Age | Commit message (Collapse) | Author | |
---|---|---|---|
2016-11-15 | Support the ident protocol | louiz’ | |
fix #3211 | |||
2016-11-11 | Add missing cstring include for strerror | louiz’ | |
2016-11-11 | Move all the connect() logic from TCPSocketHandler into a subclass | louiz’ | |
This way, TCPSocketHandler only deal with the message sending/receiving, not the connect() or anything else. This will be used for implementing servers (because when a client is accepted, we don’t need all the connect() and dns resolution stuff). | |||
2016-11-07 | Remove a never reached (and non-sensical) error | louiz’ | |
2016-11-07 | Workaround for debian and other old OS that do not have std::put_time | louiz’ | |
2016-11-07 | Add a missing <chrono> include | louiz’ | |
2016-11-07 | Add get-irc-connection-info adhoc command | louiz’ | |
fix #3171 | |||
2016-11-01 | Refactor load_certs() | louiz’ | |
2016-10-31 | Some cleanups | louiz’ | |
2016-10-31 | Remove unused roster code | louiz’ | |
2016-10-31 | Trivial cleanup | louiz’ | |
2016-10-31 | Make AddrinfoDeleter a class | louiz’ | |
2016-10-31 | Rename a variable that shadows a class member | louiz’ | |
2016-10-27 | Refactor remove_invalid_xml_chars to use correct types directly | louiz’ | |
2016-10-27 | Directly use Botan::byte instead of char, to avoid an unnecessary cast | louiz’ | |
2016-10-26 | Refactor the sha1 digest into its own function, and do not use sprintf | louiz’ | |
2016-10-20 | Optimize tcp_socket::on_send by using vector::erase() only once per call | louiz’ | |
2016-10-20 | Very little optimization by using a simpler scope_guard when possible | louiz’ | |
The version with the vector, that can be disabled etc, is “very” slow, so we use unique_ptr when we don’t need to disable it, and when it only contains one function | |||
2016-10-12 | On EINPROGRESS, we need to also check for read events (because openBSD lies) | louiz’ | |
“It is possible to select(2) or poll(2) for completion by selecting the socket for writing” Yeah, sure, “writing”… | |||
2016-10-12 | Fix an off-by-one issue in the POLL code | louiz’ | |
2016-10-11 | Conditionally use strptime if we don’t have std::get_time | louiz’ | |
2016-10-11 | Include a private and no-copy nodes in private <message/> to avoid carbon ↵ | louiz’ | |
duplication | |||
2016-10-11 | Parse the timezone myself, instead of using the broken strptime | louiz’ | |
See https://lab.louiz.org/louiz/biboumi/issues/3215 https://github.com/andikleen/glibc/blob/master/time/strptime_l.c#L746-L747 for why strptime() sucks We use std::get_time now, to parse the date and time. And we parse the timezone by hand. fix #3215 | |||
2016-10-07 | Correctly set status="110" in the presence for the target of a kick | louiz’ | |
2016-10-06 | Fix the argument of strerror after bind() | louiz’ | |
fix coverity CID 134470 | |||
2016-10-06 | Remove a branch that execute identical code in both cases | louiz’ | |
fix coverity CID 134469 | |||
2016-10-04 | Respond to muc#traffic requests | louiz’ | |
fix #3069 | |||
2016-10-03 | Avoid an exception due to some bad logic in the DNS resolution mechanic | louiz’ | |
fix #3207 | |||
2016-10-03 | Fix some compilation warning/errors that appear on FreeBSD | louiz’ | |
2016-10-03 | Fix parse_datetime by always using a 'z' as the timezone | louiz’ | |
Because some plateform accept Z and z, but some only accept z… | |||
2016-10-03 | Implement result-set-management for LIST queries | louiz’ | |
ref #2948 | |||
2016-09-29 | Re-add the ad-hoc command the was removed by mistake in the previous commit | louiz’ | |
Thank you, e2e tests | |||
2016-09-29 | Add AdhocCommandHandlers::add_command to simplify the usage of this class | louiz’ | |
And make things a little bit clearer | |||
2016-09-29 | Look for uuid/uuid.h instead of just uuid.h | louiz’ | |
Avoids a conflict between /usr/include/uuid.h and /usr/local/include/uuid/uuid.h on freebsd | |||
2016-09-28 | Remove unused code in sha1 module | louiz’ | |
2016-09-17 | Fix the parse_datetime by using %Z instead of %z | louiz’ | |
If anybody knows why fedora accepts both, but it only works with %z on debian, please tell me. | |||
2016-09-04 | MAM results can be filtered by start and end dates | louiz’ | |
2016-08-31 | Add parse_datetime | louiz’ | |
2016-08-31 | Fix to_string(time_t) and write a unit test for it | louiz’ | |
2016-08-29 | Add the missing includes back | louiz’ | |
2016-08-28 | Respond to MAM requests on a channel JID | louiz’ | |
At the moment, result-set-management is not implemented, the whole history (well, at most 1024 messages) is returned. | |||
2016-08-23 | Use Z instead of z in the timestamp format | louiz’ | |
2016-08-23 | Don't use put_time() because it’s not in gcc 4.9, in shitty-debian ↵ | louiz’ | |
“stable” Use strftime instead | |||
2016-08-22 | Add missing ctime include | louiz’ | |
2016-08-22 | Move get_first_non_empty to louloulibs/utils | louiz’ | |
2016-08-22 | When joining a channel, send the most recent history found in the database | louiz’ | |
2016-07-28 | Fix the timeout test, now that we don't wait 1ms too much everytime | louiz’ | |
2016-07-28 | Do not add 1ms to the timeout of our poller | louiz’ | |
Can’t remember why I did this, but that must be a stupid reason. Everything must work even with a timeout of 0. | |||
2016-07-28 | Always remove all the DNS sockets on an c-ares event | louiz’ | |
Because c-ares may close one of its socket, even if it’s not the one that has the event. Otherwise we may not know when a socket has been removed from our poller (automatically, when close()ed) and this leads to bugs. | |||
2016-07-27 | Revert "Don’t use unique_ptr to store dns socket handlers" | louiz’ | |
This reverts commit 5328d0806fdc5becb9344b4d4320787a2b7c0712. |