2 ms·
> there’s also a lot of value in confusion/configuration-free usability > it’s between Signal or nothing for the majority of nontechnical folk. This is a fals
by rglullis 6y ago
> there’s also a lot of value in confusion/configuration-free usability
> it’s between Signal or nothing for the majority of nontechnical folk.
This is a false dichotomy. Nothing about federated services that make it intrinsically less usable.
> pretty much Just Works
Until it stopped working. And it failed spectacularly. For everyone.
I might have to repeat this until I go hoarse, but switching from one centralized service to another is not progress. It doesn't make us free. You are still at the mercy of an selected few to take things in the direction they want, everyone is still dependent on their success and the more people adopt it the more difficult it becomes to escape it case the leaders subvert their power to go against the will of the people.
- faitswulff 6y ago> Nothing about federated services that make it intrinsically less usable. I didn't say that. I would say that Signal is more usable than Element, though. While I was commenting, I installed Element and the UX is just not as smooth as Signal - even the quick start instructions aren't as simple: "go to matrix site, click through to Element, read instructions to return to site to configure, install app, register username/pwd, forget about configuration step because it's apparently not necessary after all (?)" vs "go to signal.org, install app, verify phone number." It's not terrible for technically capable folk, but the friction is why it's not as popular as Signal. And while argumentum ad populum is a logical fallacy, practically speaking it is a boon to be popular in most things. Judging by your last paragraph, it comes down to a matter of values - is decentralization important to you? Then hands down, Signal will lose. But for a large number of people who don't care about decentralization, they will do what's popular and easy. OWS, I presume, wants to make sure the easiest, most popular solution is also secure.
- johnchristopher 6y ago> Until it stopped working. And it failed spectacularly. For everyone. The outage wasn't global and didn't impact everyone ? It kept working for me and others in my contact list but some couldn't send messages. Also, if @matrix.org fails or @something fails then it impact users who wants to interact with both those addresses. So in the end it's not that much different for users. I think you comment about the will of the people is a bit off the rails. If anything the whatsapp exodus shows it's not true since we are seeing people moving from whatsapp to signal. When I run my own matrix instance I am still dependent on the matrix guys and what they want to do with the protocol and the de facto only matrix server's code. I can't fix anything with my instance but basic proxy/ssl/dns errors. I can't get the ship to stir in any directions. There are no matrix board where I could voice my concerns about stickers or the default colourscheme. Just like Signal.
- rglullis 6y agoYour argument smells of the "I can't code or don't know Linux, so there is no benefit for me in adopting free software" BS excuse given by those that are to comfortable with Windows/Mac and do not want to give up that away. I am not a "Matrix guy", don't work for Element or the foundation. Yet, I can provide Matrix hosting services for you. If you and more of my customers start asking for changes/improvements (customizations to the clients, better solutions for storage, different kinds of integrations, etc) I can go and develop the changes myself AND DEPLOY THEM on a server for you to use, without needing to ask permission from any of the Matrix devs or Element or the Foundation. Just a very, very real example: I started looking into how I could integrate Hub20 with a Matrix homeserver to allow people to send/receive payments by messaging. It's the kind of functionality that would be super interesting for some folks, it does not require change in functionality of the core protocol in any way and it is very aligned with the Hub20's principle of facilitating access to crypto for the masses that do not want to deal with the complexities of blockchain/wallet/etc. Such an integration will never happen on Signal. Signal being centralized, there is no separate instance that could try out this integration, or make it work only for the accounts that it can manage. Telegram is the same, WhatsApp is the same. Instead of looking simply what is given to you "for free" in the different offerings from the market, start looking at what we can build on them. Don't let your lack of imagination become everyone else's weight to carry. [0]: https:/docs.hub20.io
- faitswulff 6y agoStarting with an ad hominem, are we, now? Most people can't program and are comfortable with Mac/Windows. > Don't let your lack of imagination become everyone else's weight to carry. No one here is shackling you to Signal. You're free to use and customize Matrix, but your argument that people using Signal is no progress at all based solely on being a centralized service is zealotry that is blinding you to the experiences of laypeople. Starting with your root comment - most people use centralized services that have gone down at some point. In many cases they are still using those services, not because they're foolish muggles who can't program, but because it's worth more to use than it is to switch to something else. Centralized or not, people don't really care. This entire thread has been an effort to convince you that most people don't have the same values you do, hence the argumentum ad populum that you so despise being actually a valid metric of success for a project. Popularity is a flawed argument if you're in a formal debate, but it is useful to nearly every other venture. So is Signal the best possible messenger it could be? No, probably not. I, too, would like to see a decentralized, federated protocol take off. But is it measurably better than plaintext SMS or Facebook Messenger? Yes, probably.