Safari 4 Released, Claimed "30 Times Faster Than IE7" 465
CNETNate writes "Apple has released the beta version of Safari 4 for Mac and PC, with claims that its Nitro rendering engine is '30 times faster than IE7,' and three times faster than Firefox 3. Other new features include 'Top Sites,' which shows users the most frequently visited Web pages, 'Full History Search' for searching through not only the URLs and titles of visited pages, but also the complete text within the page itself — something Opera has been doing for a while."
More Fun Demos (Score:5, Interesting)
Falling Leaves Animation: http://webkit.org/blog-files/leaves/index.html [webkit.org]
Bouncing Box Animation: http://webkit.org/blog-files/bounce.html [webkit.org]
Rotate and Fade Animation: http://webkit.org/blog-files/pulse.html [webkit.org]
CSS Recipes for Effects: http://developer.apple.com/safari/articles/webcontent/cssrecipes.html [apple.com]
CSS Gradients: http://developer.apple.com/safari/library/documentation/InternetWeb/Conceptual/SafariVisualEffectsProgGuide/Gradients/chapter_2_section_1.html#//apple_ref/doc/uid/TP40008032-CH7-SW11 [apple.com]
Video tag (requires Quicktime): http://webkit.org/blog/140/html5-media-support/ [webkit.org]
CSS Gradients: http://webkit.org/blog/175/introducing-css-gradients/ [webkit.org]
Background Shaped Clipping: http://webkit.org/blog/164/background-clip-text/ [webkit.org]
Local Database Example: http://webkit.org/misc/DatabaseExample.html [webkit.org]
Impressions (Score:5, Interesting)
- Scrolling this /. page is extremely slow in safari.
- The tabs in the window's title bar is just plain annoying and feels really out of place.
- Just like Google's Chrome this browser also doesn't blend in well with MS Windows UI. It's feels alien to the other programs.
Re:Classic Apple performance claim inflation (Score:3, Interesting)
Apple loves to put in meaningless benchmarks with no real-world meaning to hype their products.
For example, the "3 times faster than a Pentium II" claims back in some of the older PowerPC days - this was true for a single Photoshop operation that at that point had Altivec optimizations on PPC but was running straight scalar code (no MMX) on a P2.
If your going to spout blind FUD, do your homework. Altivec didn't even exist on PowerPC's back when the Pentium II was around. The current PowerPC CPUs were the PowerPC 604 and the newcomer was the G3.
Altivec didn't arrive until the G4 and by then the Pentium III was out and selling.
At the same clock rate, the PowerPC really was quite a bit faster. Not by rediculous "3x" margins but it really was quite a bit faster. The PowerPC is also a much cleaner and well-thought-out architecture. Anybody that still does any ASM can definitely vouch for this.
Just because IBM/Moto/Apple didn't have the R&D dollars to polish a turd until it hit 4GHz doesn't mean the PowerPC sucked. It was and still is an awesome architecture.
For nearly all other applications, the P2 was equal to or faster than the PPC.
No, it wasn't. I ran several real-world benchmarks as I owned an Apple B&W G3 tower and a Pentium II at the time.
Are they as fast as Apple claimed? Hell no. Were they genuinely faster? Yes.
For nearly all users, the network is the bottleneck.
Now that is very accurate. For what most people use a computer for, a single-board 1.6Ghz atom machine with a GMA950 is more than they'll ever need for web browsing, e-mail, playing youtube videos and running Word. A faster machine doesn't make you type faster or make web pages load faster.
Safari's improvements though are very welcome as they free CPU cycles for more useful things. A more efficient app is always a welcome change.
To read more about safari's legendary stability... (Score:1, Interesting)
Just type data:%80,; into your address bar!
Re:Notes on New Features (Score:1, Interesting)
If we're going to be throwing custom fonts onto the intarwebs, then CSS3 sure as fuck better be getting a revision to allow widths of boxes to be set to "big enough to fit its contents at whatever font and zoom level the user decides to use" because you know 90% of the browsers are going to either not support it or the user will have overridden it because they are offended by anything less generic than Arial. And no, measuring in em is bullshit, especially when you don't know how proportional the font is going to be and some user decides to make their username |||||||||||| just to piss you off.
While I'm whining about the fact that CSS can't cope with user-generated data, I'll continue with float, which needs an "all-the-way-left/right" setting so that when the block cannot fit on the current line, rather than moving it down juuuuust enough to make it fit (leading to a shitty stair-steps look whenever someone writes two lines worth of content where everyone else wrote one... or if I force everything to have a specific height, then shitty overlapping content), it clears all the current floats and starts at the left/right margin.
Re:Notes on New Features (Score:2, Interesting)
Scriptable plug-ins is listed as a feature.
This is now my default Windows browser (Score:4, Interesting)
For anyone that has both a Mac and PC, one of the minor frustrations you face is constantly having to remember to use different keyboard shortcuts when you move back and forth. Safari on the PC was an option for me for this reason alone. Sadly, the Mac-look, odd window handling, terrible font rendering and random long pauses (something to do with advertisements I think) made it an option only - I had to keep Chrome and FF around for some sites.
No longer. Safari 4 is now my default Windows browser. And not just because of the keystrokes, it's faster than any of the other (always up-to-date) browsers on this machine, renders everything perfectly (Chrome still has serious problems here), the font problems are gone (now Chrome is the one that looks bad), the random pauses are missing, etc.
So basically Safari now does everything any of the other browsers does, plus more, plus its faster, AND it has the same keystrokes.
Still not perfect though: I'm still trying to get the font sizes right (the readable text above is fine, but this editor has HUGE text) and I want to remove the Chrome-like tools menus (I like real menu bars, thanks), and there's some oddity when scrolling long pages. But nevertheless the problems are less than those in Chrome and the speed of FF in comparison makes me willing to overlook them.
Maury
Re:Classic Apple performance claim inflation (Score:3, Interesting)
'Full History Search' (Score:2, Interesting)
In other news... (Score:4, Interesting)
Re:How does firefox maintain competitive advantage (Score:4, Interesting)
During the early days of Safari 3.0, I was in charge of making sure my companies product was compatible with Safari.
I have built WebKit from their xcode project. I have submitted bugs. And I know that sometimes the fix arrives in Safari months before WebKit.
I have much respect for that development team, but to say that Apple (as close-lipped and proprietary as they are) isn't holding anything back is just naive.
Safari still a memory pig? Crash protection? (Score:4, Interesting)
I switched to Firefox for two reasons. One is that Safari is a major memory hog. It can use like 3x the memory as Firefox for the same thing. (And I'm talking about fresh starts. I know all about how VMs can swap unused pages to disk.)
The other missing thing from Safari was something as basic as session saving and crash protection. You have to buy Saft for that. With Firefox, it's free.
I wonder if Apple has done anything about these issues.
Re:In other news... (Score:3, Interesting)
I gave safari a run on my web app, which uses a lot of clientside scripting and has been designed to "work" on FF, IE7, chrome. I did not optimize anything for any browser, it was just a test to make sure I would make mac users happy. I was amazed by performances, really. The JS runtime is way better than anything else I've tested, and even beats chrome which is also really good. More importantly, it seems almost immune from memory leaks, compared to ff3 which needs a restart when approaching 1GB.
Re:Classic Apple performance claim inflation (Score:3, Interesting)
Re:To read more about safari's legendary stability (Score:2, Interesting)
Re:Removes existing installations (Score:2, Interesting)
I've put together some docs on creating a standalone Safari 4: http://vasi.dyndns.org:3128/svn/SafariBeta/ [dyndns.org] . It runs exactly like the beta, only it does not replace the system's WebKit library and does not replace Safari 3. So you can continue to use the old Safari, and your applications will not use the new WebKit (and potentially break because of it).
It also has a tiny patch allowing the use of an auxiliary preferences file. This lets you disable incompatible InputManager hacks for Safari 4 only, while Safari 3 will still use them.
A couple of responses to miscellaneous comments in this thread:
- Following the "Looking for Safari 3" link will just end up overwriting Safari 4 and its WebKit. Congratulations, you've reverted back to where you were before! But you still can't run Safari 3 and Safari 4 side-by-side.
- Seriously, WebKit nightlies include WebKit inside their bundle, and other apps therefore don't see the new WebKit. This works. It is a standard technique on OS X, do not be surprised.
- It is indeed a good thing that the Safari 4 beta upgrades the WebKit library, things really need to be tested before Safari 4 final is released and millions of users have their apps break. However, there's no excuse for Apple not providing a standalone Safari 3 so we can test in both versions. Also, this public beta is quite different from the last semi-private developer release of Safari 4--Apple really should have provided the beta as a dev release first, so folks could fix their WebKit-using apps.