Age | Commit message (Collapse) | Author |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
It is possible for the cap to be null when a DEL for an unloaded
module is sent out so we should not blindly call it.
|
|
|
|
This avoids announcing a bare sts cap when the config is wrong.
|
|
|
|
|
|
|
|
See also: #1749.
|
|
This is a partial reversion of 687778b72e.
See also: #1749.
|
|
|
|
This is similar to the oragono.io/nope capability only instead of
killing the connection it just rejects the request. This should be
less intrusive for users.
|
|
|
|
|
|
|
|
This fixes a minor bug in the filter module where the target would
be blank in messages when a server-targetted message matches a
filter.
|
|
|
|
|
|
|
|
|
|
This fixes a confusion created in c299adb31a where users without
the message-tags cap would get errors instead of being silently
ignored.
|
|
|
|
|
|
- Ensure that the config is actually valid. This found an invalid
entry in the default helpop configs.
- Columize the index page and build it whilst reading the config
instead of at request time.
|
|
|
|
|
|
- Cache the codepage name on load and restore on unload.
- Make origisnick constant.
- Fix the size of the casemap in memcmp call.
|
|
|
|
Closes #1745.
|
|
Closes #1744.
|
|
|
|
|
|
This name is a lot lessambiguous.
|
|
Sending historic CTCPs to clients can only end badly.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
This message exists for an incredibly rare issue and just confuses
the vast majority of people.
|