[dieselville]

This is older news, but I just want to get your thoughts on this...

Yesterday, we reported that RIM�s strategy to add Android compatibility to its QNX platform might prove to be a blessing in disguise for the company. But according to Dan Frommer , there might be a twist in the tale. RIM�s plan to run Android apps through Android �app player� may face performance issues as they run through an emulator. According to Jim Balsillie, Co-CEO of BlackBerry, "You're just not going to get things like gaming and multimedia, you're not going to get the speed going through a VM interface". Thus it is expected that heavy Android games and apps like Netflix will work poorly or might not work at all.

Do you think we going to be disappointed with the speed on certain games/apps?

[papped]

Android uses a VM to run android apps....

[therapyreject174]

It all depends how they implement it. If they give it full hardware access there should be very little performance degradation. Since the Android Player is a RIM developed program only available for their devices, I believe the integration will be very good.

The reason something like Windows Virtual PC gets crummy gaming and multimedia performance is because it doesn't get full access to some hardware. If I recall, it emulates a 4MB video card with SVGA resolution.

[dieselville]

Right on, thanks guys..

[stillzmatik]

I would rather the PB not have the Android Market if they cannot make the transition seemless. RIM should create incentives for developers to code for the QNX platform instead of creating a bridge from Android.

Posted from my CrackBerry at wapforums.crackberry.com

[mandony]

Nobody has seen the final release version including the developers. Why speculate about Andy-royd? Why be 'concerned?'

As the expression goes: 'que sera sera', 'what will be, will be'

[mithrazor]

It'll be alright. It'll run the same as on Android. It's using the same VM Android is using. It's just been ported over to QNX.

[samab]

According to Jim Balsillie, Co-CEO of BlackBerry, "You're just not going to get things like gaming and multimedia, you're not going to get the speed going through a VM interface". Thus it is expected that heavy Android games and apps like Netflix will work poorly or might not work at all.

Do you think we going to be disappointed with the speed on certain games/apps?

Google is also telling Android developers the SAME thing. Google tells you not to use the dalvik vm engine for the heavy lifting for gaming, use the Android NDK instead.

[lawguyman]

This is what Balsillie actually said (whether it can be trusted at all is another matter):

First of all, what we announced is Gingerbread. This is not Honeycomb. I don't know what the number of Honeycomb apps is, but it's not very many. Whereas Gingerbread they've got lots of them. You've got the volume of the handset apps, so if you're looking for the tonnage of apps, or some kind of long tail stuff, you've got it.

At the end of the day, people are going to want performance. You're just not going to get things like gaming and multimedia, you're not going to get the speed going through a VM interface. If you want content, or Flash type stuff, or you're looking at AIR-type, evolving web-type assets, that's what you're going to do.

There's no compromise here. You've got the tonnage of apps. And you've got the performance. Do I think the tonnage is overplayed? Yes.

But if you think it's about having a couple hundred thousand apps, there you go.

Do we believe it's about super high performance? Yes. Do we believe it's about full web fidelity? Yes. These are concepts that were really relegated as not technically possible, which we're doing here. This is a no compromise environment.

If you want to work on Android, great. Do we think people will want to migrate web assets? Yes. Do we think they're going to want super high performance native assets with the SDK? Absolutely. You think they're going to want to use their Flash based stuff for an offline Flash/AIR type environment? Yes.

I'm just not interested in these sort of religious application tonnage issues. I really think we put that issue to bed. And if you think the whole world's going to want to develop for Gingerbread, fine. Do I think that's going to happen? Then why is there a different environment for a tablet? And you know about the performance issues and you know about the app volume issues, cause it's tough. And that's why QNX matters.

That's why people are saying, Is this stuff going to go more in the browser and HTML 5 and more native? These are going to be strong trends. But if you want these app players for different VMs -- and don't forget we have 25,000 BlackBerry 6 apps. So, at the end of the day, we believe this is going to be about performance. It's going to be about enterprise greatness. Things like multi-threaded capability, symmetric multiprocessing. We believe it's about an uncompromised web. We believe it's about enterprise security. True multitasking, not with suspension -- and that matters because you're going to want to run these things in the background.

But I'm out of the religious war on tonnage, which I'm delighted.

... (Lots of repetition.)

I think it's very important to understand that this idea of "no compromise" matters. And this idea that you can pick whichever one you want.



Read more: RIM CEO Basically Admits That Android Support Is There To Check The "Lots Of Apps" Box

[BuzzStarField]

This is what Balsillie actually said (whether it can be trusted at all is another matter):

And whether or not you can understand what Balsillie is saying is a whole 'nother matter. What we really need from these guys is a well-written white paper about where RIM is going with their various "players" and "native" SDKs. And more importantly why there are going in that particular direction. People like me who jumped from Java to the AIR SDK are shaking their heads in dismay. It would be nice to know how to plan our migration path from the "old" handset to the announced QNX devices. Right now it's a confused mess and brain dumps like this one are no help whatsoever!

[joshua_sx1]

If Android apps and games will run smoothly and flawlessly in PlayBook, the BlackBerry PlayBook Apps World will be doomed... the less developers will be interested in developing applications and games that run in QNX OS... and the future BlackBerry phones that will have QNX OS are destined to be doomed too...

So, the 2012 prediction will be true then... but only to RIM...

[BuzzStarField]

If Android apps and games will run smoothly and flawlessly in PlayBook, the BlackBerry PlayBook Apps World will be doomed... the less developers will be interested in developing applications and games that run in QNX OS... and the future BlackBerry phones that will have QNX OS are destined to be doomed too...

So, the 2012 prediction will be true then... but only to RIM...

Well, the "announcements" would indicate that Android apps would have to be "repackaged", signed by the RIM signing authority, approved by RIM and posted in App World.

But the prospect of a flood of Android apps swamping App World and eclipsing "native" apps has definitely put a chill on developing AIR apps or improving current offerings in App World. The NDK has not even been released yet but if Android apps are going to run "smoothly and flawlessly" what are the prospects of new C+ apps competing on a level playing field? I have no idea if I am typical, but I'm playing a wait and see game right now trying to figure out how I con complete my app given that AIR apps can't access the magnetometer at this time. Lack of information from RIM makes it impossible to plan my next move. It's ridiculous.

[samab]

The NDK has not even been released yet but if Android apps are going to run "smoothly and flawlessly" what are the prospects of new C+ apps competing on a level playing field?

It depends on what kind of apps you are writing. You are not going to write a native c++ fart app --- just because you know c++. You can write the fart app in BASIC and you still won't see any performance difference.

The ONLY Android apps that the Playbook can run are the PURE dalvik apps. Google even tells you that you shouldn't write a PURE dalvik app if you want to write a performance app. You have to write all the heavy lifting part of the code with Android's native NDK --- which the Playbook will not be able to run.

So you are competing with dinky little Android apps written purely within the dalvik vm engine --- only your Flash/AIR Playbook app would look "prettier" (if you have the artistic skills).

[BuzzStarField]

It depends on what kind of apps you are writing.

...
So you are competing with dinky little Android apps written purely within the dalvik vm engine --- only your Flash/AIR Playbook app would look "prettier" (if you have the artistic skills).

You probably did not intend to insult me but your're being a bit dismissive here. My AIR app is not a dinky little app and could be (with a total re-write) ported to Java/Android and would have access to the magnetometer. In fact, a lot of augmented reality apps and other sophisticated apps are ideally suited to either AIR or Android. Unless you need direct access to graphics or you want to write drivers, NDK is overkill. Nevertheless, my only other option for my dinky app would be to port my app to the yet to be released NDK. Again it would require a complete re-write.

My point is that because RIM has not given developers a road map, we have no idea if or when AIR will ever be fully competitive with Android apps with respect to performance and capabilities.

And not to put too fine a point on it, I have a feeling that RIM engineers are burning the midnight oil to ensure that the player cannot be hacked to support the Android Market. Could this be the cause of delays in releasing the player? In any event, I wish they could or would have spent some time to build the missing API's into the AIR SDK.

[pkcable]

Jim may not even remember SAYING it in a few months!

[BuzzStarField]

Jim may not even remember SAYING it in a few months!

He's probably forgotten already!

[joshua_sx1]

Now we need an application for "Alzheimer's disease"...