We're trying to provide multiple platforms here for those who wish to use other mechanisms (e.g. liberapay and BTC); although interested to know what the concern is with Patreon? We're thinking of adding BountySource too.
Brian Smith had a good string of tweets that summed up my thoughts on it, but I can't find it now. There's two parts to why Patreon feels inappropriate for me: The units are too small to be realistic for a market rate salary for a software developer. If a single developer's salary is six figures, trying to fund a whole team this way a dollar at a time spreads the "crowd" too thin. Second: There's no established deliverable associated with the donation. It makes sense for people who are doing something like making YouTube videos on a regular schedule infinitely into the future, but for a software project, you want to see more specifically how the money will be spent, how much it will cost to complete a project, and what the expected timeline is. Kickstarter's model is much more appropriate for this. The developer can give an honest estimate, "We want to build X. It will take us Y time, and we need Z dollars to make it happen." Then everyone donating can better judge how much to spend and know what they can expect for their money. This is more fair to both the developers and the donors.
What advantage does Matrix offer over IRC?
Server-side history storage, optional end to end encryption, mobile clients with push notifications (the client doesn't have to maintain a persistent connection), voice/video chat, file uploads (and embeds for certain filetypes)... among other things.
Also it cleanly bridges to IRC so it's not a one-or-the-other situation. Was really impressed when I looked at it a while back.
I personally use IRC for text chat, and other tools for, say video chat (Skype), file uploads (a wiki, website, ftp, S3, etc), and so on.
You can also use OTR on IRC for end-to-end encryption, you can use IRC bouncers for persistent connections, bots to record history, etc.
I could see how someone might want all that integrated in one software package or service, though.
well, end to end crypto for one
You can use OTR to provide end-to-end encryption on IRC.
Aside from requiring everyone to install a plugin to do OTR, the crypto behind Matrix's Olm (based on Signal's Axolotl) is much better suited to asynchronus communication than OTR (and MegaOlm actually supports group chats).
There are many more benefits to Matrix, but just making clear that Matrix's end-to-end crypto is better than just plain OTR.
I generally don't think Patreon is a good way to fund software development, but I will be contributing anyway because the work the Matrix team is doing is critical. Open, secure communication on the Internet is one of the most important and relevant issues today.
The Matrix team has been great to work with in developing Ruma. As Matthew mentions in the post, our development has been slowed somewhat by lack of updates to the Matrix spec, but they've had the Ruma team's full support in prioritizing end-to-end encryption in Riot, as it's simply a more urgent priority than enabling independent implementations of the spec. It's been clear to us that the only thing preventing improvements to the spec is the lack of bandwidth on the Matrix team, and in that sense, today's news is actually good news, because now we have a chance to help increase their bandwidth beyond what their parent company could provide.
[0] https://www.ruma.io/docs/matrix/ [1] https://www.ruma.io/docs/matrix/why/