Hacker News Reader

Comment by Ericson2314 | Hacker News Reader

As I wrote in https://news.ycombinator.com/item?id=25795575

- WhatsApp: Oh wait, SMS etc. is completely insecure

- Signal: Oh wait, WhatsApp is structurally unable to be a force for privacy

- Matrix: Oh wait, even benevolent centralization is an unnecessary risk

It's not that worse is better, but the general public's imagination can only grow so fast. We need to coax people along. As such, I do think all 3 serve a purpose in certain times and places.

I took advantage of the Signal downtime to get a few people onto Matrix. There will be more opportunities.

View on HN

In fairness, I think WhatsApp's value proposition is and was as little more than 'just' oh wait, sms etc. is completely insecure. I don't think the majority really cared about that (especially since end-to-end encryption was a later addition). I think most people primarily started using WhatsApp because a) it was (is) free b) cross-platform c) worked very reliably

It wasn’t free in the beginning but it was cheaper to send a picture message from it than from MMS

Also for all of the messaging apps out there it's surprising how much feature differentiation they still have. Once you get used to a certain feature set it's hard to move to something with a different set even if it's overall probably better.

Simple Groups and painless multimedia was what gave WhatsApp the edge in the early days

Liked "Simple Groups and painless multimedia was what gave WhatsApp the edge in the early days"

-------

^ That's and ultra low res photo sharing what you get when you group SMS/MMS with folks in the Apple ecosystem. You don't get it in WhatsApp, etc.

I'm not sure what this comment is saying.

It's mocking iMessage trying to convey some extra features (like liking) over SMS which doesn't work very well.

And general costs. Whilst SMS might be free (or effectively free, plans including thousands of messages per month are not unknown) for vast swathes of people, but MMS is most definitely not.

I really doubt it was about security. My bet is on convenience and the fact that it's free for Non-North Americans while SMS is not.

E2E encryption is certainty a reason that non-techies use WhatsApp, particularly in preference to Facebook Messenger since 'Facebook reads all those messages' as a friend said. Which I don't think is technically possible any more, but remains a common suspicion.

Most people profoundly do not understand or care about the vagaries of how their messaging is encrypted.

They do care about SMS fees, especially across international borders.

They do care about sending group texts between Android and iOS and not having things go slowly haywire as messages don’t come across.

They do care about sending messages over wifi if they’re somewhere without a cell signal. (OK, bit of a Canadian thing, but the cottage has wifi and no cell signal.)

> They do care about sending messages over wifi if they’re somewhere without a cell signal. (OK, bit of a Canadian thing, but the cottage has wifi and no cell signal.)

That you have to explain why you might be without cell signal shows that you are blessed up there. Today I talked with someone a few miles outside of Frankfurt (Germany) - huge financial hub - and the line broke down almost 3 times due to bad signal...

A lot of people I knew that used WhatsApp used it because international text used to be prohibitively expensive and it was notoriously unreliable. So most of my friends used it in Europe. Also you can still use it if you aren't on cellular while traveling. Although I'm sure the cross-platform, freeness, and reliability were all pluses.

One thing that I think will always give Signal and WhatsApp an edge over something like Matrix is that it's simply tied to a phone number. You don't "make an account" and have to memorize a password or anything, it just falls into place like any other texting app. The barrier to entry is about as close to frictionless as possible.

There's nothing inherent about this to Matrix, I don't believe.

I mean, maybe? There could certainly exist a phone-verified Matrix server, obviously, but unless the default server that is used for Matrix does that, it's still going to be a bit of friction entering in the server information.

Or even more critically, a discovery server that supports phone number registration so your existing contacts can find you again after they switch.

The only major discovery server on Matrix is currently the vector.im one, and it only supports email identity registration. As many people have pointed out, most users care far more about usability than they do security/privacy, so a service that allows an optional registration of identity using the currently-standard identity tracking system of phone numbers is pretty much a requirement.

Maybe so, but that is completely irrelevant for now.

I have been able to convince quite a few non-tech friends to use Signal, but that would have been impossible with Matrix.

This is the worst thing about Matrix fans gloating over the outage: do you really think the outage helps Matrix become widespread? No, it will make people think twice before potentially taking that next step. And rightly so, because the same problem can happen in Matrix as well. A sudden surge in popularity can make matrix.org go down, and federation or not, that's not going to be a good look.

But alas, growing pains are part of the game. Let's hope they're not fatal for either.

I'm not trying to move people off signal. I of course wish the outage hadn't happened, but I'm not to avoid finding a silver lining when I can.

I'm not positive, but I think the signal protocol and signal messenger came before Moxie worked with Whatsapp to provide e2e encryption.

Not refuting your point at all, just providing additional context.

Yeah I'm charting the public's realizations, not security experts' realizations.

Matrix is merely federated, right? Why is it unlikely that a situation like email or the Internet will emerge, where the network still ends up massively centralized because that's just more convenient?

matrix.org is already 30% of network I read once? For the general public I expect this to continue (at least until we with NixOS making self hosting things far easier someday :P).

However, just as regular users want to get off the Facebook ecosystem, business's should want to get away from GSuite (Really, it's embarrassing if you are anything close to a google competitor) and I have some optimism there. The "element matrix services" of course isn't political decentralization, but is a helpful stepping stone, and still more isolated per tenant than GSuite (e.g. completely distinct accounts).

your causality is wrong here; on day 1, matrix.org was 100% of the network, because there weren't any other servers. since then it's been decreasing steadily - at around 30% atm.

But when Matrix gains massive adoption from the mainstream, wouldn't you expect those people to primarily rush to matrix.org?

When Matrix enters the mainstream, it will be helped along by technical friends telling people about it. Hopefully friends technical enough that hosting their own server for their group isn't too big a deal.

We're proactively working on P2P matrix, as per https://matrix.org/blog/2020/06/02/introducing-p-2-p-matrix/, to prevent this risk.

Thanks for the official response; I look forward to P2P Matrix.

Me too

This is literally the saving grace for Matrix in my opinion. As can be seen with Mastodon, federation tends toward centralization of the servers if there's even an incremental benefit to doing so and nothing encouraging otherwise.

Matrix at least seems to have fewer barriers to usability/functionality when self-hosting, which decreases the detriments of self-hosting and slightly lowers the drive to centralize, but until P2P is default standard for clients the convergence is likely to remain with a mostly centralized federation infrastructure target than a heavily decentralized one.

Thank you, btw. I know it's a hard problem, but there is no substitute for going P2P as far down as you can. To my understanding, there is profound moral merit in your work.

While e-mail usercount is not evenly distributed across many servers, the fact that it's open an interoperable gives you several things:

- you can use whatever client/server you want (or host your own)

- you can still communicate with your friends if they're on another server

- servers are run by separate entities in politically distinct regions

None of these things depend on there being a uniform spread of users across servers.

As I recently said in another thread, an email sort of scenario is exactly what I'd like, or at least it would be many times better than what we currently have for im systems. The most important thing to me (and I suspect many other people in favor of decentralization/federation) is the ability to host my own service. I couldn't care less if 90% of email users are on gmail as long as I can run own email server or pay someone else to run one for me while still communicating with everyone.

I'm sure some people would disagree with my priorities above but for my experience as a user, that is what makes the most difference to me - not being boxed into what someone else thinks my client should look/act like or being locked into their server. Naturally I think more decentralization is better, but that's more a philosophical preference whereas wanting to run my own server or pick my own client is more about usability and control.

Reply on news.ycombinator.com