don't casefold extban characters #307
Conversation
There was a problem hiding this comment.
This is not sufficient. There is another extban_table[(unsigned char) irctolower(*p)] below and there is also code that normalizes the extban character to lowercase when setting a ban. Perhaps this normalization should be skipped only if the uppercase extban exists, to avoid breaking existing behaviour.
Also, the uppercase and lowercase extbans need to know about each other since it's not possible to have two bans in a channel that only differ in case. Some clients such as irssi track ban lists so this may well be a client restriction as well.
|
I don't mind about the client angle given other popular IRCd software does this |
(though I'm willing to lose this debate) |
|
It seems like it would be best to make only the extban character case sensitive, so users manually typing in ban masks for Another alternative might be allowing extbans to be specified by whole words. I guess this was originally not done to avoid having too many ;-) |
is this not what the code in this PR currently does? |
There was a problem hiding this comment.
This looks alright, but,
Is there anything that expects this to always be casefolded outside of itself?
Does this have a plan to migrate or otherwise mitigate extbans that were set previously? Possibly checking if an existing extban would match one if casefolded?
No description provided.