I’ve tinkered a bit with Stalwart, and Cyrus is, of course, a known entity. I’d be hesitant to introduce a big Java thing to the stack.
JMAP is still a draft, AFAIK. I’ve been on the mailing list for like a decade, and I don’t think I’ve seen it being accepted as a standard. That doesn’t necessarily mean it isn’t useful or reasonable to deploy (but, it’s still a question whether it is useful yet, as there isn’t great client support…no point in having a JMAP server if nobody connects to it).
And, worth thinking about whether JMAP actually provides anything valuable for non-web based clients. For web-based clients it makes a lot of sense, because the client can query the JMAP server directly (should such clients exist, but they currently don’t really, AFAIK the webmail clients that I’m aware of don’t have production-ready JMAP support, and Usermin certainly doesn’t), but for something like Thunderbird, IMAP is already quite efficient…JMAP probably can’t beat it, it just changes the format of the communication back and forth. Newer isn’t necessarily better in this case.
So, I like the idea of JMAP for webmail clients, but none of the webmail clients I pay attention to has support for JMAP.
As you know (but sharing for others), JMAP is not just for email. It is also for calendars, contacts, Sieve filters, and more. The email part was stabilized first and the rest has seen more recent work.
We started in 2019 but after, our progress stalled. JMAP was a low priority and our active devs worked on some major architecture improvements and many new features. But now, things have changed. We have more devs and JMAP made it to the top of the priority list. Also, the project is globally more healthy. We have moved the project to a GitHub organization, and split up the responsibilities among a new team. In Cypht 2.0.0, we had 14 new contributors! (vs the previous version)
As of now, we only have one open issue related to JMAP within Cypht: https://github.com/cypht-org/cypht/issues/931. In theory, everything else works. Since this is so new, I do expect we’ll encounter issues and we will work in priority on all issues related to JMAP.
Thanks, Marc. It would be great to have a demo available in www.cypht.org/demo to log in and look around.
I also checked Cypht installation instructions and couldn’t find any for installing Cypht under a user domain. If it’s possible (and it might be), we could consider adding an automated script installer for it.
JMAP is definitely hindered because systems with huge market shares are not supporting it (namely Roundcube and Dovecot). Time will tell but now that we have 3 servers supporting JMAP, I think it will grow.
So it looks like Cypht is the first/only Open Source webmail client supporting JMAP. While it’s nice to be the trailblazer, I hope many others will join the club
On a related note, here is a page to mainly compare all active and mature Open Source (thus self-hostable) webmail solutions that you can install and manage email from any server with standard email protocols (IMAP, SMTP, JMAP). And Virtualmin is listed for reasons explained on the page. https://wikisuite.org/Webmail-and-groupware-comparison
That’s not how I read those threads. To me it looks like a bunch of people say they want it, but no one wants to develop it or pay for it, so nothing has happened (and it proves people don’t really want it, because if they did they’d do something about it).
I would be surprised if Thunderbird gets JMAP support within the next year, or even two, given the rate of progress and the amount of time/attention being spent on it.
If you want it urgently, I think it’s going to require you to do it or pay for it to be done.
