4 ms·
The problem is that it doesn't do a super good job of explaining why non-nerds -- or arguably, nerds! -- should care about "convergence" in this regard, and how
by chipotle_coyote 5y ago
The problem is that it doesn't do a super good job of explaining why non-nerds -- or arguably, nerds! -- should care about "convergence" in this regard, and how it improves over existing solutions. I mean, I'm over here in Apple-land, walled garden etc. etc. etc., but there are a lot of applications that I have that exist in both macOS and iOS versions. This is true for most of Apple's own apps, and actually true for a lot of third-party apps, too: in a practical sense, I've had "convergence" for many years now.
Sure, these aren't usually literally the same pieces of software running on the Mac, the iPhone, and the iPad. But from what I can tell, that's not literally true on Purism, either, because the software is going to have to be recompiled for different CPU architectures. So now we're actually a lot closer to what Apple's been doing the last few years with Catalyst (and kind of slouching toward a grander strategy with SwiftUI). And again, not to put too fine a point on it, but I don't actually care in practice how much code is shared between the versions of my prose editor I'm running on the desktop and on the phone, given that they function the same way and edit the same files. (And share them all completely transparently, because they're using iCloud. Walled garden etc. etc., but this is a real selling point.)
Purism's real argument is free-as-in-freedom software. And that's a good argument. But I think they need to get a lot better at reaching the "casual but concerned" audience, and that means getting better at writing press releases and product announcements. Speaking as a professional writer, the PureOS 10 announcement page is really, really bad. You may not think it's fair to judge the quality of the product based on that, and it isn't, but it's still going to happen. (Judging from the comments here they also need to get better at, well, shipping hardware, which is an even more serious problem, granted.)