Actually, you can pretty much blame that on Java - it's a disaster on highly parallel gear.
Do you have references for that? In my experience, Java handles threads mostly well. The last experience I have was on a 16CPU Sun beast (don't remember the model unfortunately). Is that a deficiency of Java on this architecture?
I'm fine, thanks. People are doing websites with HTML5 by the hundred of thousands, every week. All browsers support said standard. So whether the standard is draft or not, it really only matters from a rhetoric and theological standpoint. For all practical purposes, HTML5 is alive and breathing. It is doing fine.
Flipao said "changes in the spec might make the version supported by ALL browsers OBSOLETE overnight". This is pure nonsense. First, it will never happen. There would be no point. Second, all my HTML5 work would still work and would most likely survive much longer than "overnight". Third, I highly doubt any browser editor will remove support for something they just added, works, is compatible with other browsers, and is widely used in the real world.
Ok, you are speaking about theory. My argument is that in practice IT DOESN'T matter. The standard IS HERE, albeit in a draft form. And all major browsers have implemented said standard. It is used in the real word AS WE ARGUE. And no, none of the major browser vendors are going to remove all traces of HTML5 or CSS3 anytime soon. So, for all practical purposes, we have a standard, and we have an implementation.
In theory we have nothing. But hopefully, we don't live in theory.
The closest thing to that is new CSS media types that can apply to different devices. You can for example specify a CSS stylesheet for devices with less than 320 pixels of height. Or for devices with exactly 640px width.
This doesn't quite fit the bill completely, but it is a good start. It actually allows one page to display differently on different window sizes.
There you go. I am giving out examples, and people nitpick. How hard is it to understand that page layout CANNOT be the same on a 7 inch screen and on a 30 inch screen?
And no, CSS is NOT here to do page layout. It can help to some extent, but it cannot do everything one would need as far as page layout is concerned. You might even choose to put less information on every page and break down your site into more smaller pages. Try to do that with JavaScript or CSS.
I try to use javascript exclusively for user interaction.
Ah, and thank you for the "old" qualification, and the assumption that I can't adapt nor can I see anything better than my old habbits. Basically, you're calling me grandpa and patting me on the back. Do you even see how rude and incredibly stupid that is?
There is no "No Child Left Behind Act" anymore. We now call it the "No Child Goes Ahead Act". It does a better job at describing the idiocy of the whole thing.
Ok, my sentence was wrong, and it should have read "What the fuck does anyone care about the draft status of the standard?"
My point was that the standard is there, it is followed by ALL browser manufacturers. The fact that it is a draft is brought up so often that it gets on my nerves. Apologies.
People answering to "You have two choices" by "Why do you try to impose YOUR choice?" tend to annoy me. And no, this isn't the worst flamewar ever on slashdot, by a pretty fat margin.
Here at Slashdot, we have productive, intellectual discussions
The GP failed harder than me at that rule, since he didn't even read my post. Note that he said his eyes were hurt, hence maybe giving us a hint at why he didn't read what I wrote in the first place.
Please answer this question honestly: How old are you?
38. I don't see how it's relevant though. And I've been building website professionally for the last 15 years.
I ask merely because your comments come off as if written by somebody who is 12 or 13 years old, and who thus missed out on experiencing a large portion of the WWW's history.
Like others have pointed out, changing the content you're serving based on the User-Agent header value has always been a bad practice, and should be avoided. Every experienced web developer knows this.
Contrary to what you say, the User-Agent header was not originally meant to be used as you describe. Some of the earliest attempts at creating specifications, such as those from 1992 found at http://www.w3.org/Protocols/HTTP/HTRQ_Headers.html#user-agent, clearly state that it was meant purely "for statistical purposes and the tracing of protocol violations." Neither of those advocate, nor even suggest, that the response should be modified depending on the header value.
Point taken, I wrote too fast.
Furthermore, you have clearly misunderstood the purpose of the WWW. It's a system for delivering documents that can link to one another. It is not about creating "UIs", like you mistakenly appear to believe.
A document is displayed to the user through a UI. Printing is on paper is a form of UI. Displaying it on a screen is a form of UI. By saying you don't need a UI, you're saying that you don't need the users to view your document.
When used properly, the exact same content should render perfectly fine on all sorts of devices with a wide range of capabilities and display media.
There's fine, and there's nice, which is pleasant to use. I guess your websites all look black & white (account for BW printers) with no images (account for Lynx users) and just Hn, b, u, i tags. And even those are bad. strong, em, quote should be used instead. Talk about living in the 20th cetury.
It's only when you misuse the WWW that you start running into the problems that you describe, and have to start attempting the horrible practices that you advocate. Indeed, different devices do need different UIs. The only tenable solution to that problem is to create separate native applications.
There are other options that work perfectly well. It's your right to choose not to use them of course. I respect that, contrary to you who disrespect what I do without actually knowing what I do for that matter. And you don't even want to know apparently.
All attempts at providing cross-platform UIs have failed,
For you.
regardless of who developed the technology, which platforms were targeted, and when it was attempted. Trust me, you won't succeed any better than they did, and will likely produce a result that is far worse.
Thanks for the tip. I have websites that work fine for years serving a smaller Logo to iPhones because, you know, they're slower devices with smaller screens. I guess that's a failure and my iPhone users should see a huge logo that takes forever to load. Or my desktop users should see a 100px by 50px logo. Because you said so.
Re:Mego is dead, Webos is dead ...
on
Intel Drops MeeGo
·
· Score: 1
It would look as it a mobile OS is only reachable by big megacorps (iOS, Android, Windows). Community based devlopment will never work here as every company has their own agenda to push and want it now. This is a recipe for disaster.
Users lost with a game that consumes 200% more CPU than it would need because the developer was too lazy to code it in the native language.
In the language of which device?
I think the discussion was about the iPad. So I'd say the iPad.
Say I'm trying to develop an application and make it available for two different devices. I already know I have to code separate front-ends for the two devices to fit into their respective user interface paradigms, but ideally the domain logic should be shared among all platforms. For example, a game's domain logic would include its physics and NPC behaviors. This concept has been called I/O abstraction, model-view-controller, or multitier. Now say one device runs only Objective-C and standard C++ (plus JavaScript in its web browser) and another runs only C#, F#, VB.NET, and other verifiably type-safe languages (plus JavaScript in its web browser). In such a case, in what language do I code the domain logic?
You basically have two choices here: (A) You can develop everything twice. That is if you place a high value on the quality of your app for each platform. More development, higher quality because said devlopment fit the platform. (B) You can choose some language that is abstracted on both platforms. Depending of the maturity of said language, you'll get more or less overhead. Less development, less hassle, lower quality for both platforms.
It was the top app for 3 days in a row and it is probably already fading into oblivion.
Be careful with that word, or you might end up on the business end of a lawsuit from Bethesda:p
I'm ready for their lawyers as we speak. That's why I used the word 'probably', so I'm covered as I expressed my own opinion and did not claim it was a fact;-)
A "user agent" is a client of the HTTP protocol. This is the terminology used in the RFCs. I never proposed to sniff the "User-agent" field of an http request.
What the fuck does anyone care about the standard? What we want is a technology with a sufficiently large install base so that it has a meaning to actually use it. HTML5 is this. AND it has a standard backing it up, which is an added bonus more or less guaranteeing a common set of features & API across all browsers
If they change the standard overnight (which will never happen, you know it) my code won't stop working... so why should I care? Why should anyone care? You really think IE, Firefox, Chrome, Opera, Safari will all release a new version within month breaking all websites using their actual implementation?
Vaporware is something that doesn't exist. HTML5 can be used TODAY on 60% of the browsers out there. And moreover it is really easy to make it downgrade gracefully on the 40% remaining IE8-
It's getting towards the end of 2011, and people are still advocating the use of user-agent sniffing? What the hell? The past 20 years of experience with this being a horrible approach hasn't convinced you to avoid it?
Hmmm. You haven't been very involved in web development lately it would seem. Different devices make it for different UIs. Maybe you missed that part. If you think your website can display itself as nicely on a 320x480 screen through Edge as on a 1920x1200 screen through a 20MB/s DSL connection, well, you've made a heck of a lot of compromises. These devices are best recognized by their User Agent, which by the way is the way the web was designed to work from the start.
Why would you say Steve lost? Users lost with a game that consumes 200% more CPU than it would need because the developer was too lazy to code it in the native language. I guess sometimes Good Enough is good enough.
And this crappy game is nowhere near "the number 1 paid app on the ipad" as you put it. It was the top app for 3 days in a row and it is probably already fading into oblivion.
I know you're just a troll, but I still want to set the record straight.
As for user agent sniffing, if you think that was the good old days, you've been living in a cave for the last 5 years.
Slashdot Mirror
The important thing is that Windows Phone 7.5 is out. All four users of Windows Phone 7 should be notified at once. Anyone knows their email address?
Actually, you can pretty much blame that on Java - it's a disaster on highly parallel gear.
Do you have references for that? In my experience, Java handles threads mostly well. The last experience I have was on a 16CPU Sun beast (don't remember the model unfortunately). Is that a deficiency of Java on this architecture?
I'm fine, thanks. People are doing websites with HTML5 by the hundred of thousands, every week. All browsers support said standard. So whether the standard is draft or not, it really only matters from a rhetoric and theological standpoint. For all practical purposes, HTML5 is alive and breathing. It is doing fine.
Flipao said "changes in the spec might make the version supported by ALL browsers OBSOLETE overnight". This is pure nonsense. First, it will never happen. There would be no point. Second, all my HTML5 work would still work and would most likely survive much longer than "overnight". Third, I highly doubt any browser editor will remove support for something they just added, works, is compatible with other browsers, and is widely used in the real world.
Ok, you are speaking about theory. My argument is that in practice IT DOESN'T matter. The standard IS HERE, albeit in a draft form. And all major browsers have implemented said standard. It is used in the real word AS WE ARGUE. And no, none of the major browser vendors are going to remove all traces of HTML5 or CSS3 anytime soon. So, for all practical purposes, we have a standard, and we have an implementation.
In theory we have nothing. But hopefully, we don't live in theory.
The closest thing to that is new CSS media types that can apply to different devices. You can for example specify a CSS stylesheet for devices with less than 320 pixels of height. Or for devices with exactly 640px width.
This doesn't quite fit the bill completely, but it is a good start. It actually allows one page to display differently on different window sizes.
There you go. I am giving out examples, and people nitpick. How hard is it to understand that page layout CANNOT be the same on a 7 inch screen and on a 30 inch screen?
And no, CSS is NOT here to do page layout. It can help to some extent, but it cannot do everything one would need as far as page layout is concerned. You might even choose to put less information on every page and break down your site into more smaller pages. Try to do that with JavaScript or CSS.
I try to use javascript exclusively for user interaction.
Ah, and thank you for the "old" qualification, and the assumption that I can't adapt nor can I see anything better than my old habbits. Basically, you're calling me grandpa and patting me on the back. Do you even see how rude and incredibly stupid that is?
Please, come on. Let's get real over here.
There is no "No Child Left Behind Act" anymore. We now call it the "No Child Goes Ahead Act". It does a better job at describing the idiocy of the whole thing.
Ok, my sentence was wrong, and it should have read "What the fuck does anyone care about the draft status of the standard?"
My point was that the standard is there, it is followed by ALL browser manufacturers. The fact that it is a draft is brought up so often that it gets on my nerves. Apologies.
People answering to "You have two choices" by "Why do you try to impose YOUR choice?" tend to annoy me. And no, this isn't the worst flamewar ever on slashdot, by a pretty fat margin.
Here at Slashdot, we have productive, intellectual discussions
The GP failed harder than me at that rule, since he didn't even read my post. Note that he said his eyes were hurt, hence maybe giving us a hint at why he didn't read what I wrote in the first place.
Which part of "You have two choices" (emphasis mine) didn't you get you dumbass?
Please answer this question honestly: How old are you?
38. I don't see how it's relevant though. And I've been building website professionally for the last 15 years.
I ask merely because your comments come off as if written by somebody who is 12 or 13 years old, and who thus missed out on experiencing a large portion of the WWW's history.
Like others have pointed out, changing the content you're serving based on the User-Agent header value has always been a bad practice, and should be avoided. Every experienced web developer knows this.
Contrary to what you say, the User-Agent header was not originally meant to be used as you describe. Some of the earliest attempts at creating specifications, such as those from 1992 found at http://www.w3.org/Protocols/HTTP/HTRQ_Headers.html#user-agent, clearly state that it was meant purely "for statistical purposes and the tracing of protocol violations." Neither of those advocate, nor even suggest, that the response should be modified depending on the header value.
Point taken, I wrote too fast.
Furthermore, you have clearly misunderstood the purpose of the WWW. It's a system for delivering documents that can link to one another. It is not about creating "UIs", like you mistakenly appear to believe.
A document is displayed to the user through a UI. Printing is on paper is a form of UI. Displaying it on a screen is a form of UI. By saying you don't need a UI, you're saying that you don't need the users to view your document.
When used properly, the exact same content should render perfectly fine on all sorts of devices with a wide range of capabilities and display media.
There's fine, and there's nice, which is pleasant to use. I guess your websites all look black & white (account for BW printers) with no images (account for Lynx users) and just Hn, b, u, i tags. And even those are bad. strong, em, quote should be used instead. Talk about living in the 20th cetury.
It's only when you misuse the WWW that you start running into the problems that you describe, and have to start attempting the horrible practices that you advocate. Indeed, different devices do need different UIs. The only tenable solution to that problem is to create separate native applications.
There are other options that work perfectly well. It's your right to choose not to use them of course. I respect that, contrary to you who disrespect what I do without actually knowing what I do for that matter. And you don't even want to know apparently.
All attempts at providing cross-platform UIs have failed,
For you.
regardless of who developed the technology, which platforms were targeted, and when it was attempted. Trust me, you won't succeed any better than they did, and will likely produce a result that is far worse.
Thanks for the tip. I have websites that work fine for years serving a smaller Logo to iPhones because, you know, they're slower devices with smaller screens. I guess that's a failure and my iPhone users should see a huge logo that takes forever to load. Or my desktop users should see a 100px by 50px logo. Because you said so.
It would look as it a mobile OS is only reachable by big megacorps (iOS, Android, Windows). Community based devlopment will never work here as every company has their own agenda to push and want it now. This is a recipe for disaster.
Users lost with a game that consumes 200% more CPU than it would need because the developer was too lazy to code it in the native language.
In the language of which device?
I think the discussion was about the iPad. So I'd say the iPad.
Say I'm trying to develop an application and make it available for two different devices. I already know I have to code separate front-ends for the two devices to fit into their respective user interface paradigms, but ideally the domain logic should be shared among all platforms. For example, a game's domain logic would include its physics and NPC behaviors. This concept has been called I/O abstraction, model-view-controller, or multitier. Now say one device runs only Objective-C and standard C++ (plus JavaScript in its web browser) and another runs only C#, F#, VB.NET, and other verifiably type-safe languages (plus JavaScript in its web browser). In such a case, in what language do I code the domain logic?
You basically have two choices here:
(A) You can develop everything twice. That is if you place a high value on the quality of your app for each platform. More development, higher quality because said devlopment fit the platform.
(B) You can choose some language that is abstracted on both platforms. Depending of the maturity of said language, you'll get more or less overhead. Less development, less hassle, lower quality for both platforms.
It was the top app for 3 days in a row and it is probably already fading into oblivion.
Be careful with that word, or you might end up on the business end of a lawsuit from Bethesda :p
I'm ready for their lawyers as we speak. That's why I used the word 'probably', so I'm covered as I expressed my own opinion and did not claim it was a fact ;-)
A "user agent" is a client of the HTTP protocol. This is the terminology used in the RFCs. I never proposed to sniff the "User-agent" field of an http request.
You are correct. RFC 1945 officially introduced and recognized HTTP V1.0 in 1996. That's still pretty old, and it was designed for this very purpose.
What the fuck does anyone care about the standard? What we want is a technology with a sufficiently large install base so that it has a meaning to actually use it. HTML5 is this. AND it has a standard backing it up, which is an added bonus more or less guaranteeing a common set of features & API across all browsers
If they change the standard overnight (which will never happen, you know it) my code won't stop working... so why should I care? Why should anyone care? You really think IE, Firefox, Chrome, Opera, Safari will all release a new version within month breaking all websites using their actual implementation?
Get real.
Vaporware is something that doesn't exist. HTML5 can be used TODAY on 60% of the browsers out there. And moreover it is really easy to make it downgrade gracefully on the 40% remaining IE8-
It's getting towards the end of 2011, and people are still advocating the use of user-agent sniffing? What the hell? The past 20 years of experience with this being a horrible approach hasn't convinced you to avoid it?
Hmmm. You haven't been very involved in web development lately it would seem. Different devices make it for different UIs. Maybe you missed that part. If you think your website can display itself as nicely on a 320x480 screen through Edge as on a 1920x1200 screen through a 20MB/s DSL connection, well, you've made a heck of a lot of compromises. These devices are best recognized by their User Agent, which by the way is the way the web was designed to work from the start.
Why would you say Steve lost? Users lost with a game that consumes 200% more CPU than it would need because the developer was too lazy to code it in the native language. I guess sometimes Good Enough is good enough.
And this crappy game is nowhere near "the number 1 paid app on the ipad" as you put it. It was the top app for 3 days in a row and it is probably already fading into oblivion.
I know you're just a troll, but I still want to set the record straight.
As for user agent sniffing, if you think that was the good old days, you've been living in a cave for the last 5 years.
Thnhk along the line of "This is a website, so it is best viewed with a web browser".
As I said, the web is back.
SlideShare is now viewable on every kind of mobile device
As long as they support that very flavor of HTML5 which, in turn, is still a draft!
It is still a draft. So? What does this mean? We shouldn't endorse it? Especially now that ALL browsers support it?
Wake me up when youtube ditches Flash.
Hello? Are you there? It's time to wake up. Type youtube.com on an iDevice to see it at work.
And we hate flash and it's good to see these moves away from proprietary tech to open standards. The web is back.
Sure. But you can always serve the proper video to the proper user agent. This is no rocket science.