5 ms·

It's sad to see the decline in values that had me in love with joining this industry. This is now a financialized company, keeping things consistent, and shippi

It's sad to see the decline in values that had me in love with joining this industry. This is now a financialized company, keeping things consistent, and shipping a credit card, is much more important than bold HCI moves, or letting any % of profit dip in service of searching for a wider revenue basin. I worked in Android at Google, and I think the year-long shipping schedules are inimical to SW exploration. Communication is poor enough at BigCo that you'll have a feature approved in August, in flight by October, duplicated and shipped by a completely different team in January, and have to throw something together to be ship-worthy in April. And if you don't, you didn't get any work done this year. It's not really "you need to launch to promo" as "if you don't launch, welcome to the performance shitlist. hope you don't become part of the 5% we'll sacrifice to Wall Street this year!" Then, the schedule offset between the top and bottom means by the time you're done shipping the barely working version, its time to do whatever thing the VPs were kicking around the last few months by themselves and fell in love with.

mike_hearn 2y ago
Perhaps, and that's an accurate description of Google nowadays, but I think I disagree in the case of Apple. We should be fair to Cook and the rest of the Apple team here: Vision Pro is nothing but bold HCI moves, and the very opposite of what you get from a yearly shipping schedule. It's probably been in development for over a decade. There are people who joined Apple, spent years working on Vision Pro, were promoted and then left before it was even announced! In terms of execution quality and ethos it reflects the Apple of the Jobs era in every way. The problem is simply that breaking into new markets is hard, especially if the market is a price/quality point that nobody else is attempting. Maybe there's just no demand in the end. That's R&D for you. My argument is really the opposite: Apple doesn't need moves as bold as Vision Pro, or cars or robots. There's plenty of room for innovation within existing product segments, price points and form factors. If I ran the Core OS group at Apple I'd have a dozen research ideas to experiment with by next week, and I'm just one guy who thinks about operating systems more than is healthy. If you look across the span of their existing businesses there must be hundreds of ways they could capture market share in innovative ways.
- ghaff 2y ago
  I'm not sure what major user-visible innovation in desktop operating systems really would look like these days. It presumably wouldn't be something that fundamentally changed the user experience. I have some suspicion that convergence of mobile and desktop/laptop will happen with time but a lot of users would probably hate it, Microsoft already sort of whiffed on it, and convergence probably isn't really in Apple's business interests anyway so long as they can sell someone both a laptop and a tablet.
  - mike_hearn 2y ago
    Why not experiment with fundamentally changed user experiences? The WIMP paradigm goes back to the 1980s and much of it is rotting away, getting less and less well matched to the modern world yet without being rethought. Random examples of eminently fixable HCI issues that bother me every day: 1. The clipboard is invisible, historyless and lost on reboot unless you install third party utilities. Yet clipboard operations are a key part of any professional user's day. 2. The Downloads folder is basically an indirect cache for remote URLs because apps can't open them directly, but I have to manually manage it by cleaning things out every so often. Also, if I open the same link twice it'll just download the same file twice which is all wrong. Possible fix: the OS should have a kind of "belt" of recently used "things", whether those are objects copied from apps, files accessed over HTTP or apps. They should persist across reboots. The OS should act as if these don't use any disk space even though they do, and delete old items automatically as the disk fills up. 3. Window management can be upgraded significantly. Multi-screen is sort of hacked on, the way Cmd-Tab works for instance when >1 screen gets involved is quite unintuitive. People need layouts that they can switch instantly without animation delays, and the only way macOS offers to do that is by using Cmd-Tab. Safari gets this right: switching and closing tabs is instant, why can't the WM work do that? 4. Browsers are barely integrated and so the UI is very confusing. My wife sometimes accidentally saves PDFs of Google Sheets because that's what happens if you click File > Save in Safari. Why are there two menu bars and what's the difference, eh, life is short and computer history is long. Don't worry honey, I'll do it. Also see: starting Zoom meetings from iCalendar (opens a web page which opens an app). Possible fix: Any action triggered by the keyboard should not animate (or only does it once or twice to teach the user what's happening and then stop). Make Safari/browser the default desktop experience, allowing apps to launch into tabs with a better/more keyboard-able splitting implementation. Get rid of the default keyboard shortcuts for things like save/print in Safari (do you really need keyboard shortcuts for such rare operations), and let web apps take that over. If the web app doesn't, try to auto-wire them to web app UI using Apple Intelligence or DOM matching heuristics. And so on. This is like a fraction of what could be fixed. Those are small things. Competing harder with Windows would not just involve innovation, it'd also involve stuff that doesn't come naturally to Apple e.g. maybe a UI mode that's more like Windows to make switchers more comfortable, or Macs that are sold with Office out of the box, or integrated support for running some apps using Wine or built in support for Parallels-style Coherence virtualization.

mike_hearn 2y ago

Perhaps, and that's an accurate description of Google nowadays, but I think I disagree in the case of Apple. We should be fair to Cook and the rest of the Apple team here: Vision Pro is nothing but bold HCI moves, and the very opposite of what you get from a yearly shipping schedule. It's probably been in development for over a decade. There are people who joined Apple, spent years working on Vision Pro, were promoted and then left before it was even announced! In terms of execution quality and ethos it reflects the Apple of the Jobs era in every way. The problem is simply that breaking into new markets is hard, especially if the market is a price/quality point that nobody else is attempting. Maybe there's just no demand in the end. That's R&D for you. My argument is really the opposite: Apple doesn't need moves as bold as Vision Pro, or cars or robots. There's plenty of room for innovation within existing product segments, price points and form factors. If I ran the Core OS group at Apple I'd have a dozen research ideas to experiment with by next week, and I'm just one guy who thinks about operating systems more than is healthy. If you look across the span of their existing businesses there must be hundreds of ways they could capture market share in innovative ways.

ghaff 2y ago
I'm not sure what major user-visible innovation in desktop operating systems really would look like these days. It presumably wouldn't be something that fundamentally changed the user experience. I have some suspicion that convergence of mobile and desktop/laptop will happen with time but a lot of users would probably hate it, Microsoft already sort of whiffed on it, and convergence probably isn't really in Apple's business interests anyway so long as they can sell someone both a laptop and a tablet.
- mike_hearn 2y ago
  Why not experiment with fundamentally changed user experiences? The WIMP paradigm goes back to the 1980s and much of it is rotting away, getting less and less well matched to the modern world yet without being rethought. Random examples of eminently fixable HCI issues that bother me every day: 1. The clipboard is invisible, historyless and lost on reboot unless you install third party utilities. Yet clipboard operations are a key part of any professional user's day. 2. The Downloads folder is basically an indirect cache for remote URLs because apps can't open them directly, but I have to manually manage it by cleaning things out every so often. Also, if I open the same link twice it'll just download the same file twice which is all wrong. Possible fix: the OS should have a kind of "belt" of recently used "things", whether those are objects copied from apps, files accessed over HTTP or apps. They should persist across reboots. The OS should act as if these don't use any disk space even though they do, and delete old items automatically as the disk fills up. 3. Window management can be upgraded significantly. Multi-screen is sort of hacked on, the way Cmd-Tab works for instance when >1 screen gets involved is quite unintuitive. People need layouts that they can switch instantly without animation delays, and the only way macOS offers to do that is by using Cmd-Tab. Safari gets this right: switching and closing tabs is instant, why can't the WM work do that? 4. Browsers are barely integrated and so the UI is very confusing. My wife sometimes accidentally saves PDFs of Google Sheets because that's what happens if you click File > Save in Safari. Why are there two menu bars and what's the difference, eh, life is short and computer history is long. Don't worry honey, I'll do it. Also see: starting Zoom meetings from iCalendar (opens a web page which opens an app). Possible fix: Any action triggered by the keyboard should not animate (or only does it once or twice to teach the user what's happening and then stop). Make Safari/browser the default desktop experience, allowing apps to launch into tabs with a better/more keyboard-able splitting implementation. Get rid of the default keyboard shortcuts for things like save/print in Safari (do you really need keyboard shortcuts for such rare operations), and let web apps take that over. If the web app doesn't, try to auto-wire them to web app UI using Apple Intelligence or DOM matching heuristics. And so on. This is like a fraction of what could be fixed. Those are small things. Competing harder with Windows would not just involve innovation, it'd also involve stuff that doesn't come naturally to Apple e.g. maybe a UI mode that's more like Windows to make switchers more comfortable, or Macs that are sold with Office out of the box, or integrated support for running some apps using Wine or built in support for Parallels-style Coherence virtualization.