Well, no the W3C does not have any enforcement authority, but they could certainly pick a reference implementation, and it would be known which browsers are "compliant" or not.
Assuming, for the sake of argument, that half of IE users install the Google plugin for some reason, then, based on your numbers, from that alone, WebKit ends up with about 53% marketshare, while IE either remains the same (in which case the totals are greater than 100%, of course the totals are really already greater than 100% because many people run multiple browsers) or decreases to about 33% (depending on how you count it and how users use it.) WebKit market share is increasing on it's own, so, even if it's not half, but some smaller number, the impact could be significant.
As far as mobile goes (and this applies to non-mobile as well) the important number is not the installed base, but the amount of actual web use, which most surveys seem to indicate is overwhelmingly WebKit based.
And, yes, of course I am speculating.
EDIT: My mistake, I misread your numbers, so, in correction: WebKit goes to ~40%, not 53%. The other numbers remain the same.
My last 2 cents...
If they did pick one of the vendors engines as a reference standard, the others would never accept it.
These are useage statistics, not installed base. The numbers are based on actual http requests to web pages.
If they did pick one of the vendors engines as a reference standard, the others would never accept it.
These are useage statistics, not installed base. The numbers are based on actual http requests to web pages.
Well, accepting it doesn't mean they have to adopt it, just implement the feature set of the reference implementation so that it behaves in the same way. I don't expect FF is likely to drop their rendering engine any time soon, however, they might not have a problem with matching a reference implementation feature set. (Except of course for the video issues.)
I don't think numbers showing Opera with the highest number of mobile http requests can be correct, or perhaps those numbers are based on something else? They would certainly contradict all of the numbers I have seen.
Anyway, I don't actually think the W3C would ever adopt a specific implementation as a reference, and my post was mainly to discuss the pros and cons, which is how I got to the 4th hand.
Conveniently the reference implementation would be WebKit as shipped by Apple, right? All other browser vendors would be at the mercy of the WebKit developers and their whims which parts of a standard to implement and not. This idea is hopeless since it would be never accepted by other vendors, especially if the W3C would chose an engine which is only used by a minority of people on this planet. I'd rather opt that the browser vendors work on an extensive test suite which could be used as a reference for how standard compliant a browser is. This would be neutral enough and every browser vendor could contribute to such a test suite independently.
If we waited for standards to be completed and ratified before putting them into production then we?d only have 802.11n routers for a couple weeks now, yet MS has been including 802.11n drivers in Windows for some time now.
I have only superficial knowledge on this topic but was this not caused by hardware producers selling hardware which were based on a draft version of the standard as they needed a shiny new product they could sell? If that would be the case Microsoft was practically forced to include drivers in order to satisfy customer expectations.
Quote:
That just isn?t how the technology tends to work. With HTML5, it?s a very complex system. When one part is completed there is no reason that a browser engine developer shouldn?t try to implement it, especially when that addition can assist in the speed and usefulness of the browser. If it?s a popular enough it will get picked up by other developers and site developers. No need to wait 5 years before we put these efficient and useful additions into the browsers. Some of these HTML5 inclusions may need to change, but for the most part each part will be finished enough that they can included as needed.
On the other hand implementing half-finished standards creates facts which are hard to change, especially if a browser vendor used a lot of resources to implement the draft specification. This can be problematic as suddenly all other browser vendors are forced to implement substandard specifications because it's already used in the wild and we and up in a mixed compatibility hell. Take for instance Microsoft who provided constructive feedback on the HTML5 standard for the first time. If their criticism is accepted some browser vendors will have to rewrite their engines and worse, customers of the engine feature as well.
Quote:
We?re finally seeing the momentum pick up in browser development after such a long stagnation. This is a very good thing.
We can all thank Firefox for that for breaking the iron grip Microsoft had on the browser market.
Conveniently the reference implementation would be WebKit as shipped by Apple, right? All other browser vendors would be at the mercy of the WebKit developers and their whims which parts of a standard to implement and not. This idea is hopeless since it would be never accepted by other vendors, especially if the W3C would chose an engine which is only used by a minority of people on this planet. I'd rather opt that the browser vendors work on an extensive test suite which could be used as a reference for how standard compliant a browser is. This would be neutral enough and every browser vendor could contribute to such a test suite independently.
Yeah a comprehensive test suite would be a good way to go.
We can all thank Firefox for that for breaking the iron grip Microsoft had on the browser market.
Absolutely - it was the first browser to kick IE's ass in performance by every measure. I still think it's the best browser out there, especially if you do any web development, and/or you have to use Windows. There are good choices on either platform, but on Mac you really don't need much choice because Safari is so good. On Windows... thank god for Firefox.
That means most IE users will enhance their relative security by either installing the Chrome Frame plugin or downloading Chrome 3, Safari 4, or Firefox 4 to use instead.
Or how about IE8, just because older versions have security risks it doesn't change the security of the latest one. I was liking the article up until this point as it was seeming fairly unbiased and based on facts. Nothing changes the fact that there are parts of IE8 that are more secure than the others. For example IE8 will block certain cross site scripting attacks, whereas if you attempt the same attach on the others they've just left it up to the website to protect itself. As cross site scripting is a fairly big thing, shouldn't this be noted.
Quote:
Originally Posted by AppleInsider
Additionally, since HTML 5 is designed to replace Flash and Silverlight, any studies recommending a web browser solely on the basis of security will need to factor in all of the security vulnerabilities in Flash and Silverlight against using a browser that does not need them to render rich web applications. The reports Microsoft cites in favor of IE 8 do not.
That makes no sense, you still need Flash and Silverlight for sites with Flash and Silverlight content! Contrary to what seems many peoples beliefs you also don't need HTML5 to make a rich web application. You can do it in even IE6, it was after all Microsoft that introduced the essential elements that make AJAX work.
On the plus side does this now make IE8 the most compatible browser? Given that it now has the IE7, IE8 and WebKit rendering engines.
The main objection seems to be about control though, which is usually the reason you find projects with similar goals fragmented. The comment about hacking webkit a lot is perhaps exaggerated as they will mean hacking it to do what Firefox does now like building the interface in XUL and allowing plugins. I don't think they'd have to go to those lengths or at the least they would build a plugin system around the webkit engine the way Safari does.
It means they can focus their development effort more on the application experience, which as I said earlier is mainly why people pick one browser or another it's not really the layout, which is just expected to be correct and that burden is placed on web developers.
Quote:
Originally Posted by Erunno
All other browser vendors would be at the mercy of the WebKit developers and their whims which parts of a standard to implement and not. This idea is hopeless since it would be never accepted by other vendors, especially if the W3C would chose an engine which is only used by a minority of people on this planet.
It's a hypothetical though and one not based on the stats. Webkit has been consistently the first and best for adherence to web standards out of any engine and building an engine is no trivial task, which is why there are so few major ones. You've got IE's Trident engine, Mozilla's Gecko, Webkit and Opera's Presto.
It makes sense that you'd pick one of the big ones and there's no way IE's track record for compliance wins it any favors nor is it open source. Opera's engine isn't open source either and in low usage. Firefox's engine has a lot of code bloat that has an impact for mobile usage. The choice points heavily in favor of Webkit.
I consider a browser engine to be like a graphics driver. NVidia and ATI generally don't give OS vendors a spec sheet, they write the drivers and that way guarantee that drawing calls to their driver work correctly. I know that Apple choose to right parts themselves but that's a good example of where their implementation falls short of Windows counterparts in terms of support for various features like anti-aliasing, hardware decoding etc.
I definitely have mixed feelings on this story. IE is a nightmare to develop for, Flash and Silverlight need to go away
Just out of interest why (serious question)?
There seems a lot of talk about negative affects of this for Microsoft and Adobe. But what about Apple? Silverlight and Flash are both capable of running applications at the same level as any desktop app. Flash even installs as a runtime for desktop apps, and Silverlight uses the same language and libraries as a desktop app. If HTML5 is really good enough to replace Silverlight and Flash, doesn't that also mean its good enough to completely replace any need to develop native iPhone apps?
Why pay Apple 30% for every product you sell, and require your developers to learn how to program native iPhone apps when theoretically you could be programming for any device HTML5 capable in a standard language, that could also be as good as a native iPhone app? I would have though Apple has more to loose than anyone.
Adobe's monopoly on Flash and having to buy the developer tools from Adobe might just be one thing.
Flash is proprietry users are left to Adobe's whim as to when updates etc become available. e.g. a few years ago having to wait for 64bit versions especially for Linux.
HTML 5 is open, WebKit developed by Apple is open.
Quote:
Originally Posted by timgriff84
Just out of interest why (serious question)?
There seems a lot of talk about negative affects of this for Microsoft and Adobe. But what about Apple? Silverlight and Flash are both capable of running applications at the same level as any desktop app. Flash even installs as a runtime for desktop apps, and Silverlight uses the same language and libraries as a desktop app. If HTML5 is really good enough to replace Silverlight and Flash, doesn't that also mean its good enough to completely replace any need to develop native iPhone apps?
Why pay Apple 30% for every product you sell, and require your developers to learn how to program native iPhone apps when theoretically you could be programming for any device HTML5 capable in a standard language, that could also be as good as a native iPhone app? I would have though Apple has more to loose than anyone.
Why pay Apple 30% for every product you sell, and require your developers to learn how to program native iPhone apps when theoretically you could be programming for any device HTML5 capable in a standard language, that could also be as good as a native iPhone app? I would have though Apple has more to loose than anyone.
Apple is getting 30% while the developer is getting 70% for a free SDK, no CC fees, free hosting, etc. While an app built with Obj-C will be worlds better than an HTML-based app, you are correct that there are many apps on the apps store that could easily be web-based, but how would they profit off of them if they weren't for sale. This seems more like a free-app solution where Apple gets 30% of zero-zip-nada amd still does the hosting and update revisions for free.
HTML5 benefits Apple despites it having local DB caching so you don't even need to be connected to the Internet and their work in getting WebGL incorporated into WebKit so that the browser can even use HW acceleration. For the user we can also get app that Apple won't. Authorize for various reasons, but which all lead back to protecting their own ass, like with ae appropriate material (porn), potentially illegal material (torrents), etc. As a publically traded company they have do have to maintain a certain "family" imagine and try to get sued when it can Be avoided. It's jay good business.
Comments
Well, no the W3C does not have any enforcement authority, but they could certainly pick a reference implementation, and it would be known which browsers are "compliant" or not.
Assuming, for the sake of argument, that half of IE users install the Google plugin for some reason, then, based on your numbers, from that alone, WebKit ends up with about 53% marketshare, while IE either remains the same (in which case the totals are greater than 100%, of course the totals are really already greater than 100% because many people run multiple browsers) or decreases to about 33% (depending on how you count it and how users use it.) WebKit market share is increasing on it's own, so, even if it's not half, but some smaller number, the impact could be significant.
As far as mobile goes (and this applies to non-mobile as well) the important number is not the installed base, but the amount of actual web use, which most surveys seem to indicate is overwhelmingly WebKit based.
And, yes, of course I am speculating.
EDIT: My mistake, I misread your numbers, so, in correction: WebKit goes to ~40%, not 53%. The other numbers remain the same.
My last 2 cents...
If they did pick one of the vendors engines as a reference standard, the others would never accept it.
These are useage statistics, not installed base. The numbers are based on actual http requests to web pages.
My last 2 cents...
If they did pick one of the vendors engines as a reference standard, the others would never accept it.
These are useage statistics, not installed base. The numbers are based on actual http requests to web pages.
Well, accepting it doesn't mean they have to adopt it, just implement the feature set of the reference implementation so that it behaves in the same way. I don't expect FF is likely to drop their rendering engine any time soon, however, they might not have a problem with matching a reference implementation feature set. (Except of course for the video issues.)
I don't think numbers showing Opera with the highest number of mobile http requests can be correct, or perhaps those numbers are based on something else? They would certainly contradict all of the numbers I have seen.
Anyway, I don't actually think the W3C would ever adopt a specific implementation as a reference, and my post was mainly to discuss the pros and cons, which is how I got to the 4th hand.
If we waited for standards to be completed and ratified before putting them into production then we?d only have 802.11n routers for a couple weeks now, yet MS has been including 802.11n drivers in Windows for some time now.
I have only superficial knowledge on this topic but was this not caused by hardware producers selling hardware which were based on a draft version of the standard as they needed a shiny new product they could sell? If that would be the case Microsoft was practically forced to include drivers in order to satisfy customer expectations.
That just isn?t how the technology tends to work. With HTML5, it?s a very complex system. When one part is completed there is no reason that a browser engine developer shouldn?t try to implement it, especially when that addition can assist in the speed and usefulness of the browser. If it?s a popular enough it will get picked up by other developers and site developers. No need to wait 5 years before we put these efficient and useful additions into the browsers. Some of these HTML5 inclusions may need to change, but for the most part each part will be finished enough that they can included as needed.
On the other hand implementing half-finished standards creates facts which are hard to change, especially if a browser vendor used a lot of resources to implement the draft specification. This can be problematic as suddenly all other browser vendors are forced to implement substandard specifications because it's already used in the wild and we and up in a mixed compatibility hell. Take for instance Microsoft who provided constructive feedback on the HTML5 standard for the first time. If their criticism is accepted some browser vendors will have to rewrite their engines and worse, customers of the engine feature as well.
We?re finally seeing the momentum pick up in browser development after such a long stagnation. This is a very good thing.
We can all thank Firefox for that for breaking the iron grip Microsoft had on the browser market.
Conveniently the reference implementation would be WebKit as shipped by Apple, right? All other browser vendors would be at the mercy of the WebKit developers and their whims which parts of a standard to implement and not. This idea is hopeless since it would be never accepted by other vendors, especially if the W3C would chose an engine which is only used by a minority of people on this planet. I'd rather opt that the browser vendors work on an extensive test suite which could be used as a reference for how standard compliant a browser is. This would be neutral enough and every browser vendor could contribute to such a test suite independently.
Yeah a comprehensive test suite would be a good way to go.
We can all thank Firefox for that for breaking the iron grip Microsoft had on the browser market.
Absolutely - it was the first browser to kick IE's ass in performance by every measure. I still think it's the best browser out there, especially if you do any web development, and/or you have to use Windows. There are good choices on either platform, but on Mac you really don't need much choice because Safari is so good. On Windows... thank god for Firefox.
That means most IE users will enhance their relative security by either installing the Chrome Frame plugin or downloading Chrome 3, Safari 4, or Firefox 4 to use instead.
Or how about IE8, just because older versions have security risks it doesn't change the security of the latest one. I was liking the article up until this point as it was seeming fairly unbiased and based on facts. Nothing changes the fact that there are parts of IE8 that are more secure than the others. For example IE8 will block certain cross site scripting attacks, whereas if you attempt the same attach on the others they've just left it up to the website to protect itself. As cross site scripting is a fairly big thing, shouldn't this be noted.
Additionally, since HTML 5 is designed to replace Flash and Silverlight, any studies recommending a web browser solely on the basis of security will need to factor in all of the security vulnerabilities in Flash and Silverlight against using a browser that does not need them to render rich web applications. The reports Microsoft cites in favor of IE 8 do not.
That makes no sense, you still need Flash and Silverlight for sites with Flash and Silverlight content! Contrary to what seems many peoples beliefs you also don't need HTML5 to make a rich web application. You can do it in even IE6, it was after all Microsoft that introduced the essential elements that make AJAX work.
On the plus side does this now make IE8 the most compatible browser? Given that it now has the IE7, IE8 and WebKit rendering engines.
If they did pick one of the vendors engines as a reference standard, the others would never accept it.
Yeah, the Mozilla VP of engineering said so here:
http://arstechnica.com/open-source/n...to-gecko.ars/2
The main objection seems to be about control though, which is usually the reason you find projects with similar goals fragmented. The comment about hacking webkit a lot is perhaps exaggerated as they will mean hacking it to do what Firefox does now like building the interface in XUL and allowing plugins. I don't think they'd have to go to those lengths or at the least they would build a plugin system around the webkit engine the way Safari does.
It means they can focus their development effort more on the application experience, which as I said earlier is mainly why people pick one browser or another it's not really the layout, which is just expected to be correct and that burden is placed on web developers.
All other browser vendors would be at the mercy of the WebKit developers and their whims which parts of a standard to implement and not. This idea is hopeless since it would be never accepted by other vendors, especially if the W3C would chose an engine which is only used by a minority of people on this planet.
It's a hypothetical though and one not based on the stats. Webkit has been consistently the first and best for adherence to web standards out of any engine and building an engine is no trivial task, which is why there are so few major ones. You've got IE's Trident engine, Mozilla's Gecko, Webkit and Opera's Presto.
It makes sense that you'd pick one of the big ones and there's no way IE's track record for compliance wins it any favors nor is it open source. Opera's engine isn't open source either and in low usage. Firefox's engine has a lot of code bloat that has an impact for mobile usage. The choice points heavily in favor of Webkit.
I consider a browser engine to be like a graphics driver. NVidia and ATI generally don't give OS vendors a spec sheet, they write the drivers and that way guarantee that drawing calls to their driver work correctly. I know that Apple choose to right parts themselves but that's a good example of where their implementation falls short of Windows counterparts in terms of support for various features like anti-aliasing, hardware decoding etc.
I definitely have mixed feelings on this story. IE is a nightmare to develop for, Flash and Silverlight need to go away
Just out of interest why (serious question)?
There seems a lot of talk about negative affects of this for Microsoft and Adobe. But what about Apple? Silverlight and Flash are both capable of running applications at the same level as any desktop app. Flash even installs as a runtime for desktop apps, and Silverlight uses the same language and libraries as a desktop app. If HTML5 is really good enough to replace Silverlight and Flash, doesn't that also mean its good enough to completely replace any need to develop native iPhone apps?
Why pay Apple 30% for every product you sell, and require your developers to learn how to program native iPhone apps when theoretically you could be programming for any device HTML5 capable in a standard language, that could also be as good as a native iPhone app? I would have though Apple has more to loose than anyone.
Flash is proprietry users are left to Adobe's whim as to when updates etc become available. e.g. a few years ago having to wait for 64bit versions especially for Linux.
HTML 5 is open, WebKit developed by Apple is open.
Just out of interest why (serious question)?
There seems a lot of talk about negative affects of this for Microsoft and Adobe. But what about Apple? Silverlight and Flash are both capable of running applications at the same level as any desktop app. Flash even installs as a runtime for desktop apps, and Silverlight uses the same language and libraries as a desktop app. If HTML5 is really good enough to replace Silverlight and Flash, doesn't that also mean its good enough to completely replace any need to develop native iPhone apps?
Why pay Apple 30% for every product you sell, and require your developers to learn how to program native iPhone apps when theoretically you could be programming for any device HTML5 capable in a standard language, that could also be as good as a native iPhone app? I would have though Apple has more to loose than anyone.
Why pay Apple 30% for every product you sell, and require your developers to learn how to program native iPhone apps when theoretically you could be programming for any device HTML5 capable in a standard language, that could also be as good as a native iPhone app? I would have though Apple has more to loose than anyone.
Apple is getting 30% while the developer is getting 70% for a free SDK, no CC fees, free hosting, etc. While an app built with Obj-C will be worlds better than an HTML-based app, you are correct that there are many apps on the apps store that could easily be web-based, but how would they profit off of them if they weren't for sale. This seems more like a free-app solution where Apple gets 30% of zero-zip-nada amd still does the hosting and update revisions for free.
HTML5 benefits Apple despites it having local DB caching so you don't even need to be connected to the Internet and their work in getting WebGL incorporated into WebKit so that the browser can even use HW acceleration. For the user we can also get app that Apple won't. Authorize for various reasons, but which all lead back to protecting their own ass, like with ae appropriate material (porn), potentially illegal material (torrents), etc. As a publically traded company they have do have to maintain a certain "family" imagine and try to get sued when it can Be avoided. It's jay good business.
once you go mac you dont go back
Here's to hoping Google getting it into 99% of IE users!