Setting AuthOnlyViaModule on a user causes CheckPass to never return true,
causing all authentication attempts using the configured password to fail, both
on IRC connections and for webadmin. This is useful in situations where an
external module (cyrusauth, certauth, imapauth) handles authentication. Setting
the global AuthOnlyViaModule option causes similar behavior across every
user. If AuthOnlyViaModule is set to true globally, it cannot be overridden
per-user.
Close#1474Close#331
I noticed that the default network flood rates are wrong (for most
networks). It has been very standard since the beginning of IRC to use a
value of 2 seconds per line, with a 10 line head start. Almost all
networks work this way, and if you don't follow, having a lot of
channels or a lot of clients connected to ZNC results in excess flood
disconnections and much frustration for the user.
I think you should make the default value for these 2 and 9 instead of 1
and 4 so that this nasty bug doesn't hit people who will not know how to
debug it. You find these same types of settings built into every IRC
client. 2 and 10 is the standard. 2 and 9 provides a slight buffer so
there is no mistake.
Thanks
-Rubin
Close#1416
This also alters PutClient such that the CMessage variant handles
sending messages, rather than the CString variant. As a side bonus, this
gives callers better information on whether the message was sent to the
client. Additionally, it eliminates the need for a hook to let modules
set the tags sent to a client, as that can now be done inside
OnSendToClientMessage.
The output style of tables was recently changed. This broke the
ControlpanelModule integration test. Adapt some of the patterns used so
that they work again.
Signed-off-by: Uli Schlachter <psychon@znc.in>
Previously, "more or less unrelated" things would fail. After this
change, also a message is printing that gives a hint on what went wrong.
Signed-off-by: Uli Schlachter <psychon@znc.in>
A few different implementations of computing the current time were
spread out through the code base, most of them using gettimeofday().
This centralizes the logic in CUtil::GetTime() for easier maintenance,
and also allows all call sites to get the benefit of the clock_gettime()
code path on systems that support it.
The gettimeofday function returns 0 for success, not for failure. As a
result of the inverted logic we were losing millisecond precision when
parsing incoming messages on non-HAVE_CLOCK_GETTIME systems (macOS).
