Age | Commit message (Collapse) | Author |
|
|
|
According to `man 2 getrlimit` resource limits are inherited by
the child when a process forks.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Closes #546.
|
|
|
|
|
|
|
|
|
|
We only build on Xenial nowadays so this check is unnecessary.
|
|
|
|
|
|
|
|
|
|
As recommended at https://www.freedesktop.org/wiki/Software/systemd/NetworkTarget/.
Closes #1729.
|
|
This message exists for an incredibly rare issue and just confuses
the vast majority of people.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
This might be causing issues for some people?
|
|
Tell the user when they are extbanned rather than incorrectly say
that the channel mode is set.
Refactored the logic in m_nonotice to match that of the others.
|
|
Now that casemapping is configurable in the core and we set that
Config variable rather than just modifying the ISupport output each
time, we need to change the variable back when being unloaded. So we
save the current value when loading and set it back when being
unloaded. We also need to call the ISupport builder a second time as
the core calls it before we destruct.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
- The connection timeout can now be customised with <database:timeout>.
- <database:port> is now limited to the 1-65535 range.
- The MySQL library will now not install a SIGPIPE handler as it
would override the default InspIRCd ignore behavopur.
- Errors caused by setting the default character set and executing
the initial query are now no longer ignored.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Many clients don't reconnect automatically and it can take a while
for them to reconnect.
|