Live data from Hacker News

Apple OS X El Capitan

apple.com

131–140 of 146 posts

Re: Apple OS X El Capitan

#131
post #110
post #30

Earlier quoted context omitted.

This benefits game engines in general. I fail to see why it would favour only DX 12. Apple saved OpenGL from extinction when it was chosen as the iOS 3D API, as the API never had a foothold on game consoles. Apparently now they are done with it and with no plans on adopting Vulkan.

Do you have a source for "no plans on adopting Vulkan"? Apple is definitely on the Vullkan working group (logo is part of the image at the bottom of https://www.khronos.org/vulkan ).

It is a gut feeling from an old dog that has seen corporations change their stances in technology too many times.

Why adopt Vulkan given their investment in Metal, to the point that they have moved their accelerated 2D stack to Metal.

What is the point to rewrite it again for Vulkan?

Even if they adopt Vulkan, most likely would be a layer on top of Metal.

As for being in the group, Sony is also and they don't use OpenGL on their consoles.

Re: Apple OS X El Capitan

#132
post #85

Since no one already mentioned about it, here's an update to WebKit in Safari 9.0[1]. Just to pick a few: * Force Touch Trackpad Mouse Events * Content Blocking Safari Extensions (bytecode-compiled content blocklist for both iOS and OSX) * SFSafariViewController for iOS (as an alternative to UIWebKit and WKWebKit) * ECMAScript 6 support * CSS properties are now unprefixed (including flexbox!) I'm excited about the co…

I'm very curious about Content Blocking Safari Extensions as well.

The old mechanism, that is the non-standard "beforeload" DOM event + synchronous messaging via canLoad, had it's issues. But the new one is entirely different from the hooks Chrome and Firefox provide to intercept requests. So I'm certainly not a fan of it either, as this makes it even more elaborate to port extensions relying on content blocking to Safari, and massively reduces the amount of code that can be shared between the respective Chrome, Firefox and Safari extension.

But even more importantly, I fear that it will have some shortcomings when it comes to blocking requests depending on its context (e.g. frame hierarchy). At least I'd be very surprised if they do a better job than Google did with Chrome's declarativeWebRequest API.

If somebody knows any more details about that new API please let me know.

Re: Apple OS X El Capitan

#133

Earlier quoted context omitted.

>The incentive from Apple is very strong to install 10.10 And what incentive is that? Unless you need to latest Xcode to develop iOS apps, I don't see a reason to use Yosemite. I, for one, am happy to have stayed on 10.9 and will stick with it until El Capitan proves to be an "upgrade".

> And what incentive is that? For one thing, they won't sell you Mavericks anymore. If you didn't buy it while it was out, it is simply unavailable for you to purchase. They obviously could sell it to you, but they won't. It's a matter of intentional policy to make sure you don't buy in at any point earlier than the current one. (Take that in, BTW. That means someone at Apple made the judgment that it is literally mo…

Ah come on. Yosemite is not a stable, reliable OS? That's just hyperbole... Running 10.10.3 on half a dozen machines, mostly older ones, and it's stable, reliable and working just fine. MBPro's, Mac Mini's, MBAir, iMac.

Re: Apple OS X El Capitan

#134
post #114
post #94

Earlier quoted context omitted.

> The whole point of walled garden is locking people in. No, the whole point for me is keeping the weeds out and have a comprehensive experience where stuff works together well. It's 2015, there's nothing much to keep me locked in in any "walled garden". I can use another client for my IMAP emails (or export and re-import them in 20 ways), I can move my photos out of Photos app, I use Chrome to begin with but I could…

The very start of this thread gave an example of lock-in, being a custom API to get 3D performance to match other vendors. Judging by the size of the gaming industry, 3D performance is something people are interested in.

So Apple's grand plan is to get better gaming performance to lock people in?

First of all, gamers don't prefer Mac in the first place.

Second games are transient. You play them, and then you mostly forget them as new ones come out. You could jump to a new platform (that would be Windows) and buy new games and it wouldn't be a problem that your old Apple stuff doesn't work in Windows. So that's not much of a lock-in.

Third, if with the custom API they get "3D performance to match other vendors" then that's not a tempting proposition that can lock people in: it's "a match".

If, OTOH, they got performance many times that of other vendors with Metal, then that would be the GOOD kind of lock-in, the one where a platform is so much better that you use that.

Re: Apple OS X El Capitan

#135
post #15

Welp... it looks like OS X is about to discover tiling window management lite. ;)

On a serious note: if you need automatic window tiling in OS X, you should check out Amethyst [0] (available via homebrew cask). I mention this everywhere I can because I spent a lot of time searching for and testing various window management software for OS X myself, and Amethyst is only easy, fully functional alternative I've found that works out of the box. [0]: https://github.com/ianyh/Amethyst

Or get Hammerspoon[0] and roll your own window manager with Lua.

[0]: http://www.hammerspoon.org/

Re: Apple OS X El Capitan

#136
post #10

gotta love the adoption rate being percentage and not numbers. osx would show as a flat line...

Using numbers for adoption rate would make no sense at all. There would be no context for it. The comparison is how many installs are on the latest version. It would have to be X number out of Y number of installs in order to find out how many installs are running the same version. And then if you're doing X out of Y, you might as well normalize Y to 100, then you get a percent. If you're looking to see how many of y…

you develop for people. not percentages. the numbers means everything.

If 100% of mac users are on the latest version but that accounts for 1million users, and windows is split between 7 and 8. and heck, even XP... but the lowest is already 5 billion, then it is a no brainer that supporting one or two versions of windows makes more sense.

Re: Apple OS X El Capitan

#137
post #72

Earlier quoted context omitted.

Same, I'm still on Mavericks, and Yosemite still feels like "bleeding edge" to me.

I have to support an elementary school full of Macs, and Yosemite has been my ongoing nightmare. (Well, that and the bad joke known as OS X Server.) I won't update during the year for obvious reasons, but I'm still trying to decide whether or not I'm going to have to fragment my environment again (after just getting it all together with 10.9, finally). The incentive from Apple is very strong to install 10.10, but it…

> it just runs like utter garbage on all our older machines. interesting, it runs better on my dads 2011 11" MBA than on my 2013 13" rMBP, probably due to the retina display

Re: Apple OS X El Capitan

#138

Earlier quoted context omitted.

For a computer that costs thousands of dollars. Is that too much to ask that Apple make proper drivers for the one new model of discrete graphics cards they ship every year?

I’ve talked about this with one of the guys at Apple who writes graphics drivers. Basically vendors give them their Windows drivers and they change them to work against IOKit. No idea what my point is, besides, “it’s probably not that easy”. But they are on the Vulkan team and probably now have experience providing a similar API, so I’m hopeful the Macs get great adoption for Vulkan.

It's surely not easy, but that's still an important task. Hopefully El Capitan will also optimize these.

Re: Apple OS X El Capitan

#139
post #15

Earlier quoted context omitted.

On a serious note: if you need automatic window tiling in OS X, you should check out Amethyst [0] (available via homebrew cask). I mention this everywhere I can because I spent a lot of time searching for and testing various window management software for OS X myself, and Amethyst is only easy, fully functional alternative I've found that works out of the box. [0]: https://github.com/ianyh/Amethyst

Sorry, amethyst is nowhere near real tiling window managers. Hopefully with this we'll see real tiling window managers.

I haven't had a chance to try reverse engineering the APIs, but on the surface nothing has changed with the accessibility framework, which means we're not going to get much progress towards better window management.

Re: Apple OS X El Capitan

#140

Earlier quoted context omitted.

> And what incentive is that? For one thing, they won't sell you Mavericks anymore. If you didn't buy it while it was out, it is simply unavailable for you to purchase. They obviously could sell it to you, but they won't. It's a matter of intentional policy to make sure you don't buy in at any point earlier than the current one. (Take that in, BTW. That means someone at Apple made the judgment that it is literally mo…

Ah come on. Yosemite is not a stable, reliable OS? That's just hyperbole... Running 10.10.3 on half a dozen machines, mostly older ones, and it's stable, reliable and working just fine. MBPro's, Mac Mini's, MBAir, iMac.

Not sure if you're being sarcastic or not ...

Overall OS stability has gone down with Yosemite. I have to restart far more often than I should -- usually after coming out of sleep mode.

Post reply on HN