1131
Samsung joins Google in RCS shaming Apple
(www.theverge.com)
This is a most excellent place for technology news and articles.
Why should anyone care about RCS? The trend has been to get everything into data instead of carrier owned services for two decades now, we don't need another SMS (it will likely always be a fallback). What we should move onto is a carrier and device type angnostic universal standard protocol over TCP / QUIC like XMPP or Matrix, with SMS as the backup.
When you get a phone you can get an phone system account and a telephone number already. Modern apps in the Google ecosystem should already recognize you are already signed in with Google and sync your contacts. Since almost everyone is already in the Google ecosystem, if Google supported it they could have extended their XMPP implementation in Hangouts to allow messaging directly via XMPP to those contacts and SMS for anyone not yet in the system (similar to how Signal did, Apple does, and Google does now with RCS). Unlike Apple, since its just XMPP, users can still add friends and be added by friends on other XMPP servers (ex. their ISPs, their own, or a third party). They could have supported or jumpstarted a new very simple open source alternative app for that portion for AOSP if the EU complained. Eventually Carriers could have supported passthroughs for those still on feature phones and other users of SMS to use the number@carrier accounts to hit XMPP users with generated SMS numbers for non-SMS users (pushed either by business necessity or part of a government / teleco org like GSMA staged removal of SMS and telephone numbers). It's all data at the end of the day.
Instead, they developed a whole new protocol to fluff the telecos and keep the now badly managed telephone number system even more necessary allowing spammers and allow the problems of legacy SMS to continue.
Apple, Google, and Samsung should all be shamed for not supporting fully open protocols and necessitating dependency on user harming stacks.
This sounds nice at a superficial level, but there's a lot of reliability and backwards compatibility issues being ignored. During natural disasters and emergency situations, internet and cellular data are the first to fail. It's not casual. For the phone and SMS (GSMA) protocols are sturdy enough that they can operate with very simple, low energy consuming and highly reliable machines. Internet data services on the other hand consume way more electricity (more expensive to have them operate with backup generators, for example) and are more delicate and prone to failure. They also need to be replaced more often. 100% of national emergencies systems run on phone and SMS tech, that could reliably operate for several decades with little maintenance that would cost billions to replace them with internet based system that were as reliable and durable. And then on top of it all, wired phones can even operate without electricity and connect with cellular terminals to contact other phones and cellphones. Only the tower needs to have power. There's just a lot banked of that reliability that most modern conveniences don't have.
In case of emergency it runs or it doesn't run. No matter if cellular or data.
Best were something like Briar's local wifi mesh standardized for emergency anyway.
I totally agree we can't simply drop SMS immediately, but what am I missing in supporting backwards compatibility (for example via my pseudo number solution, like how VOIP works) preventing us from moving forward during a stagged shutdown in the span of decades? MMS and RCS both would also fail under cellular data loss, and SMS itself hasn't always been available during major disasters. I'm not sure I buy the argument you can't have similarly low energy towers (even with net neutrality states, you can still cap all bandwidth per user), and a simpler tower that only does data should be far more reliable than a tower that provides multiple carrier services given the simplicity (and it's very rare to have towers that only do voice + SMS anymore).
Neither has running water or electricity. And SMS isn't actually the last fall-back (over here), that's FM radio which has better reach and crank-powered receivers start at like 10 bucks. Also there's a ton of generator-powered receivers around (called cars). Oh, dang, no, that's not actually the last fall-back that'd be megaphone trucks and cars practically all emergency service vehicles have some kind of PA system.
Solar storm killed the electrics of the new vehicles? There's a 60-year old Unimog still standing around getting moved once a year to keep it operational and I bet you'll find an analog megaphone in storage somewhere. It's astonishing how little stuff gets thrown away, we once stumbled across a stash of field telephones, half of them with swastikas ground off, the others still intact. Those require a crank and a copper cable to operate, nothing more. We used them to organise parking for a summer camp before the days of mobile flatrates.
The actual upside of plain ole GSM is that practically everyone carries around a receiver all the time, and there's reception literally everywhere. Better reach and better signal bandwidth than sirens, though of course nothing beats the oh fuck oh fuck hear it in your marrow aspect of sirens.
Catastrophe relief isn't an area where you ever want to have a single strategy because absolutely nothing is 100% fail-safe. In principle something like TETRA would be better than GSM but civilian phones don't speak it. (TETRA uses mesh networking, you can do direct handset to handset calls, drive around base stations in trucks to extend reach, etc)
Great points!
I don't know for certain. But one point to consider is that you have to qualify your “simply” statements with the fact that we are talking about millions of towers and hundreds of millions of repeaters over millions of square miles. While RCS works on top of the backbone that's already there and fallsback to SMS by design. So it might actually be simpler. The big up is that the server is on the carrier, not centralized, which makes it entirely different than what you are talking about and giving it more resilience.
There is so much nonsense being said about RCS, it will not fail under "cellular data loss".