That's true, the time taken rebuilding the context depends a lot on the complexity of the context and yes the switching of resolutions and color depth plays a part - a big contribution comes from the monitor switching modes too.
The only difference now is that a lot more people bought Droids, N1s, and Galaxies, and now they're feeling the pain that all the first-gen owners felt when their platforms came along.
How are they 'feeling the pain'? Because of a rumor that a future OS that doesn't even have a release date may not support their devices?
In any case it's not the makers of Android but the handset manufacturers and carriers that handle the updates, so it isn't an Android problem but a HTC or Samsung or Motorola or Archos...etc... problem.
I wouldn't throw stones at Apple though. While I prefer Android to iOS, the fact remains that Apple has supported old hardware for upwards of three years - FAR longer than anybody has supported Android-based phones.
Just because it runs doesn't mean it will run well, there have been heaps of issues with the older 'supported' devices like the 3G and 3GS running the latest OS. There's little point supporting devices that won't run the system well anyway.
Im not disagreeing with you, personally i don't really have an opinion on java performance as I rarely use it and where I see it is in places that you can't isolate Java as being the cause of a performance bottleneck. But it occurs to me that the linked article seems to compare message passing in objective-c to function calls in java, which isn't really a fair comparison and the overhead should be pretty low since the dynamic lookup should only occur once per method after which it should be a hash lookup, which is quite fast.
I don't even see how this is possible. From a processor standpoint, a 1800MHz single core is *roughly* equivalent to a 900MHz dual core.
Probably going to get better battery life and responsiveness in multitasking since you'll spend less time context switching.
TFA is claiming that Google cares about the chip and not the relative performance? That doesn't make sense.
Getting the same performance for the same software out of different chips is very difficult even if they are capable of the same performance. Optimising for a particular instruction set? Optimising the scheduler and power management for multi-core CPUs?
I don't actually know, i'm just throwing out some guesses.
That model doesn't work for me. $600 is way to much to be spending on a mobile device with such a limited scope. I can get a good laptop for that kind of money.
So you bought it under the expectation that it would be able to upgrade for how many releases? Or for how long? I mean this doesn't affect the lifespan of the device or anything.
Absolutely - this really kinda ticks me off - I just paid $600 for a tablet that has been effectively dead-ended. I wonder how many of the original customers they screwed on the first round will feel like bending over to take another jab when they fork out another $600 to upgrade their OS. What a freaking waste.
But Android is open source, isn't that the beauty of it? Also you'll still upgrade to Gingerbread and who knows if there'll be a 2.4 so it's hardly 'dead-ended'. I mean Honeycomb doesn't even have a release date yet!
My point is that the performance requirement progression is pointless, useless, too fast, and stupid.
Why? We want more security, more stability, more responsiveness and more capability...these things don't come free. No-one's forcing you to upgrade so it's moving too fast for you then stick with what you've got.
Gingerbread is leaving behind almost every device already on the market.
That's a lot more forced progression than I'm used to seeing from any OS.
Well firstly it doesn't even have a release date yet and secondly of course at some point the OS will leave most existing devices behind, look at all the problems with the older iPhones running the latest OS, and the original 2G and early iPods don't get it at all. They want to start fresh to provide a consistent user experience and this is the way to accomplish that, seems reasonable since there are already some devices on the market that will support the OS even though the OS doesn't have a release date yet.
Google is trying to eliminate the fragmentation that everyone is complaining about, and also trying to reduce the crappy tablets out there giving Android a bad name.
And that's a fantastic thing to do, just like MS is doing with WP7, it provides a consistent experience as the developer intended.
It doesn't seem right. It's just out of fucking line that a cellphone OS would require a dual-core processor. Somebody needs to trim some bloat.
That 'bloat' you speak of is 'applications' and people want to run applications on their devices. People want to run multiple applications on their devices simultaneously so naturally a multi-core CPU is the ideal choice.
It's funny how some people are so ignorant they just assume that because system requirements go up it means the software is wasting clock cycles. Do you think added stability, security and features all come with no computational cost?
The problem with using the GPU is that every context switch requires a complete reinitialization of the GL context, even on a PC, alt tabbing into and from fullscreen games takes ages - it's fine when specific applications which requires the speed use it directly, but it's not when going from one activity to another gives you a loading screen.
Odd, I've never noticed such problems in Mac OS X, even though every window in the OS is composited by the GPU. Maybe it's not so much that GL is inherently flawed so much as that the Windows implementation was hobbled by Microsoft so that game developers would use DirectX instead.
Read the text he wrote - in fact it's the text you quoted - you're not comparing the same scenario.
No matter what MS likes to claim.NET is not portable in any meaningful way. The day I see MS made virtual machines for linux and mac is the day.NET becomes portable.
Yeah, cos linux and mac are the only other platforms. I suppose you haven't heard of the xbox360? or WP7? Get your head out of the sand.
"Following this, the team declared Sony's security to be EPIC FAIL!"
Is it really necessary for everybody to talk like complete dicks nowadays?
For elitist cockbags there is no other choice. But you'd think they'd tone it down a bit particularly when it's taken a number of YEARS to break said security.
If you want an idea of how little R&D spending is, you only need to look at medicine
Why would you look there and infer numbers for the tech industry - which is the context of this discussion - instead of just looking at the tech industry itself? In the tech sector the average spend is around 12-15% of total revenue, so as a percentage of their costs it is generally significantly higher.
It's really not possible to "invent" anything anymore as an individual. You'll inevitably violate some company's patent. You need your own war chest of patents to defend yourself.
What do you mean by 'as an individual'? Because I'm still seeing plenty of startups - that don't have large patent portfolios - succeeding.
Slashdot Mirror
That's true, the time taken rebuilding the context depends a lot on the complexity of the context and yes the switching of resolutions and color depth plays a part - a big contribution comes from the monitor switching modes too.
The only difference now is that a lot more people bought Droids, N1s, and Galaxies, and now they're feeling the pain that all the first-gen owners felt when their platforms came along.
How are they 'feeling the pain'? Because of a rumor that a future OS that doesn't even have a release date may not support their devices?
In any case it's not the makers of Android but the handset manufacturers and carriers that handle the updates, so it isn't an Android problem but a HTC or Samsung or Motorola or Archos...etc... problem.
I wouldn't throw stones at Apple though. While I prefer Android to iOS, the fact remains that Apple has supported old hardware for upwards of three years - FAR longer than anybody has supported Android-based phones.
Just because it runs doesn't mean it will run well, there have been heaps of issues with the older 'supported' devices like the 3G and 3GS running the latest OS. There's little point supporting devices that won't run the system well anyway.
That all depends on which vendor you've chosen to go with, not all vendors have included hardware lockdowns.
Gingerbread is already out.
Not for all devices yet.
http://www.javarants.com/2010/05/26/android-dalvik-vm-performance-is-a-threat-to-the-iphone/
Im not disagreeing with you, personally i don't really have an opinion on java performance as I rarely use it and where I see it is in places that you can't isolate Java as being the cause of a performance bottleneck. But it occurs to me that the linked article seems to compare message passing in objective-c to function calls in java, which isn't really a fair comparison and the overhead should be pretty low since the dynamic lookup should only occur once per method after which it should be a hash lookup, which is quite fast.
But it isn't a given either that it will run on phones.
Just like how iOS has tablet and phone versions.
The consensus is that this stupid rumor is false. It makes absolutely no sense to require any particular number of cores to run Android.
Or perhaps they are requiring a Cortex A9 instruction set and don't think the 'low power ... into cost sensitive devices' option would suffice?
Isn't 2.3 Gingerbread?
It says it "supports phone functionality (as speaker phone, via provided wired ear piece or Bluetooth earpieces.)"
Isn't that essentially what the GP wrote " cell phone capability ", as opposed to the Archos tablets?
I don't even see how this is possible. From a processor standpoint, a 1800MHz single core is *roughly* equivalent to a 900MHz dual core.
Probably going to get better battery life and responsiveness in multitasking since you'll spend less time context switching.
TFA is claiming that Google cares about the chip and not the relative performance? That doesn't make sense.
Getting the same performance for the same software out of different chips is very difficult even if they are capable of the same performance. Optimising for a particular instruction set? Optimising the scheduler and power management for multi-core CPUs?
I don't actually know, i'm just throwing out some guesses.
That model doesn't work for me. $600 is way to much to be spending on a mobile device with such a limited scope. I can get a good laptop for that kind of money.
So you bought it under the expectation that it would be able to upgrade for how many releases? Or for how long? I mean this doesn't affect the lifespan of the device or anything.
The last major iPad OS upgrade was free for all iPad owners and capable of running on current hardware, and it's likely the next one will be as well.
Gingerbread for Android tablet owners is a free upgrade too. In fact many Android tablets have gone through more free OS updates than the iPad.
Absolutely - this really kinda ticks me off - I just paid $600 for a tablet that has been effectively dead-ended. I wonder how many of the original customers they screwed on the first round will feel like bending over to take another jab when they fork out another $600 to upgrade their OS. What a freaking waste.
But Android is open source, isn't that the beauty of it? Also you'll still upgrade to Gingerbread and who knows if there'll be a 2.4 so it's hardly 'dead-ended'. I mean Honeycomb doesn't even have a release date yet!
My point is that the performance requirement progression is pointless, useless, too fast, and stupid.
Why? We want more security, more stability, more responsiveness and more capability...these things don't come free. No-one's forcing you to upgrade so it's moving too fast for you then stick with what you've got.
Gingerbread is leaving behind almost every device already on the market.
That's a lot more forced progression than I'm used to seeing from any OS.
Well firstly it doesn't even have a release date yet and secondly of course at some point the OS will leave most existing devices behind, look at all the problems with the older iPhones running the latest OS, and the original 2G and early iPods don't get it at all. They want to start fresh to provide a consistent user experience and this is the way to accomplish that, seems reasonable since there are already some devices on the market that will support the OS even though the OS doesn't have a release date yet.
Google is trying to eliminate the fragmentation that everyone is complaining about, and also trying to reduce the crappy tablets out there giving Android a bad name.
And that's a fantastic thing to do, just like MS is doing with WP7, it provides a consistent experience as the developer intended.
It doesn't seem right. It's just out of fucking line that a cellphone OS would require a dual-core processor. Somebody needs to trim some bloat.
That 'bloat' you speak of is 'applications' and people want to run applications on their devices. People want to run multiple applications on their devices simultaneously so naturally a multi-core CPU is the ideal choice.
It's funny how some people are so ignorant they just assume that because system requirements go up it means the software is wasting clock cycles. Do you think added stability, security and features all come with no computational cost?
Portable tends to mean platforms from more than one vendor.
No it doesn't, it has nothing to do with the vendor.
Odd, I've never noticed such problems in Mac OS X, even though every window in the OS is composited by the GPU. Maybe it's not so much that GL is inherently flawed so much as that the Windows implementation was hobbled by Microsoft so that game developers would use DirectX instead.
Read the text he wrote - in fact it's the text you quoted - you're not comparing the same scenario.
No matter what MS likes to claim .NET is not portable in any meaningful way. The day I see MS made virtual machines for linux and mac is the day .NET becomes portable.
Yeah, cos linux and mac are the only other platforms. I suppose you haven't heard of the xbox360? or WP7? Get your head out of the sand.
On Mac, it's called CoreGraphics. On Linux, it's called QPainter or Cairo. On Windows, there are so many APIs that I don't feel like naming them all.
lol...open graphics library
"Following this, the team declared Sony's security to be EPIC FAIL!"
Is it really necessary for everybody to talk like complete dicks nowadays?
For elitist cockbags there is no other choice. But you'd think they'd tone it down a bit particularly when it's taken a number of YEARS to break said security.
If you want an idea of how little R&D spending is, you only need to look at medicine
Why would you look there and infer numbers for the tech industry - which is the context of this discussion - instead of just looking at the tech industry itself? In the tech sector the average spend is around 12-15% of total revenue, so as a percentage of their costs it is generally significantly higher.
It's really not possible to "invent" anything anymore as an individual. You'll inevitably violate some company's patent. You need your own war chest of patents to defend yourself.
What do you mean by 'as an individual'? Because I'm still seeing plenty of startups - that don't have large patent portfolios - succeeding.
Why innovate when you can litigate?
What specific innovation is this legal action stifling?
Though this is still almost a year old