4 ms·
Hmm ... - https://upload.wikimedia.org/wikipedia/commons/d/d5/IRCd_software_implementations3.svg https://upload.wikimedia.org/wikipedia/commons/d/d5/IRCd_sof..
by gpmcadam 9y ago
Hmm ...
- https://upload.wikimedia.org/wikipedia/commons/d/d5/IRCd_software_implementations3.svg https://upload.wikimedia.org/wikipedia/commons/d/d5/IRCd_sof...
- https://xkcd.com/927/ https://xkcd.com/927/
- prophesi 9y agoThere's nothing wrong with there being more implementations of ircd. Each one is based on the same protocol, with their own additional features. The XKCD comic would be relevant if they were implementations of -different- protocols.
- progval 9y agoA problem with IRCds is that a lot of them are very similar (charybdis, ratbox, ircd-seven, hybrid), which duplicates the efforts needed to fix bugs and add features, for very little benefits. And the benefits each fork has generally is extra non-standard modes, which complicates work on the client-side. For instance, +q is a channel owner (= someone very powerful) on UnrealIRC, but it's a muted user (= kind of a ban) on ircd-seven
- prophesi 9y agoTrue, anything non-standard will complicate things. But it shouldn't complicate much, since the core functionality is standardized. Accidentally alerting the user that he's now a channel owner instead of a muted user is a trivial issue. EDIT: Also, there's work on getting modern irc server features added to the protocol, which could possibly standardize those additional modes. https://ircv3.net/ https://ircv3.net/
- progval 9y ago> Accidentally alerting the user that he's now a channel owner instead of a muted user is a trivial issue. But it's not if you accidentally make someone an owner instead of muting them. > EDIT: Also, there's work on getting modern irc server features added to the protocol, which could possibly standardize those additional modes. https://ircv3.net/ https://ircv3.net/ Yeah, the IRCv3 Working Group is doing great work