Mac OS X Lion's Safari is Snappier and More Stable Thanks to WebKit2

[linux2mac]

I agree with a browser having the least UI is the best, that being said, Safari is quite UI-free as well. Firefox seems to be quite crowded compared to them both.

I agree. I have no idea what he was talking about Safari not being so UI free. I was looking at the top of this web page, and all I see is an address bar, search bar, and few bookmarks under the address bar. I suppose I could remove the search bar and bookmarks and then I would just be left with an address bar. LOL

 

[zipster]

I hope so. Safari is garbage on my Mac. Chrome is faster, snappier, lighter, etc.

 

[LouieSamman]

I hope so. Safari is garbage on my Mac. Chrome is faster, snappier, lighter, etc.

Being that Safari is the native browser for OS X it's sad that we have to revert to another browser.

Reason for this is that google is a software company and Apple is a software and hardware company so it's harder for Apple to compete with a company that does only software.

I feel that Apple will really pull forward soon and show a better browser than chrome. Better mailing service with MobileMe than Google Chrome.

They will add great UI along with speed. Apple just need time.

Making MobileMe free is a big step that will also introduce a revamp of MobileMe.

Safari 4.1 on OS X Lion has got a lot of people happy over how snappy the browser is.

We will have to just wait and see what they will introduce.

 

[Aisys]

Safari 5.1 preview does not entirely solve Safari's huge memory usage issues, but it does improve the situation, due to the new process model.

I submit a lot of bug reports to Webkit Bugzilla, including some high profile memory issues, and I can note the following from my experience there:

- The Webkit team *does* consider memory usage in Safari pretty awful and something they need to seriously address.
- They *do* have open bug reports for several issues that are causing the browser memory usage to grow unbounded.
- Fixes for those memory leaks tend to arrive very slowly, unfortunately... but they're actively worked on nevertheless.
- You can track the state of memory leaks in Webkit at the following url, which they've recently developed to help track these down. There are a number of improvements being worked on to further improve the tool's ability to assist in that work. Ironically using the tool leaks a lot of memory. http://build.webkit.org/LeaksViewer/

 

[bobeff09]

Safari Web Content process

Why is this process burning CPU cycles (55% per Activity Monitor). It spins up the fan and the whole Safari process becomes jerky, unstable.

Platform
mac book pro
OS 10.6.8
Safari 5.1

 

[AidenShaw]

That's why Macs are so cool to begin with, when a program crashes, only that program crashes.

Welcome to the world of Windows NT 3.1, as well as all of the UNIXen and other proprietary operating systems.

Come to think of it, virtually every OS except 16-bit Windows and Mac OS (classic) had memory protection as a foundation.

Now, only a portion of the program will crash? GO Apple.

Go Apple indeed - copy features from Internet Explorer 7 and other applications.

 

[munkery]

Application sandboxing and mandatory access controls (MAC) in OS X are the same thing. More specifically, applications are sandboxed in OS X via MAC. Mac OS X uses the TrustedBSD MAC framework, which is a derivative of MAC from SE-Linux. This system is mandatory because it does not rely on inherited permissions. Both mandatorily exposed services (mDNSresponder, netbios...) and many client-side apps (Safari, Preview, TextEdit…) are sandboxed in Lion.

Windows does not have MAC. The system that provides sandboxing in Windows, called mandatory integrity controls (MIC), does not function like MAC because it is not actually mandatory. MIC functions based on inherited permissions so it is essentially an extension of DAC (see #1). If UAC is set with less restrictions or disabled in Windows, then MIC has less restrictions or is disabled.

http://www.exploit-db.com/download_pdf/16031 -> article about Mac sandbox.

http://msdn.microsoft.com/en-us/library/bb648648(v=VS.85).aspx -> MS documentation about MIC.

https://media.blackhat.com/bh-eu-11/Tom_Keetch/BlackHat_EU_2011_Keetch_Sandboxes-Slides.pdf -> researchers have found the MIC in IE is not a security boundary.

 

[DadaGo]

Safari and Lion

I have used Snow Leopard and Safari since it has come out and have been very pleased with the stability and responsiveness.

Lion and Safari appear to be less stable with Safari frequently shutting down and asking to re-open. In addition other software from iPhoto to iTunes to the app store and Mail suffer from degraded stability in Lion.

This is a differential statement to what I got used to with Snow Leopard and Leopard before that. Snow Leopard was truly a faster operating system and more stable.

I have a 13" MacBook Pro 2010 version with all the bells and whistles and have not made any hardware changes before switching to Lion.

 

[SamIchi]

I'd would try safari but it doesn't have a middle click to close tab option. Why don't they do this?!?!

Firefox has it, chrome has it.

The cool thing about Safari 5 is the slick back and forward swipe animations. Chrome has lost the swipe to go back/forward in Lion, hopefully they'll have an update that address that.

 

[rockinred]

To be honest, I love safari on the mac.

But, the version that's shipped in Mac OS X Lion seems to have more bugs than Windows Vista. It's a very buggy release.

- Not snappier at all
- There's a simple question here: Sometimes the webpages refresh when toggled; similar to iOS's safari. Did anybody notice this themselves?
It would be sad if Apple made it more like iOS to save the onboard memory.

Agreed... nothing snappy about it. I like the whole swipe thing. Can't stand how buggy Safari is. All these years I have never had this kind of slow problems and issues with Safari. I deleted the cache.db and still no help. I am about to give up on it.

 

[KingCrimson]

That's why Macs are so cool to begin with, when a program crashes, only that program crashes. Now, only a portion of the program will crash? GO

Google Chrome on Windows has had that functionality for 2 years.

 

[AidenShaw]

Google Chrome on Windows has had that functionality for 2 years.

IE also....

 

[vitzr]

I don't have a problem with Safari.

It's been stable, smooth, just not very fast, not a slug by any means, but just not as fast as Firefox or Chrome.

That said, I must admit that Chrome has become so fast, and reliable that it's far and away my favorite. Also the kinds of extensions they offer now are really great.

 

[munkery]

Google Chrome on Windows has had that functionality for 2 years.

IE also....

IE only for web process but not for plugins. Plugins run in web process.

Firefox only for plugins. Rendering and scripting engines are not isolated from main process.

Chrome for web process and plugins.

Safari same as Chrome (prior to 5.1 only for plugins).

So, all popular browsers have included some implementation of process separation for several years.

Also, take into account that sandbox implementations vary across OSs

Application sandboxing and mandatory access controls (MAC) in OS X are the same thing. More specifically, applications are sandboxed in OS X via MAC. Mac OS X uses the TrustedBSD MAC framework, which is a derivative of MAC from SE-Linux. This system is mandatory because it does not rely on inherited permissions. Both mandatorily exposed services (mDNSresponder, netbios...) and many client-side apps (Safari, Preview, TextEdit…) are sandboxed in Lion.

Windows does not have MAC. The system that provides sandboxing in Windows, called mandatory integrity controls (MIC), does not function like MAC because it is not actually mandatory. MIC functions based on inherited permissions so it is essentially an extension of DAC (see #1). If UAC is set with less restrictions or disabled in Windows, then MIC has less restrictions or is disabled.

http://www.exploit-db.com/download_pdf/16031 -> article about Mac sandbox.

http://msdn.microsoft.com/en-us/library/bb648648(v=VS.85).aspx -> MS documentation about MIC.

https://media.blackhat.com/bh-eu-11/Tom_Keetch/BlackHat_EU_2011_Keetch_Sandboxes-Slides.pdf -> researchers have found the MIC in IE is not a security boundary.