Email sucks

Oh yeah! I heard about that. That looks great! Will definitely try it next time I have the need.

Hmm... OK. I could have sworn I was required to give a phone number when I tried. But maybe I'm remembering wrong? Good to know, then!

Getting locked out is a real risk, and one that creeps up on you, as you start with an account you don't care about and use it more and more over the years.

I hope they're printing out backup codes and have a safe place to keep them.

Really? When I tried it, it wouldn't allow me to do anything without entering an SMS-capable number. Maybe it's changed since then? Or maybe my private browsing set off some sort of "scam" detection?

Thanks for your comments! Also, my bad...when i read both your posts, i failed to click through on the link to I2P-Bote (I think i focused too much on the blockchain angle and not enough on the email aspect), so of course failed to read the I2P Bote features for email replacement. Quite interesting; and thanks again!

Another fun fact, Twitter's original 140 character limit was to allow use via SMS (with 20 chars reserved for person's handle in DMs).

I miss when Twitter was functionally just the "Current Status" bar in AIM as a standalone service. I kind of wish that the company focused on being some kind of mass market Identity as a Service platform to work as a sort of universal login instead of the ersatz blogging platform it turned into. It's something the internet really could have benefitted from, and would have allowed Twitter to work as a functional alternative to Facebook by enabling a federated network of interlinked blogs and multiple different kinds of social media sites.

I mean, they could even have built some of those modules for blogging or photo sharing or whatever too, just making it interoperable with others. Alas, what could have been.

Damn, you are correct. I apologise for spreading false info.

I swear, I remember seeing a presentation about the history of e-mail protocol and there they also in short discussed how SMS are (loosely) based on the e-mail protocol as well, showing different similarities. But that was years ago and I can’t find it anymore, so it seems I either misremembered or the presentation overstated the (few?) similarities.

As a return favour, I can confirm that this is my memory of the Twitter reason as well.

I'll add that matrix is still in early days, and i really believe that it still needs evolving...but i'm a believer in its flexible nature, and that it can provide a neat, decentralized, self-hostable, secure messaging offering for numerous use-cases.

The potential problem with this argument is that I can remember the same arguments being made about XMPP many years ago, and yet here we are, 20 years later, and those hypothetical clients and interfaces outside of the instant-messaging paradigm never seemed to develop.

In part, my frustration about this is that people have been criticizing email on these points for decades, while pointing to protocols that could hypothetically become alternatives, but never do. Meanwhile, email continues to stagnate.

And e-mail proper is hardly used for long, carefully-written letter-like messages.

I'd say this is a culture problem. Internally, within my team, I try to maintain a norm where quick questions and general chatter are done in Slack but all formal requests and notices are sent as emails. Often, if we have a conversation in Slack, I will send an email to the participants summarizing the topic of the conversation, the positions argued for or against, the group consensus, and next steps for each person in a separate email.

Synchronous chat is just too chaotic and disorganized to make for easy reference, especially if you're looking back long after the fact when you've forgotten the context or if you're trying to get a third party to weigh in.

Email does get overused for pointless missives, notices, promotions, quick-questions, and status-updates. IMO all those would be better served by direct messaging, a phone call, a chatbot, or some kind of RSS feed. But if you treat emails like a tool for writing formal memos and letters it works quite well at that. Most of the complaints people have about email come from senders abusing it and not respecting the recipients' attention. I think this is, largely, a consequence of work hierarchies. Managers were the last to catch on to email and are the most likely to use it in inconsiderate ways that are disrespectful of their recipients' time and attention. It's created a culture where we've normalized a lot of conventions around how people use email that are also, consequently, disrespectful of peoples' time.

Now, maybe there are some people who are actually using e-mail as their primary mode of communication, but I know I don't, and probably neither does the majority of e-mail users

In my experience, the vast majority of the academic/scholarly research community uses email as their primary mode of communication outside of live meetings, and in many cases, do so through relatively long, letter-like emails.

Our research group did try to use Matrix, but everyone switched back to email, despite my insistence on continuing to try it. It's enormously clunky for longer-message discussions, and discussions that don't fit around relatively UI-expensive rooms; I also seem to recall some considerable frustrations with the inflexibility of MXC as implemented, and the poor support for equations/etc.

Also, you are mostly talking about the features, or lack thereof, of some Matrix clients:

I'm not sure that I'd agree (and would also mention that what I wrote certainly applies to all Matrix clients I'm aware of). While there is always the hypothetical-client argument that was also often made for XMPP, the Matrix protocol itself is primarily built around rooms and a room-first, rather than message-first, architecture. It seems possible that the DAG architecture of events could be used for some form of threading if handled somewhat differently, but messages would still go to rooms. In all clients, rooms are relatively "heavy", and this is the case in server implementations as well: creating a new room for each new topic of discussion, inviting everyone, having everyone join, etc, would seem problematic, particularly when room deletion is generally not reasonably implemented in Synapse.

On the other hand, however, seeing a room as a conversation, and restricting addressing to the entire conversation rather than individual messages, could potentially be a significant improvement over the common problem of people not understanding when people have been added to/removed from emails, and what quoted material is still present, a problem that I've frequently experienced, sometimes with very problematic consequences, when boards use email for discussions.

Actually, all of this now makes me wonder about the viability of an email-like Matrix client. I would imagine this might involve:

• Rooms reenvisioned as conversations with topics rather than names, and sortable/filterable by date/etc, along with abilities to archive
• Conversation invitations dealt with in some more efficient way: eg, auto-accepting invitations from some list of trusted users, and listing unaccepted invitations as unopened conversations in some sortable/filterable way.
• Split message-list/message-content view, and message threading, rather than an instant-message UI.

Incidentally, I already write all my e-mails long-form content in an external editor: the lack of my Emacs shortcuts is maddening.

So do I, actually, often using mu4e. To sound particularly horrible, this is in part to be able to write reasonable HTML emails via org-mode or pandoc, so that I can include figures and equations.