News for hackers. Stuff that matters.
Re:Looking back at how Signal works by ComputerGuru on Friday June 05, 2020 @10:29PM (#23436299)
Backwards compatibility against a previous set of features isn't easy but it's certainly not impossible and graceful degradation is a thing.
Re:Looking back at how Signal works by Paul-ish on Friday June 05, 2020 @10:59PM (#23436430)
It isn't clear if graceful degradation is possible in a security app.
Re:Looking back at how Signal works by X6S1x6Okd1st on Saturday June 06, 2020 @12:43AM (#23436889)
There have been tons of problems with that concept and TLS
Re:Looking back at how Signal works by ComputerGuru on Friday June 05, 2020 @11:59PM (#23436713)
Cases where a client is completely broken are never the problem: users will be forced to switch to a different client. It sucks, but it's no worse than the current state of affairs. A security-mandated change in protocol/behavior would fairly fall under that category.
Re:Looking back at how Signal works by walrus01 on Friday June 05, 2020 @10:54PM (#23436407)
You can always define backwards compatibility that only goes to a certain lowest common denominator feature set, and no lower. For instance I have a number of httpd that support TLS1.2+ and specifically disallow SSLv3, TLS1.0 and TLS1.1. The population of browser user agents that don't understand TLS1.2 is infinitesimal at this point.