Safari 4 Released, Claimed "30 Times Faster Than IE7"
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."
Anyone know if this is a new engine or just Squirrelfish renamed?
Looks like Safari might be the first Acid 3 browser to the market. Opera's version 10 is Acid 3 compliant, but it's still in Alpha testing.
I noted this feature in Opera 10. The results shown in the demos were rather impressive. The web pages had more of a print-layout look to them without the classic trick of relying on images to cover all the content. This has the potential to completely change the look of the web for the better.
I'm still trying to figure out how being able to use Canvas as a style to apply to web elements is useful, but the idea definitely sounds cool. I suppose one could always set a fixed web page background as a canvas, then make it look like they're on an acid trip as they scroll. :-P
I'm downloading the beta now. If it lives up to the hype that Apple is giving it, it will be an amazing piece of software.
Javascript + Nintendo DSi = DSiCade
Here's the actual claims from Apple's website:
"Using the new Nitro Engine, for example, Safari executes JavaScript up to 30 times faster than Internet Explorer 7 and more than 3 times faster than Firefox 3 based on performance in leading industry benchmark tests: iBench and SunSpider.
In addition to superior JavaScript performance, Safari offers top-flight HTML performance -- the best on any platform -- loading pages 3 times faster than Internet Explorer 7 and almost 3 times faster than Firefox 3."
I'm not too familiar with either of these benchmarking programs, so I can't really pick at the results too much, but the actual claim is 'up to 30 times faster' which means that for some function it's 30 times faster, but for most it's probably not at that level of magnitude. It seems as though some of this important information was lost in the game of telephone that is internet news.
Also, I'm more interested in how it stacks up against Firefox, Opera, and Chrome. Comparing it to IE7 is a little bit like Ford comparing their new car to a horse and cart. No offense meant to the browser, but from every chart I've seen it's the bottom of the barrel in terms of speed.
http://www.apple.com/safari/features.html - Safari 4 introduces the Nitro JavaScript engine, an advanced bytecode JavaScript engine that makes web browsing even faster. In fact, Safari 4 executes JavaScript up to 6 times faster than Internet Explorer 8 and up to 4 times faster than Firefox 3.1.
So if Safari has this great performance, how can the FF figure out how Safari does it?
By heading over to WebKit.org and downloading the open source rendering engine it uses?
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.
For nearly all other applications, the P2 was equal to or faster than the PPC. But Apple hyped their systems based on that one single meaningless-for-most-people benchmark. (As opposed to AMD's speed rating system which for the Athlon XPs was based on a suite of benchmarks and their average comparison to a similarly clocked P4, which was typically pretty accurate.)
Here, how is Apple magically eliminating network latency and providing infinite network bandwidth with browser changes? For nearly all users, the network is the bottleneck.
retrorocket.o not found, launch anyway?
I've tried pointing that out before, but you're probably wasting your breath. The tin-foil hat crowd here at slashdot seems to think that Apple is keeping all the juiciest enhancements for themselves. I know it's not true because I run Safari on my macs and have run some webkit browsers like midori on my linux machines, they're about as fast, certainly faster than firefox. I'd use midori as my full time browser, but it's not as full featured as firefox and is unstable (or was last version I downloaded, like 0.0.21 or so).
Gentlemen! You can't fight in here, this is the war room!
Link for Safari 3:
http://support.apple.com/downloads/#internet
Clearly labeled as "Looking for Safari 3? Download here" at bottom of the Safari 4 download page.
The latest safari nightly (r41176) compiles and runs just fine on my stock Ubuntu Hardy box. The only pain I encountered was the libsoup 2.25 library dependency which I had to pull down and compile myself instead of using the older library supplied from the Hardy repository.
Not only is WebKit open-source, it also seems like Apple has gone to great length to ensure that this piece of software is portable to other operating systems. The key is to actually do the "wget, autogen.sh, make, make install" steps yourself. It's really not that hard.
As a developer working on WebKit, this is completely wrong and more than a little insulting.
The versions of WebKit included with Safari releases are built directly from the public tree. There is no secret version of WebKit that Apple fixes bugs in for Safari releases before eventually landing the changes in the WebKit tree. The WebKit tree is Apple's official WebKit tree, and is where all of Apple's development on WebKit for Mac OS X and Windows takes place.
For sake of reference http://trac.webkit.org/browser/releases/Apple/Safari%204%20Public%20Beta contains the exact source code of WebKit that was built and released as Safari 4 Public Beta earlier today. There are no secret changes in the version of WebKit that Apple shipped. The changes are all there in the open for the world to see.