| | | | | Since the processing always succeeds and registration cannot be undone, this doesn't seem to fix any real issue.
* kike: do nothing on equivalent renicksPřemysl Janouch2018-08-011-0/+4
|
* kike: code cleanupsPřemysl Janouch2018-08-011-4/+1
|
* kike: fix grammar in hostname validationPřemysl Janouch2018-08-011-1/+2
| | | | | This has an entry in RFC 2812 errata, although it's held for document update. We can afford the strictness.
* kike: fix grammar in config item descriptionPřemysl Janouch2018-08-011-1/+1
|
* kike: add a comment about identifier encodingPřemysl Janouch2018-08-011-0/+7
|
* Relicense to 0BSD, update mail addressPřemysl Janouch2018-06-211-3/+2
| | | | | | | | | I've come to the conclusion that copyright mostly just stands in the way of software development. In my jurisdiction I cannot give up my own copyright and 0BSD seems to be the closest thing to public domain. The updated mail address, also used in my author/committer lines, is shorter and looks nicer. People rarely interact anyway.
* kike: disable TLS session reusePřemysl Janouch2018-01-091-0/+3
|
* kike: handle accept() errors betterPřemysl Janouch2018-01-091-9/+14
| | | | Might prevent some denial of service attacks.
* kike: thorough review, no functional changesPřemysl Janouch2018-01-091-3/+15
|
* kike: introduce cstr_set()Přemysl Janouch2018-01-081-23/+12
|
* Bump liberty, avoid fall-through warningsPřemysl Janouch2018-01-071-0/+1
|
* Rework constructors/destructorsPřemysl Janouch2017-06-221-12/+8
|
* Bump libertyPřemysl Janouch2017-06-221-138/+81
|
* kike: fix two memory leaksPřemysl Janouch2017-06-221-0/+2
|
* Avoid the "poller_fd::closed" featurePřemysl Janouch2017-05-061-4/+3
| | | | | | | Reliability enhancement for Linux. This feature was created for ponymap, however we don't care about an extra syscall in most places. Doing it right even saves lines.
* Bump libertyPřemysl Janouch2017-01-231-85/+85
|
* CleanupPřemysl Janouch2016-12-281-10/+1
|
* Add support for OpenSSL 1.1.0Přemysl Janouch2016-12-281-0/+5
|
* Bump libertyPřemysl Janouch2016-10-111-12/+12
|
* Better support for the KILL commandPřemysl Janouch2016-09-231-0/+5
|
* ZyklonB, kike: Use pledge(2) in OpenBSDPřemysl Janouch2016-03-301-0/+6
| | | | | | degesch has something like "stdio wpath cpath inet tty proc exec" but given that it's user-extensible and very annoying for users to have it crash, I'm leaving it unrestricted for now.
* kike: fix daemonization on *BSDPřemysl Janouch2016-03-131-0/+2
| | | | Bump liberty.
* Fix typosPřemysl Janouch2016-02-091-1/+1
|
* kike: add support for IRCv3.2 server-timePřemysl Janouch2016-01-311-3/+18
|
* Bump libertyPřemysl Janouch2016-01-171-46/+1
|
* Bump libertyPřemysl Janouch2016-01-161-1/+0
|
* kike: more cleanupPřemysl Janouch2016-01-161-12/+8
|
* kike: allow messages before protocol establishmentPřemysl Janouch2016-01-161-9/+4
| | | | We can just queue them.
* kike: asynchronous address resolutionPřemysl Janouch2016-01-161-58/+132
| | | | As well as some refactoring and cleanup.
* Bump liberty, enable TLS SNIPřemysl Janouch2016-01-041-0/+1
| | | | | | | Involves some rewrites to fit the new APIs. SNI has been implemented Mostly just because we can, I don't think it's widely in use and kike doesn't support this feature of the protocol either.
* CleanupPřemysl Janouch2015-12-311-1/+1
|
* Enable TCP_NODELAYPřemysl Janouch2015-11-221-0/+7
|
* Bump libertyPřemysl Janouch2015-08-171-4/+4
|
* More SSL -> TLS renamingPřemysl Janouch2015-07-281-22/+22
|
* SSL -> TLS; fix error handlingPřemysl Janouch2015-07-151-19/+23
|
* Unfuck OpenSSL error messagesPřemysl Janouch2015-07-151-9/+4
| | | | We still take just the first error message in the queue, though.
* Bump libertyPřemysl Janouch2015-07-151-0/+3
|
* Bump liberty, fix map iteratorsPřemysl Janouch2015-07-141-13/+12
|
* Disable SSL 2 and 3Přemysl Janouch2015-07-131-0/+3
|
* CleanupPřemysl Janouch2015-07-121-2/+0
|