Transparency is the killer for HTML 5 and Flash. Because transparency is such an easy thing to do in Flash and the results are visually pleasing, you see it everywhere. The power required to add the shadowBlur and render the underlying objects while animating the top layer is where the whole thing goes south.
yup, exactly. One good reason why if I do something in flash, or anything else for that matter, I try to avoid alpha tweens or anything involving heavy blurred shadows etc. I'll use any trick in the book besides.
I certainly don't want to excuse the crimes of adobe's mac flash plugin, but, one needs to keep things in perspective about the fact that regardless of the tool, certain things are very cpu intensives regardless of the delivery method, and there's going to be reams of developers who will code badly to take out your browser.
Is there going to be a huge outcry over html5 if that gets out of control and is annoying the crap out of users with flying ads too? (remember animated gifs....)
Transparency is the killer for HTML 5 and Flash. Because transparency is such an easy thing to do in Flash and the results are visually pleasing, you see it everywhere. The power required to add the shadowBlur and render the underlying objects while animating the top layer is where the whole thing goes south.
Why, do you suppose that is?
Isn't ActionScript very similar to JavaScript?
Aren't both interpreted?
Don't the latest versions of each use some sort of JIT interpretation/compilation?
Is this (Transparency) just an isolated instance where Flash excels?
Could it be due to the relative maturity of implementations of Flash vs HTML5?
Do you see any reason why HTML5 cannot eventually equal or exceed the Flash implementation?
On a forum such as this, your "facts" are as good as mine!
Which is why I provided some code that can be independently verified. In science that is what we do. You publish your findings and provide a means for the experiment to be duplicated by others. That way there is no personal bias clouding the issues.
Provide some Flash code that will spike my browser. I will analyze it and try to determine what is causing the problem.
Do you see any reason why HTML5 cannot eventually equal or exceed the Flash implementation?
HTML 5 is already faster than Flash. The primary reason is that HTML is being processed by features of the browser itself and Flash is processed by a plugin. You can compare that to the way php compiled into Apache httpd is faster than php loaded as an extension.
HTML 5, JS, CSS are like the children of Safari and the Flash plugin is the hired help who speaks some other language. A translation is involved for Flash to communicate with Safari which creates additional overhead and delay.
Mmm... It does significantly degrade performance. With only that one window open It spiked CPU at 91%...
His example is complete bollocks. It's a Canvas example, that is only small feature of HTML5, that is newer and less developed, that doesn't WebGL incorporated yet, and that he designed to push your CPU just for the sake of pushing your CPU. But, from that example we are supposed to say that the entirety of HTML5 is bad and Flash is the future?
Which is why I provided some code that can be independently verified. In science that is what we do. You publish your findings and provide a means for the experiment to be duplicated by others. That way there is no personal bias clouding the issues.
Provide some Flash code that will spike my browser. I will analyze it and try to determine what is causing the problem.
As you should understand that is a bit more difficult. What spikes my browser are applications/movies written in Flash. I do not have source code to these. Further, while I can overload the system to cause Flash to spike, that would provide little enlightenment.
I could disable Click2Flash, and likely, after a few hours the CPU would spike, Safari would hang, or the plugin task would crash. I would post the details to Apple (and I do), But I wouldn't have anything tangible to provide to you other than to copy/paste the reported status/details and email them to you.
I can say this, After installing Click2Flash and preventing automatic loading of Flash.(with an occasional override for a specific video)-- i have very few problems with Flash. Before, i could expect several hangs or crashes per week.
My goal is not to Debug Flash, rather to utilize the web for my own purposes, When it is convenient, or when I want to back up an assertion, I'll make an extra*effort (send the crash info to Apple or document CPU usage).
HTML 5 is already faster than Flash. The primary reason is that HTML is being processed by features of the browser itself and Flash is processed by a plugin. You can compare that to the way php compiled into Apache httpd is faster than php loaded as an extension.
HTML 5, JS, CSS are like the children of Safari and the Flash plugin is the hired help who speaks some other language. A translation is involved for Flash to communicate with Safari which creates additional overhead and delay.
His example is complete bollocks. It's a Canvas example, that is only small feature of HTML5, that is newer and less developed, that doesn't WebGL incorporated yet, and that he designed to push your CPU just for the sake of pushing your CPU. But, from that example we are supposed to say that the entirety of HTML5 is bad and Flash is the future?
Mmmm... from what he said I didn't infer that HTML5 was bad and Flash was good!
I find your comment on WebGL interesting, though.
I assume that you mean that in the future, HTML5 will support a form of OpenGL and, that, in implementing that, it will take advantage of GPU hardware when available.
Flash could certainly do this too, if the OS permits access to the APIs and hardware.
His example is complete bollocks. It's a Canvas example, that is only small feature of HTML5, that is newer and less developed, that doesn't WebGL incorporated yet, and that he designed to push your CPU just for the sake of pushing your CPU. But, from that example we are supposed to say that the entirety of HTML5 is bad and Flash is the future?
You are just trying to stir up some controversy. I said nothing of the sort and you know it. If you would point to specific errors in my posts I will try to either clarify them or stand corrected and admit to the error.
I said nothing of the sort and you know it. If you would point to specific errors in my posts I will try to either clarify them or stand corrected and admit to the error.
You're right, I woke up and read your comment and test Canvas before I had my coffee. The point is, Canvas should never be used to show how poor HTML5 performance is as it's a single aspect of HTML5. It's still being built so until there are development tools for it without HW Acceleration in place and sites are actively using it for more than a demo it's not a good measure of anything. it's irrelevant. The single most important aspect of Flash on the web today is video and this is considerably more costly on Flash than it is in HTML5. That's relevant.
You're right, I woke up and read your comment and test Canvas before I had my coffee. The point is, Canvas should never be used to show how poor HTML5 performance is as it's a single aspect of HTML5. It's still being built so until there are development tools for it without HW Acceleration in place and sites are actively using it for more than a demo it's not a good measure of anything. it's irrelevant. The single most important aspect of Flash on the web today is video and this is considerably more costly on Flash than it is in HTML5. That's relevant.
My one and only point is that stupid programmers can ruin even the best delivery platform. And when that happens the outcry is that the platform is at fault. That is why Apple tries to keep such a tight grip on what is and is not allowed on the iPhone. It is because if a rogue programmer compromises the device the press will be pointing their finger at Steve Jobs.
My one and only point is that stupid programmers can ruin even the best delivery platform. And when that happens the outcry is that the platform is at fault. That is why Apple tries to keep such a tight grip on what is and is not allowed on the iPhone. It is because if a rogue programmer compromises the device the press will be pointing their finger at Steve Jobs.
I wholeheartedly agree with that.
Adding to that, I have concerns about WebGL. Poor or malicious programmers using HTML and JS to get more direct access to HW. I can see a potential for this being abused.
I wonder if Apple's latest stunt will spur a wave of browser-blocking antics across the internet:
First, proponents of open standards will block Safari browsers from their websites in protest of Apple. Apple haters will start doing the same thing and block Safari browsers from their sites.
With fewer than 5% of web users bothering with Safari there's no need to block it, it's lack of appeal has already taken care of that.
His example is complete bollocks. It's a Canvas example, that is only small feature of HTML5, that is newer and less developed, that doesn't WebGL incorporated yet, and that he designed to push your CPU just for the sake of pushing your CPU. But, from that example we are supposed to say that the entirety of HTML5 is bad and Flash is the future?
I think it's Safari; Safari 4.0.5 on Win7 uses up 25% of my CPU (Core i5-750) on that example, and doesn't even run it smoothly (motion blur set to 1000), whereas Chrome 6.0 dev only uses about 10% and runs as smooth as silk. (motion blur also at 1000).
I just tried the latest nightly of Webkit, and it doesn't do any better. I could try it on my Mac, but I know it will run like crap (it's an old C2D).
It may be a newer part of HTML5, but if Apple needs GPU acceleration to do canvas, they're doing it wrong.
The single most important aspect of Flash on the web today is video and this is considerably more costly on Flash than it is in HTML5. That's relevant.
Costly in what way?
Because if by "cost" you mean cost, remember that h264 isn't free.
Sure, Steve likes to point out that it currently has no *license* fees, and when spoken exactly like that he's technically correct. But as the patent holder of h264, Apple expects vendors writing h264 codecs to pay them other fees which are not called "license" fees.
Then in less than five years they start adding in license fees too -- classic bait-n-switch -- so there's a cost now and no way to know how much dependency on h264 is going to cost us in the future.
All we know is that Apple sits on the receiving end of that cash stream.
I think it's Safari; Safari 4.0.5 on Win7 uses up 25% of my CPU (Core i5-750) on that example, and doesn't even run it smoothly (motion blur set to 1000), whereas Chrome 6.0 dev only uses about 10% and runs as smooth as silk. (motion blur also at 1000).
I just tried the latest nightly of Webkit, and it doesn't do any better. I could try it on my Mac, but I know it will run like crap (it's an old C2D).
It may be a newer part of HTML5, but if Apple needs GPU acceleration to do canvas, they're doing it wrong.
The word is that Safari 5 adds the much needed HW acceleration to Safari on Windows. I have my doubts as to how well it will help with Canvas, but it's a start.
Quote:
Originally Posted by RationalTroll
Costly in what way?
Because if by "cost" you mean cost, remember that h264 isn't free.
Sure, Steve likes to point out that it currently has no *license* fees, and when spoken exactly like that he's technically correct. But as the patent holder of h264, Apple expects vendors writing h264 codecs to pay them other fees which are not called "license" fees.
Then in less than five years they start adding in license fees too -- classic bait-n-switch -- so there's a cost now and no way to know how much dependency on h264 is going to cost us in the future.
All we know is that Apple sits on the receiving end of that cash stream.
Awesome¡ A comment about processing overhead regarding Flash v. HTML5 for video streaming and you jump into a comment about H.264 which is not germane to the topic and is an optional codec for each.
The word is that Safari 5 adds the much needed HW acceleration to Safari on Windows. I have my doubts as to how well it will help with Canvas, but it's a start.
Awesome¡ A comment about processing overhead regarding Flash v. HTML5 for video streaming and you jump into a comment about H.264 which is not germane to the topic and is an optional codec for each.
Yet you are a guy who bought up a codec, that both HTML5 and Flash can use, among other codecs, despite it not being relevant to my comment. It must be hard to be rational when you reading comprehension is such an issue for you.
Comments
Transparency is the killer for HTML 5 and Flash. Because transparency is such an easy thing to do in Flash and the results are visually pleasing, you see it everywhere. The power required to add the shadowBlur and render the underlying objects while animating the top layer is where the whole thing goes south.
yup, exactly. One good reason why if I do something in flash, or anything else for that matter, I try to avoid alpha tweens or anything involving heavy blurred shadows etc. I'll use any trick in the book besides.
I certainly don't want to excuse the crimes of adobe's mac flash plugin, but, one needs to keep things in perspective about the fact that regardless of the tool, certain things are very cpu intensives regardless of the delivery method, and there's going to be reams of developers who will code badly to take out your browser.
Is there going to be a huge outcry over html5 if that gets out of control and is annoying the crap out of users with flying ads too? (remember animated gifs....)
Transparency is the killer for HTML 5 and Flash. Because transparency is such an easy thing to do in Flash and the results are visually pleasing, you see it everywhere. The power required to add the shadowBlur and render the underlying objects while animating the top layer is where the whole thing goes south.
Why, do you suppose that is?
Isn't ActionScript very similar to JavaScript?
Aren't both interpreted?
Don't the latest versions of each use some sort of JIT interpretation/compilation?
Is this (Transparency) just an isolated instance where Flash excels?
Could it be due to the relative maturity of implementations of Flash vs HTML5?
Do you see any reason why HTML5 cannot eventually equal or exceed the Flash implementation?
.
On a forum such as this, your "facts" are as good as mine!
Which is why I provided some code that can be independently verified. In science that is what we do. You publish your findings and provide a means for the experiment to be duplicated by others. That way there is no personal bias clouding the issues.
Provide some Flash code that will spike my browser. I will analyze it and try to determine what is causing the problem.
Is this (Transparency) just an isolated instance where Flash excels?
Flash doesn't excel. They both suffer from the same issues when it comes to rendering animated transparency. More power is the only solution.
I'll use any trick in the book besides.
Love that book.
Do you see any reason why HTML5 cannot eventually equal or exceed the Flash implementation?
HTML 5 is already faster than Flash. The primary reason is that HTML is being processed by features of the browser itself and Flash is processed by a plugin. You can compare that to the way php compiled into Apache httpd is faster than php loaded as an extension.
HTML 5, JS, CSS are like the children of Safari and the Flash plugin is the hired help who speaks some other language. A translation is involved for Flash to communicate with Safari which creates additional overhead and delay.
Mmm... It does significantly degrade performance. With only that one window open It spiked CPU at 91%...
His example is complete bollocks. It's a Canvas example, that is only small feature of HTML5, that is newer and less developed, that doesn't WebGL incorporated yet, and that he designed to push your CPU just for the sake of pushing your CPU. But, from that example we are supposed to say that the entirety of HTML5 is bad and Flash is the future?
Which is why I provided some code that can be independently verified. In science that is what we do. You publish your findings and provide a means for the experiment to be duplicated by others. That way there is no personal bias clouding the issues.
Provide some Flash code that will spike my browser. I will analyze it and try to determine what is causing the problem.
As you should understand that is a bit more difficult. What spikes my browser are applications/movies written in Flash. I do not have source code to these. Further, while I can overload the system to cause Flash to spike, that would provide little enlightenment.
I could disable Click2Flash, and likely, after a few hours the CPU would spike, Safari would hang, or the plugin task would crash. I would post the details to Apple (and I do), But I wouldn't have anything tangible to provide to you other than to copy/paste the reported status/details and email them to you.
I can say this, After installing Click2Flash and preventing automatic loading of Flash.(with an occasional override for a specific video)-- i have very few problems with Flash. Before, i could expect several hangs or crashes per week.
My goal is not to Debug Flash, rather to utilize the web for my own purposes, When it is convenient, or when I want to back up an assertion, I'll make an extra*effort (send the crash info to Apple or document CPU usage).
.
HTML 5 is already faster than Flash. The primary reason is that HTML is being processed by features of the browser itself and Flash is processed by a plugin. You can compare that to the way php compiled into Apache httpd is faster than php loaded as an extension.
HTML 5, JS, CSS are like the children of Safari and the Flash plugin is the hired help who speaks some other language. A translation is involved for Flash to communicate with Safari which creates additional overhead and delay.
Very cogent answers!
Thanks!
.
His example is complete bollocks. It's a Canvas example, that is only small feature of HTML5, that is newer and less developed, that doesn't WebGL incorporated yet, and that he designed to push your CPU just for the sake of pushing your CPU. But, from that example we are supposed to say that the entirety of HTML5 is bad and Flash is the future?
Mmmm... from what he said I didn't infer that HTML5 was bad and Flash was good!
I find your comment on WebGL interesting, though.
I assume that you mean that in the future, HTML5 will support a form of OpenGL and, that, in implementing that, it will take advantage of GPU hardware when available.
Flash could certainly do this too, if the OS permits access to the APIs and hardware.
.
His example is complete bollocks. It's a Canvas example, that is only small feature of HTML5, that is newer and less developed, that doesn't WebGL incorporated yet, and that he designed to push your CPU just for the sake of pushing your CPU. But, from that example we are supposed to say that the entirety of HTML5 is bad and Flash is the future?
You are just trying to stir up some controversy. I said nothing of the sort and you know it. If you would point to specific errors in my posts I will try to either clarify them or stand corrected and admit to the error.
I said nothing of the sort and you know it. If you would point to specific errors in my posts I will try to either clarify them or stand corrected and admit to the error.
You're right, I woke up and read your comment and test Canvas before I had my coffee. The point is, Canvas should never be used to show how poor HTML5 performance is as it's a single aspect of HTML5. It's still being built so until there are development tools for it without HW Acceleration in place and sites are actively using it for more than a demo it's not a good measure of anything. it's irrelevant. The single most important aspect of Flash on the web today is video and this is considerably more costly on Flash than it is in HTML5. That's relevant.
You're right, I woke up and read your comment and test Canvas before I had my coffee. The point is, Canvas should never be used to show how poor HTML5 performance is as it's a single aspect of HTML5. It's still being built so until there are development tools for it without HW Acceleration in place and sites are actively using it for more than a demo it's not a good measure of anything. it's irrelevant. The single most important aspect of Flash on the web today is video and this is considerably more costly on Flash than it is in HTML5. That's relevant.
My one and only point is that stupid programmers can ruin even the best delivery platform. And when that happens the outcry is that the platform is at fault. That is why Apple tries to keep such a tight grip on what is and is not allowed on the iPhone. It is because if a rogue programmer compromises the device the press will be pointing their finger at Steve Jobs.
My one and only point is that stupid programmers can ruin even the best delivery platform. And when that happens the outcry is that the platform is at fault. That is why Apple tries to keep such a tight grip on what is and is not allowed on the iPhone. It is because if a rogue programmer compromises the device the press will be pointing their finger at Steve Jobs.
I wholeheartedly agree with that.
Adding to that, I have concerns about WebGL. Poor or malicious programmers using HTML and JS to get more direct access to HW. I can see a potential for this being abused.
I wonder if Apple's latest stunt will spur a wave of browser-blocking antics across the internet:
First, proponents of open standards will block Safari browsers from their websites in protest of Apple. Apple haters will start doing the same thing and block Safari browsers from their sites.
With fewer than 5% of web users bothering with Safari there's no need to block it, it's lack of appeal has already taken care of that.
His example is complete bollocks. It's a Canvas example, that is only small feature of HTML5, that is newer and less developed, that doesn't WebGL incorporated yet, and that he designed to push your CPU just for the sake of pushing your CPU. But, from that example we are supposed to say that the entirety of HTML5 is bad and Flash is the future?
I think it's Safari; Safari 4.0.5 on Win7 uses up 25% of my CPU (Core i5-750) on that example, and doesn't even run it smoothly (motion blur set to 1000), whereas Chrome 6.0 dev only uses about 10% and runs as smooth as silk. (motion blur also at 1000).
I just tried the latest nightly of Webkit, and it doesn't do any better. I could try it on my Mac, but I know it will run like crap (it's an old C2D).
It may be a newer part of HTML5, but if Apple needs GPU acceleration to do canvas, they're doing it wrong.
The single most important aspect of Flash on the web today is video and this is considerably more costly on Flash than it is in HTML5. That's relevant.
Costly in what way?
Because if by "cost" you mean cost, remember that h264 isn't free.
Sure, Steve likes to point out that it currently has no *license* fees, and when spoken exactly like that he's technically correct. But as the patent holder of h264, Apple expects vendors writing h264 codecs to pay them other fees which are not called "license" fees.
Then in less than five years they start adding in license fees too -- classic bait-n-switch -- so there's a cost now and no way to know how much dependency on h264 is going to cost us in the future.
All we know is that Apple sits on the receiving end of that cash stream.
I think it's Safari; Safari 4.0.5 on Win7 uses up 25% of my CPU (Core i5-750) on that example, and doesn't even run it smoothly (motion blur set to 1000), whereas Chrome 6.0 dev only uses about 10% and runs as smooth as silk. (motion blur also at 1000).
I just tried the latest nightly of Webkit, and it doesn't do any better. I could try it on my Mac, but I know it will run like crap (it's an old C2D).
It may be a newer part of HTML5, but if Apple needs GPU acceleration to do canvas, they're doing it wrong.
The word is that Safari 5 adds the much needed HW acceleration to Safari on Windows. I have my doubts as to how well it will help with Canvas, but it's a start.
Costly in what way?
Because if by "cost" you mean cost, remember that h264 isn't free.
Sure, Steve likes to point out that it currently has no *license* fees, and when spoken exactly like that he's technically correct. But as the patent holder of h264, Apple expects vendors writing h264 codecs to pay them other fees which are not called "license" fees.
Then in less than five years they start adding in license fees too -- classic bait-n-switch -- so there's a cost now and no way to know how much dependency on h264 is going to cost us in the future.
All we know is that Apple sits on the receiving end of that cash stream.
Awesome¡ A comment about processing overhead regarding Flash v. HTML5 for video streaming and you jump into a comment about H.264 which is not germane to the topic and is an optional codec for each.
The word is that Safari 5 adds the much needed HW acceleration to Safari on Windows. I have my doubts as to how well it will help with Canvas, but it's a start.
Awesome¡ A comment about processing overhead regarding Flash v. HTML5 for video streaming and you jump into a comment about H.264 which is not germane to the topic and is an optional codec for each.
HTML5 is not a codec.
HTML5 is not a codec.
Yet you are a guy who bought up a codec, that both HTML5 and Flash can use, among other codecs, despite it not being relevant to my comment. It must be hard to be rational when you reading comprehension is such an issue for you.