Currently we are using IRC for many functions (it's one of the backchannels for most major events, it's the one of the main places to get tech support, it's a somewhat active nontechnical support forum for various communities and the place to get urgent help from e.g. stewards, it's used for all kinds of real-time tech discussions like deploy coordination), but for all the well-known UX issues it makes it harder for new users to engage the community. We sometimes link to the Libera.chat web GUI as a way of making participation easier, but it's not great. Matrix is probably the main FLOSS alternative today - it's really its own chat network, but it can be bridged to IRC and so a Matrix client can be used as an IRC client. Most notably, their standard web client, Element (formerly called Riot), could be used. This task is about evaluating Element as 1) an alternative interface that can be linked to from a web page, 2) the "officially" suggested IRC client for people new to IRC.
See also: T186061: Evaluate Matrix / Element as the recommended chat system for Wikimedia (as a standalone communication network)
(the part below is outdated and needs to be reevaluated)
Riot roadmap from early 2019, 2019 report / 2020 roadmap (not super relevant since most of these features won't work over IRC)
Major issues:
- vector-im/riot-web#8646: room search doesn't work. Blocker for recommending as a generic client, not an issue when using it via web links (e.g. for an event). Seems like this is a regression that will be fixed soon.
- vector-im/riot-web#9264: no guest access. Blocker for using via web links.
- Freenode antispam features confuse the bridge
- matrix-org/matrix-appservice-irc#938: when joining a +r room, the bridge tells Matrix the user is unable to comment, even if the user is registered.
Minor issues:
- privileged actions don't translate well
- ops using Riot don't receive messages blocked by reduced moderation mode (+mz)
- vector-im/riot-web#11625: interactions with the bridge are confusing