Uh oh, Apple's Safari browser is trailing Froyo's performance

[Reluctant Adept]

Sometimes at 28 though I feel like I'm one of the oldest in here.

Oh no, I'm sure there are plenty of older folk around. I'm positively ancient at 42.

My point is unlike what Steve Jobs has you believing, you can have your cake and eat it too with flash player. Set it up to run on demand and it's like it's not even there.

While that makes a certain amount of sense for a seasoned user who already knows the web site in question, think about the experience from a fresh slate perspective. You visit a web site you've never been to before and there are three rectangular blocks with nothing in them. Which ones are ads? Which ones are meaningful content? Should you click on them or not? The situation with Flash in Froyo is an interesting workaround for today but it's not a desirable place to wind up in the long run.

Either Adobe will wake up and redesign Flash to be bandwidth, touch, and power-friendly or interactive content in HTML5 will sweep it into the history books. Apple is placing their bets to hurry things along to what they consider the most positive outcome for the industry, just like they've accelerated the demise of the floppy disk, parallel and serial ports, and the rise of Wi-Fi and Bluetooth.

 

[wreckshop]

N1 kills iphone at web page loading and javascript performance. Ive tested both and some pages the N1 loads much much faster than iphone. That being said, I'd much rather have iphones slower performance in exchange for its superior battery life.

 

[thatoneguy92088]

Actually, I'd have to tip my hat to the iPhone 4.

In every case the iPhone 4 loaded the brunt of the site first (the text/articles). While the Nexus did "complete" the page first, the iPhone loaded the important parts faster and then took a bit longer to load the ads/margins.

Point is, you can load the page in full faster on the Nexus One. You can browse the web (meaning actually read/use the page) faster on the iPhone.

And then when something new comes in and it moves the page and you click on a link you didn't want?

 

[Carniphage]

Surprised no one knowledgable has debunked this.

The Engadget test is not a test of javascript performance at all. It is purely a test of how fast each phone can load a web page.

The world is moving towards having web pages be very powerful, rich clients, similar to running a full application, but loaded via the browser. When Steve Jobs talks about HTML5, he also means Javascript to provide the interactivity.

That's a bit of a political statement.

The Android platform relies heavily on Java because it uses it for both Web applications and in its "native apps".

The iPhone model is not the same. Java is supported for Web apps, but remember the native application model on the iPhone is based on externally compiled native code. It does not use Java at all.

So any fair comparison of application performance should not look exclusively at Java.

The iPhone is a hybrid like a Prius. It has Java engine for light web stuff, and a native code engine for the heavy lifting. Not benchmarking that is like testing a Prius but disallowing the gasoline engine.

The Java-alone performance of Android 2.2 is currently better than iPhone. But it inaccurate to suggest this is a big deal for most users.

C.

 

[kevinof]

Totally and completely incorrect. Java <> Javascript. They are completely different technologies and used in completely different ways. Android uses a version of the Java VM to execute apps and other system code.

Javascript is ONLY used within the browser.

That's a bit of a political statement.

The Android platform relies heavily on Java because it uses it for both Web applications and in its "native apps".

The iPhone model is not the same. Java is supported for Web apps, but remember the native application model on the iPhone is based on externally compiled native code. It does not use Java at all.

So any fair comparison of application performance should not look exclusively at Java.

The iPhone is a hybrid like a Prius. It has Java engine for light web stuff, and a native code engine for the heavy lifting. Not benchmarking that is like testing a Prius but disallowing the gasoline engine.

The Java-alone performance of Android 2.2 is currently better than iPhone. But it inaccurate to suggest this is a big deal for most users.

C.

 

[Carniphage]

Totally and completely incorrect. Java <> Javascript. They are completely different technologies and used in completely different ways. Android uses a version of the Java VM to execute apps and other system code.

Javascript is ONLY used within the browser.

Yeah, you are right. Dumb error on my part.

But it is true that the Android conference conflated JavaSCRIPT performance with general performance. Or at least it was conflated in the minds of many.

C.

 

[kevinof]

I've no doubt they did (marketing is all about smoke and mirrors) but there actually is a (noticeable) increase in performance on my Desire when I went to Froyo. It was already fast on 2.1 but certainly has improved greatly in Froyo.

Yeah, you are right. Dumb error on my part.

But it is true that the Android conference conflated JavaSCRIPT performance with general performance. Or at least it was conflated in the minds of many.

C.

 

[dmelgar]

Yeah, you are right. Dumb error on my part.

But it is true that the Android conference conflated JavaSCRIPT performance with general performance. Or at least it was conflated in the minds of many.

C.

No, that are not 'conflated'.

Froyo, Android 2.2, makes HUGE improvements in speed in both Javascript performance and Java performance. Two totally separate items. Java is an interpreted language. They've added a Just In Time compiler (JIT) into Froyo which speeds up Java execution 2-5x. Thats an amazing improvement and speeds up pretty much every app running on Android.

The browser has also been improved with a new Javascript engine named V8 which improves performance 2-3x. V8 is written in C++. Presumably it doesn't benefit from the Java JIT speed improvement. V8 is open source so Apple could pick it up. V8 does a similar process to a JIT, compiling Javascript into native code during first execution to achieve its amazing speed.
http://code.google.com/p/v8/

 

[bbplayer5]

It was like 1/8 of a second faster! Who cares!

 

[Carniphage]

No, that are not 'conflated'.

Froyo, Android 2.2, makes HUGE improvements in speed in both Javascript performance and Java performance. Two totally separate items.
http://code.google.com/p/v8/

I am sure they have improved things a lot.

But the Android "running rings around" demo was only about Javascript.

Did they demonstrate the relative performance of the interpreted (JIT) Java with the native iPhone (ObjectiveC / Cocoa Touch) applications? Because that would be a useful comparison.

C.

 

[reservedegotist]

Not that I know of (JIT vs. Objective C/Cocoa). Still, an improvement on the Javascript performance will be welcome.

 

[dmorris01]

If its so cool, then why are you here on a iphone forum being a complete ass?

Just sayin'.

I didn't necessarily see him as being an ass, he was pointing out that if Apple wanted to have the 'whole' internet (as they claim they do) they could easily do it and make the Flash parts optional. Better yet, be able to turn it completely off if you want.

I'm not being an ass in making such a statement, am I? This is a discussion board.

 

[dmorris01]

I'll be honest, my nexus with froyo is insanely fast....theres just nothing fun to do with the speed. The apps are often hard to navigate, and they are rarely optimized for the display. I wish they had some decent games to show what the processor/os can do.
Apple is able to write code and design silicon that work together to provide a fast and enopjoyable experience. They were able to do this with the very first iPhone introduced in 2007, I have yet to see an android phone that can even scroll the icons smoothly when swiping screens. It's because they cant optimize the os for a single type of device....it will be very hard for google to ever compete with apple's overall user experience solely based on this.

Agreed, well stated.