Live data from Hacker News

The Matrix Trashfire

blog.koehntopp.info

191–200 of 217 posts

Re: The Matrix Trashfire

#191
post #170
post #168

Earlier quoted context omitted.

> difficult to actually federate as most XEPs are optional or not supported on your federation partners This is false.

Voice/Video is an optional XEP and if it's not supported what happens to the client exactly? "this is false" is a terribly glib statement with literally no backing and can only be said if a person has either zero knowledge of what they're talking about or they've tied themselves to a single implementation of XMPP everywhere, which is essentially standardising a bunch of XEPs.

Well if you use a matrix client that does not support video calls (Syphon), you also can't be called.

Camera and microphone are not a mandatory feature of Matrix either (;

Re: The Matrix Trashfire

#192
post #124

I really don't like the forced end to end encryption of chats. I do not care that much about the security of my chats to justify the complications of end to end encryption -- verifying devices, constant "reset" prompts that I don't even know what they do, but they seem like they are very destructive. I trust my homeserver, since I host it myself. I do not need end to end encryption. I understand this is an issue of t…

tbh the best thing about Matrix is that you have a hope in hell of actually federating it. XMPP had that promise but due the XEP situation it quickly became difficult to actually federate as most XEPs are optional or not supported on your federation partners. Everything you said is true; if you can tolerate centralisation then just stick to XMPP, it's pretty good.

Why centralization? XMPP users are identified with their JID in format username@server.example, similar to email addresses. And server to server communication is very well documented and major XMPP servers for instant messaging (ejabberd and prosody) both allow server to server message exchange. I don't see any benefits for Matrix when it comes to federation and I wouldn't agree with you that by using XMPP you tolerare centralisation at all.

It's level of centralisation is comparable to that of email (and email is very well federated in my opinion).

MUCs like the XEP you mentioned are handled by a different XEP.

Re: The Matrix Trashfire

#193
post #182

Me too I tried to get Matrix working to have a conversation. Many worthy people I know use it But I failed in a similar way Do many people actually use it? When it is so easy to use Signal or (Dog help us, WhatsApp)

We definitely have work to do on the onboarding experience, but I'm pleased to say that there are 115M addressable users on the open federation – so many people are having great success once they get past the initial friction. Aside from all the FOSS projects that use Matrix, it's also used by the German healthcare agency, French civil servants, NATO, a number of universities including MIT and TU Dresden, Moodle, and many others.

We're moving quickly to address the feedback in the blog post and will be investing more in docs and UX to address the friction.

Please don't hesitate to share your own experiences if you run into trouble! Stuff like that is a real gift. We're always looking to learn and improve.

Josh, Managing Director of the Matrix.org Foundation

Re: The Matrix Trashfire

#194
post #4

We also tried to use it, but frequently, messages will fail to decrypt with no option to retry. Threads is a mess, where messages will show as unread, but you can't actually see what message was unread. Matrix/Element is so close to a great alternative to Slack, but in it's current state it's totally unusable.

> We also tried to use it, but frequently, messages will fail to decrypt with no option to retry. This is a years-old issue with Element, which never happened to me with other sending clients such as FluffyChat. It's unbelievable that it's unfixed given it's a dealbreaker as it results in permanently unreadable messages on your end (the "waiting" in "waiting for this message" is a lie). And since this needs to be fix…

This just sounds like a description of Matrix's key sharing mechanism? Messages are supposed to be unencryptable if you don't have the keys, and bringing online another device (or having all your keys pre-shared so you don't have to) is what provides the keys. If you want to avoid this altogether, the UI prominently advertises the optional encrypted key backup service provided by the homeserver, and various manual options for sharing keys.

If FluffyChat is not having this issue, it is probably overly eager to share encryption keys instead of allowing the user fine-grained access to control keys, which is successfully hiding the complexity of the ratchet encryption but potentially exposing the user to attacks to force the sharing of keys.

Edit: I was looking around. While Matrix is well documented, Element's documentation is poor because they expect you to figure things out from popups in the UI -- fair enough, unfortunately most apps are like this, and Element's popups have gotten a lot clearer. But I did find these two pages from a university that seem to serve well as "Element's missing manual". Worth a read if you are trying Matrix for the first time, because it discusses some things that can look like bugs but are really user error.

https://docs.matrix.kit.edu/en/settings/

https://docs.matrix.kit.edu/en/faq/

Re: The Matrix Trashfire

#195
post #23
post #9

Earlier quoted context omitted.

> his comes off as the author just not doing basic reading. Perhaps a technology will not have success if its users need to do basic reading.

A chat platform for the illiterates might not be the best business proposition Jokes aside, I don't think matrix/element, at this stage, are trying to overthrow telegram or whatsapp. It seems their main approach is somewhat aimed at the people who use(d) IRC on the general audience side and institutional clients who they work with to create ad hoc solutions for employees, in which case which client to download is sli…

As the Managing Director of the Matrix.org Foundation I can assure you I'd love nothing more than to displace centralized, proprietary communication tools.

It just so happens that right now it's easier to land with folks who are patient with sharp edges and already believe in the value of FOSS, E2EE, and decentralization. Gotta start somewhere, right? :)

IDK if you caught it, but the project lead, Matthew Hodgson, gave a main stage talk at FOSDEM a couple weeks ago and offered an update on the project and, in particular, on how we're taking advantage of the push that regulators are making for interoperability. WhatsApp, in particular, gets mentioned in this context and the writers at WIRED and Tech Crunch seemed to pick up on that!

Re: The Matrix Trashfire

#196

Earlier quoted context omitted.

> Lastly I think comparing an open source project like Matrix/Element to Publicly traded corporations like Slack or Meta, is not fair. If we want to "win" (reach similar/higher adoption), we need to at least come close. It's not easy work, but not doing it and leaving the product that so much good work has ready gone into unusable for a vast number of people would be a bummer.

Not sure we want to "win". If endusers want apps "to just work" without "paying up", then I would recommend them to stay with Whatsapp. That kind of users aren't worth the hassle if they have no money. I personally have better relationships with people that enjoy learning something new, and coming up with solutions for issues themselves, eventually contributing to the ecosystem.

Oh wow, since when there is an option to "pay up" into the Matrix ecosystem and get a solution that just works? Could you point to it?

Re: The Matrix Trashfire

#197

Earlier quoted context omitted.

I always give the power to my QA team to block any release no matter what and to give higher priority to tickets than product manager. If CEO wants to override, I cover them and take the blame. This is not a guarantee that there will be no bugs in production, but it saved us a few times.

The Japanese testers were the best I'd ever seen. They never reported a "NotABug." They could back up every report, and give exact reproduction steps. They found weird, obscure corner cases, and that was by hand (they hated automation tools). They had 3,000-line Excel spreadsheets. If even one of those rows failed, the whole shooting match (like an entire product line) could come to a halt (so that meant they had to…

> They never reported a "NotABug."

"Every ticket should result in a code change, *or* a documentation change"

was one of the coolest sentences I ever read on some web page.

Re: The Matrix Trashfire

#198

Earlier quoted context omitted.

> We also tried to use it, but frequently, messages will fail to decrypt with no option to retry. This is a years-old issue with Element, which never happened to me with other sending clients such as FluffyChat. It's unbelievable that it's unfixed given it's a dealbreaker as it results in permanently unreadable messages on your end (the "waiting" in "waiting for this message" is a lie). And since this needs to be fix…

This just sounds like a description of Matrix's key sharing mechanism? Messages are supposed to be unencryptable if you don't have the keys, and bringing online another device (or having all your keys pre-shared so you don't have to) is what provides the keys. If you want to avoid this altogether, the UI prominently advertises the optional encrypted key backup service provided by the homeserver, and various manual op…

reading a manual for a chat app to fix 'user error' that results in bizarre behaviour like this is not a reasonable solution.

Re: The Matrix Trashfire

#199
post #98

There are a lot of sharp corners, but going for the beta option was not a wise move.

Unless I read it wrong, the author chose element x because it was available on all platforms. It looked like element wasn't available in the Mac app store

Ah! I see now.

Re: The Matrix Trashfire

#200
> All in all, this is a mess, and my recommendation is to avoid Matrix for at least two years.

Two more years. I've been a matrix user for almost a decade. Things have not gotten better at all.

I used to use Riot, I invested (and subsequently lost) a good amount of social capital on boarding my close circle, mostly non technical people, to using it. We used it. It was a little janky with the key sharing and verifying users and what not, but we managed. Then Element came out, and at first, whatever, but then a lot of those verification tools just stopped working. Several of our accounts became locked in some limbo where they couldn't be verified because the verification methods didn't support legacy verification. Nobody wanted to move to a new client which was less feature rich. We moved to signal eventually and have been mostly happy ever since.

I still use it with a couple of people. Its still just as janky as ever. First riot, then element. Now Element X? Dendrite... How many times are these guys going to rebuild from scratch before they realize they're in over their heads?

I just gave up on it, I still use it with the couple of people I'm connected to and I don't do new contacts on it with anyone. I treat it like I do Discord. For me now it's XMPP and email, and signal for real world contacts. I mess with some of the newer messaging tools out there like Simplex (which absolutely has a terrible onboarding flow, but they're new so they get a pass for now) and things, and I hope to get utility out of them, but I'm never going to burn myself again pushing people to use something new, I'll wait until these things prove themselves, a lesson I learned from Matrix which never did mature into anything worthwhile.

Post reply on HN