Hello everyone,
I have discovered SimpleX Chat (nothing to do with XChat or HexChat, or the favorite letter of some dumb billionaire), and it appears being a legit good effort at providing good privacy while retaining “mainstream” usability.
And it has been audited (by one company so far, it seems).
The only concern I have is with regards to battery life (given that it has to maintain roughly as many open connections as you have contacts, AFAICT).
Has anyone here used it? Any opinion?
Very good answer, thank you very much.
WRT battery life, the only device I have been able to use it with is a second hand device I just got, and have little metrics for. So, I’m not sure if the device battery is worn out, my expectations are from another time, or it has battery “problems”. Given your testimony, probably one of the two first options. Maybe both.
I have tried Tox, Signal, Briar, Amethyst (Nostr) and lastly SimpleX.
I dropped Tox because of network and battery usage, and didnt actually get to try Amethyst because of its use of Google Push.
Usability wise, and userbase wise, Signal is by far the best, but its use of phone numbers as identifiers is absolutely terrible in my opinion.
Briar has features that others don’t have, and works ok, but isn’t really feature-complete when it comes to mainstream use.
SimpleX strikes me as a good middle ground between the two, with a very complete featureset, even if, as you mentionned, it is rather hidden behind all that UI.
Also thanks for telling me about the Desktop GUI, I was searching for one and didn’t find it. So I’ll look again.
Now, concerning the absence of identifiers, the marketing material clearly mentions “[not] any user identifiers”. As I understand it, it still has identifiers, but as conversation endpoints, and they are unique to a given conversation. So, yes identifiers, but their meaning is a lot harder to infer than with user identifiers. It kind of is like with using unique cryptocurrency wallets per contact, and making transfer through exchanges, converting between currencies. It is a lot harder to track.
Agreed that having the option to run your own server is invaluable. One could for example deploy an entire SimpleX infrastructure on a different network, such as DN42. And in the event of a global apocalyptic event, the distribution of both clients and servers would allow users of local or regional networks to still have usable private messaging.
And as for people adopting the app, it is via people like me and you. I run the operations for around 6-10 people in my immediate surroundings (friends and family), and my recommendations mean a lot to them (it often influences entirely what they get to use). Besides, I also advise people professionally. And, so, assuming we each influence a dozen people on average, they will, in turn, create momentum for their own social circles. That’s exactly how gmail gained traction.
Now, I don’t see people who are using only whatsapp adopting SimpleX any time soon, but honestly, those people aren’t using Signal either (and if they are, it is very temporary), and will always be the last to move.
Well, thanks for taking the time to answer me, in turn!
Mind you, this is very recent and it’s in the releases page of their GitHub under a pre-release. It’s in the assets of the 5.3-beta release, which, now that I’ve checked, has packaging for MacOS, Ubuntu and AppImage. They’re the ones with the *-desktop affix.
Yes, I think you’ve done a better job of explaining it than me. It’s impossible, to my knowledge, to communicate without any kind of identifier, but their model is a rather ingenious one for people concerned with privacy. Couple that with onion routing, and I feel very safe talking to people on the app.
You sound more hopeful than I am, lol. But I too hope that technologies such as SimpleX take off, if only because of early adopters such as us.
Edit: also, something that SimpleX does is markdown editing, which is just… 👌
Great info, thank you!!
First, 🙏
Second, if you generate an entirely new key for every next message, appending it at the end of the current message, while merely depositing the message at a known place (deaddrop), while using tor (or similar), there is literally no way to link two messages without decrypting the first. That would forego any kind of identifier, but if a single message gets lost, communication entirely breaks. So, I’m no a cryptography expert, but I believe there are ways to do a similar design (mitigating the shortcomings), and eliminate identifiers entirely.
Yes, maybe, but also, a sudden swing in userbase can happen, look at Reddit and Lemmy. So it is important to have good, usable software (at least moderately) ready to kick in with a sudden increase in adoption (modulo server loads, this can usually be solved with more servers). And that is IMHO where our most important role is: bringing normies to the group, getting their feedback, and relaying that feedback upstream. So that when the user rush happens, the app/platform isn’t immediately cancelling the movement due to inaccessibility/poor UX.
Yes, it is quite essential, I agree there too. Signal has formatting, but you need to use the UI to set it, and that just doesn’t feel as right…