Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

saiyan

macrumors member
Original poster
Aug 5, 2010
31
2
Okay.

So after I updated iOS to 4.3 on my iPad, I tested Safari, Perfect Browser and Atomic Browser using Sunspider's Javascript benchmark. Out of the three browsers, only Safari showed significant improvement (about 33xx ms). Perfect Browser and Atomic Browser both still showed slower score (about 84xx ms). :(

Does anyone know if we need to wait for the developers of Perfect and Atomic browsers to update their apps to utilize the new Nitro Javscript engine?

Thanks..
 

MikhailT

macrumors 601
Nov 12, 2007
4,582
1,325
This is strange, you’d think they all use the same core as Mobile Safari.

It is possible Apple separated the APIs but I somehow doubt that.
 

JohnDG

macrumors member
Sep 1, 2010
62
0
Okay.

So after I updated iOS to 4.3 on my iPad, I tested Safari, Perfect Browser and Atomic Browser using Sunspider's Javascript benchmark. Out of the three browsers, only Safari showed significant improvement (about 33xx ms). Perfect Browser and Atomic Browser both still showed slower score (about 84xx ms). :(

Does anyone know if we need to wait for the developers of Perfect and Atomic browsers to update their apps to utilize the new Nitro Javscript engine?

Thanks..

The assumption I'm making is that they need to be "recompiled" (is this term showing my age?) and released to include the new webkit.

jdg
 

JulianL

macrumors 68000
Feb 2, 2010
1,657
654
London, UK
The assumption I'm making is that they need to be "recompiled" (is this term showing my age?) and released to include the new webkit.
I'm afraid not. I thought the same thing but I just emailed the developer of the browser I use, iCab. He's in Germany and I emailed him at 9pm his time so, due to time, he was a bit slow in replying - he took a whole hour and 9 minutes to get back to me (this guy's email support is just amazing)!

Anyway, he gave me permission to post his reply so here it is. My questions are preceded by ">" (and, in my defence, I was trying to enumerate all logical possibilities with my thought that a flag might need to be set, I never thought that would make much sense) ...

>I just upgraded my iPad 1 to iOS 4.3 and tried running the SunSpider
>0.9.1 benchmark in iCab Mobile. It ran slightly slower on iOS 4.3 vs
>4.2 (about 8600 vs about 8350 for iOS 4.2) whereas Safari showed the
>speedup that all the articles are talking about (about 3300 under iOS
>4.3). I hear on Macrumors forum that Atomic also doesn't see any
>speedup with 4.3

Yes, according to Apple, this is a "known issue". So this seems to be
either a bug in iOS 4.3, or maybe this is intentional and Apple will test
this exclusively in their own Apps first, before making this new
engine available for other Apps as well.

Currently, the new JS engine seems to be Safari only. No other App
can use it at the moment. We have to wait for Apple to fix this.

>I'd assumed that the new Nitro JavaScript engine was in the WebKit and
>so all third part browsers would automatically see the benefits. Is

Yes, this was my assumtion too. But it looks like this is not the case.

>this not the case, or is some code change required in iCab to set a
>flag or something to say that you want to enable nitro? Is there hope

There's no such flag. And it would not make any sense to have such a
flag. All Apps would always want to use the faster and better engine.

>that iCab will get to the same performance as Safari or have Apple
>deliberately done something to hurt third party browsers?

Apple has to fix this. Otherwise other Apps can not get the same
performance because there are no other engines allowed than the ones
which are built-in in the iOS and which the iOS provides through the
public API to the developers.

>I expect you've had this question a hundred times this week so please
>let me know if it is OK to post any answer you give on forums - it
>might save you from a few extra emails asking this question.

Yes, this is OK.

In short: Only Apple can fix this. "Nitro" seems to be exclusively used
in Safari at the moment. No other App is able to use it. I guess (and hope)
that one of the next iOS releases will fix this.


This guy is a hardcore iOS developer, blogging about detailed iOS programming techniques (http://www.icab.de/blog/), so I have great confidence in information that he gives. I think that we're at the mercy of Apple on this one.

- Julian
 

HXGuy

macrumors 68000
Mar 25, 2010
1,679
0
Damn that sucks! I hate Safari because of their stupid "tabs" and no option for full screen or touch gestures...all reasons of why I use Atomic exclusively.
 

JulianL

macrumors 68000
Feb 2, 2010
1,657
654
London, UK
Don't know anything about JavaScript but wifi has been improved tremendously with this update and iCab is just flying along with email and YouTube.

Safari is no faster than iCab on my iPad.
Yeah, it's easy to carried away and forget that JavaScript is just one part of the overall picture. I'm also finding that browsing simple pages seems quicker in 4.3. Also, this JavaScript thing might be temporary, they might open it up to everyone in a future release.

What worries me more is the whole concept of offering a faster item for Safari and not everyone. The nightmare is that this spreads and more and more tweaks that improve rendering speed get reserved for Safari only and the publicly accessible WebKit behind the official public APIs stagnates. I really hope that doesn't happen. Going back to Safari after iCab would feel a huge step backwards to me now unless Apple make some significant improvements to the Safari UI.

- Julian
 

spammerhamster

macrumors 6502
Feb 5, 2010
288
1
Strange enough... The other browser does benefit from a better memory management as they can load pages with a lot of images.
 

saiyan

macrumors member
Original poster
Aug 5, 2010
31
2
Atomic Web Browser is back in the App Store. I guess the developer has removed whatever private API calls they were using before.

Anyway. I updated to the latest version (5.5) and ran Sunspider benchmark again. It is still slow as before.

Perfect Browser and Mercury Browser are both slow too.

I wish Apple would simply implement tabs in its mobile Safari browser (at least on iPad) so I don't need to even bother with these other 3rd party browsers even though they are supposed to be using WebKit and Javascript engine built-in to iOS.
 

Maaz

macrumors 6502a
Mar 13, 2010
645
27
My web pages on atomic are loading the EXACT same as safari as of last night.
 

fertilized-egg

macrumors 68020
Dec 18, 2009
2,109
57
I'm afraid not. I thought the same thing but I just emailed the developer of the browser I use, iCab.

Wow thanks a lot for that JuianL. So like many of us even the iCab dev had assumed the new Javascript engine would automatically be enabled for other 3rd party browsers but that's not the case. That blows :(
 

anthonymoody

macrumors 68040
Aug 8, 2002
3,063
1,153
Wow thanks a lot for that JuianL. So like many of us even the iCab dev had assumed the new Javascript engine would automatically be enabled for other 3rd party browsers but that's not the case. That blows :(

Hopefully this is not just temporary but very short lived as well. I would hate for Apple to place its dev community at such a distinct disadvantage over the long haul.
 

kas23

macrumors 603
Oct 28, 2007
5,629
288
It doesn't matter for, I'll still stick with iCab Mobile. Nitro or no Nitro, it runs circles around Mobile Safari (which is really a shoddy browser)
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.