> it takes your programmers ten times as long to do as it does in Java or VB or C# or whatever.
All the web programmers out there are trying to figure out why this is a bad thing. If the business people want more expensive development, more jobs for us!:)
(And seriously, I think you underestimate the real costs of deployment and support. In a large environment it can be ridiclously expensive to solve a client problem. And with many apps living on the "extranet", web is the only way to go.)
What I am arguing against is the idea that AJAX makes webapps powerfull enough to drive all kinds of desktop app extinct, or pose a threat to Microsoft.
It doesn't pose any additional threat to Microsoft, because web applciations have dominated over most traditional Windows programming for years. You're talking about hte future, but this is all history now. The market for VB Forms (which was once something like 50% of all programming jobs) or MFC is basically dead.
Microsoft survived and prospered because they stayed ahead of the game in web application framewoks, and because their browser was significantly better than the competition for client-scripting and extendablity. The revenue from Windows Server licences has more than covered the loss of Win32-lockin.
Different JVMs should be more compatible than different Web browsers currently are.
Key word being "should". In practice, the transition from MS/NS Java 1.1 to Sun Java 1.2 broke a fuckload of applets (solution was usually to dump Java). You also hear of companies having to maintain three or four different JVMs on all their workstations due to specific Java app requirements.
In terms of Javascript, "Version 6 and above" for IE and Mozilla has been a stable platform for years now.
Microsoft didn't invent it. Nutscrape was encouraging people to do the same thing with Java Applets back when Communicator 4 was released. Before MS had XMLHttpRequest, they had their own applet.
As the other poster mentioned, the Hidden Frame trick also dates back to the v4 browser era.
As someone who had been doing "AJAX" sites since about 1998, I have to agree. The "A" stands for Asynchronous -- a browser's javascript environment is multithreaded -- when you click a button three times, you effectively fire off three different threads that run in parallel. Oh, and you have no effective way of synchronizing them.
The bottom line is that it's very easy to write something that works fine against localhost or your development server but blows up when used under realworld network and server conditions. If you are doing something complicated, you need to expect that the page state may be inconsistent at any time, and code around it.
(Note that I haven't tried any of the new AJAX frameworks, so they may have a solution for this.)
Except these sites weren't like MySpace, CDBaby, or Google -- they were mostly just lame clones of Amazon.com.
I think everyone involved knew these commercials were a long shot at best, and their real purpose was to advertise the IPOs and not the sites. The goal was to flush as much investor money down the toliet as fast as possible, and hope everyone could cash out before the house of cards collapsed. Nobody was trying to build a successful business.
I will agree that Apple went to Intel for primarily financial and business reasons rather than the technical merits/demerits of PPC v Cell v x86.
It is hilarious(and sad) to think of all that hardwork Apple engineers went through to be ready for media processing monster chips like Cell
The problem with this argument is that only a small percentage of Apple's business is related to high-end, specialized media-processing. Apple primarily markets general purpose computers to consumers. And even then, Intel provides enough raw oophf to handle most media tasks with ease.
It was funny how the first comments about the Intel Macs was how fast Safari was. The fact is that competitive Integer performance much more important to the general user experience than obscure encoding tasks.
They went your route with the G4 -- you ended up with a computer that was fast for playing videos and 3 photoshop filters, and dog slow 99% of the time. Not to mention it was much more complex and expensive to program for. The bottom line is that Apple doesn't have the market or developer support for something like Cell. I'm sure the 12 people who buy Sony's machines will enjoy them.
> It's turning into same mistake Apple made back in the day
(Aside from all the hystrical business-political rhetoric on this story)
It's possible it could be the same technical limitations that Apple is seeing right now. They are using OpenGL for desktop effects, and consequentally the game performance is teh suck.
Hey Transhuman -- Doubleclick cookies and the like are reported by Spyware apps, but aren't really spyware. If you look in Firefox, you'll have all the same cookies.
Are people angry at this? And if so, how has it hurt Microsoft? I think history has shown that treating the entire computer as a dongle had a far greater affect on consumers than copy-protection.
The hardware chip is doing public-private key encryption. If you can mimic this, there's much more profitable endevors available to you than getting OS X running on your Dell.
And if not outright outlaw Linux, at least make sure only generic whitebox motherboards from Taiwan run it. The Dell and HPs will all be locked to the copy of Windows married to their TCPA module during manufacturing.
This silly conspiracy theory is getting tiring. Why would Dell & HP prevent paying customers from running Linux or DOS or whatever the fuck they wanted to run? Both companies sell Linux and brag about how much money it makes them.
Their vendors would just build the DRM checks into the HD-DVD software or whatever they were trying to protect. Linux just wouldn't get those features, not be "outlawed" or whatever.
Apple already doesn't treat customers like scum the way Microsoft does
Assuming you are talking about Windows copy-protection... Apple has turned a blind eye to OS piracy, because with such a small installed base, it's better for everyone to have everyone on a reasonably current OS.
But recently Apple announced huge revenue from Tiger upgrades. MacOS is becoming a profit center in it's own right and not just paying it's way as it did in the past. As hardware revenues inevitably decline, Apple will look more and more to software like OS X and iLife to keep the Mac profitable. And eventually they will have to consider copy-protection to protect this revenue stream.
Java just makes you do it in a java style. I suppose this is good for someone that can't manage their own succesful style.
Well, I don't know if you were expecting an intelligent response, but yeah, that's kind of the point of Java -- there's One Way To Do Things and patterns and methodologies are standardized. That makes programmers cheaper and more replacable from a management standpoint.
For most business programming, Capital don't want style, especially unique style.
Hmm. My interpretation is that this only means that Apple hasn't finished their TPM implementation yet. The trusted boot features of TPM could prevent replacing the kernel code.
Also, it would be rather embarassing if production OS X/Intel was running partially in emulation. Shades of System 7.
I think there's merit to this theory -- A few years ago, Microsoft seemed to be shitting bricks over a potential AOL/TimeWarner or Sony closed media box. (In fact that was the main conspiracy theory explaining the XBox.)
Movie Executives (a group which Steve Jobs is a member of) are always harping on how "insecure" the PC platform is. This puts MS and Intel in a position of appeasement, where they invent bullshit technologies like encrypted monitor cables so that Hollywood will grace them with HD-DVD software.
Positioning the Mac as a "secure PC" could be a huge tactical advantage for Apple. But ultimately a temporary one, because Apple is relying on commodity parts that were originally designed for the WinTel market.
The move to Intel is all about controlling consumers.
Well, Apple has always had a very firm grip on it's userbase. The move to Intel is more about the future economics of PC hardware (remember Bill Gates' free PC prediction?), because Apple had every bit of control needed on PowerPC.
I expected Apple to hold off on anything that looks like TC until Microsoft could release it first.
It makes a lot more sense for Apple than Microsoft: Total control over hardware & software, no legacy installed base, significant revenue from media rather than just software.
Microsoft will support TPM (or any other chip that people put on mobos), but they don't really have any business reasons to push it. If their hardware customers don't use it, it doesn't really affect them.
You also have to remember that Gecko started development before there was any such thing as "strict" HTML.
Plus, the fact that XHTML is easier to implement is exactly why it shouldn't be done first. Consider the <li> tag. In traditional HTML, it didn't need to be closed. So it would be "easier" and most compatible for your renderer to not require </li> and only check for it when the spec requires that you be strict. If you assume that </li> will be present, you end up writing two different logic paths (for with and without).
Huh? The thing is ancient, recieved a slapdash, contractually-obligated upgrade for OS X, and that's it. It was widely rumored that MS had given up on IE/Mac as early as 2000. When software goes for 5 years without a major upgrade, it a bit of a clue that nobody's working on it.
MS is Apple's largest ISV, and most evidence indicates that Mac Unit works fairly closely with Apple.
MS was fairly tolerant of Apple while they didn't represent a threat
For the most part, Apple just isn't a threat to MS's business. I'm sorry if the truth hurts.
Quirksmode-first is a horribly stupid decision from a software design standpoint
The biggest problem in software design is designing something that people actually want to use. A web browser that can't browse the web would be a problem.
Also, one could consider XHTML as mostly a subset of Quirky HTML. so I could see why browsers treat it as the special case, rather than visa-versa.
And while a strict browser could be faster, Mozilla and every other major browser are designed for quirky HTML first, and XHTML second. That means it's likely largely the same rendering path, with additional code on top to enforce XHTML rules. XHTML may have some great machine-parsing applications, but as of right now, I'm not sure if Web Browsers are one of them.
As for whether a DOCTYPE is worth serving ugly pages to 70-80% of your audience... I think the bottom line in our discussion is that while W3C specs are nice, it's important to understand realwold behavior.
Slashdot Mirror
> it takes your programmers ten times as long to do as it does in Java or VB or C# or whatever.
:)
All the web programmers out there are trying to figure out why this is a bad thing. If the business people want more expensive development, more jobs for us!
(And seriously, I think you underestimate the real costs of deployment and support. In a large environment it can be ridiclously expensive to solve a client problem. And with many apps living on the "extranet", web is the only way to go.)
What I am arguing against is the idea that AJAX makes webapps powerfull enough to drive all kinds of desktop app extinct, or pose a threat to Microsoft.
It doesn't pose any additional threat to Microsoft, because web applciations have dominated over most traditional Windows programming for years. You're talking about hte future, but this is all history now. The market for VB Forms (which was once something like 50% of all programming jobs) or MFC is basically dead.
Microsoft survived and prospered because they stayed ahead of the game in web application framewoks, and because their browser was significantly better than the competition for client-scripting and extendablity. The revenue from Windows Server licences has more than covered the loss of Win32-lockin.
Different JVMs should be more compatible than different Web browsers currently are.
Key word being "should". In practice, the transition from MS/NS Java 1.1 to Sun Java 1.2 broke a fuckload of applets (solution was usually to dump Java). You also hear of companies having to maintain three or four different JVMs on all their workstations due to specific Java app requirements.
In terms of Javascript, "Version 6 and above" for IE and Mozilla has been a stable platform for years now.
Microsoft didn't invent it. Nutscrape was encouraging people to do the same thing with Java Applets back when Communicator 4 was released. Before MS had XMLHttpRequest, they had their own applet.
As the other poster mentioned, the Hidden Frame trick also dates back to the v4 browser era.
As someone who had been doing "AJAX" sites since about 1998, I have to agree. The "A" stands for Asynchronous -- a browser's javascript environment is multithreaded -- when you click a button three times, you effectively fire off three different threads that run in parallel. Oh, and you have no effective way of synchronizing them.
The bottom line is that it's very easy to write something that works fine against localhost or your development server but blows up when used under realworld network and server conditions. If you are doing something complicated, you need to expect that the page state may be inconsistent at any time, and code around it.
(Note that I haven't tried any of the new AJAX frameworks, so they may have a solution for this.)
Except these sites weren't like MySpace, CDBaby, or Google -- they were mostly just lame clones of Amazon.com.
I think everyone involved knew these commercials were a long shot at best, and their real purpose was to advertise the IPOs and not the sites. The goal was to flush as much investor money down the toliet as fast as possible, and hope everyone could cash out before the house of cards collapsed. Nobody was trying to build a successful business.
I will agree that Apple went to Intel for primarily financial and business reasons rather than the technical merits/demerits of PPC v Cell v x86.
It is hilarious(and sad) to think of all that hardwork Apple engineers went through to be ready for media processing monster chips like Cell
The problem with this argument is that only a small percentage of Apple's business is related to high-end, specialized media-processing. Apple primarily markets general purpose computers to consumers. And even then, Intel provides enough raw oophf to handle most media tasks with ease.
It was funny how the first comments about the Intel Macs was how fast Safari was. The fact is that competitive Integer performance much more important to the general user experience than obscure encoding tasks.
They went your route with the G4 -- you ended up with a computer that was fast for playing videos and 3 photoshop filters, and dog slow 99% of the time. Not to mention it was much more complex and expensive to program for. The bottom line is that Apple doesn't have the market or developer support for something like Cell. I'm sure the 12 people who buy Sony's machines will enjoy them.
> It's turning into same mistake Apple made back in the day
(Aside from all the hystrical business-political rhetoric on this story)
It's possible it could be the same technical limitations that Apple is seeing right now. They are using OpenGL for desktop effects, and consequentally the game performance is teh suck.
Granted, that's horrible HTML...
It's also a fair example, because Word-HTML can "round-trip" back to Word with no loss in fidelity. A barebones HTML file can not.
What they could do in a reasonable timeframe is give up on IE as the product, port the KHTML renderer over to Windows, and give it a new name.
It would take years to add the gazillion proprietary IE features to KHTML or any other browser code.
If what was reported yesterday is true, and IE7 will support 99% of CSS 2.1, then full ACID2 compliance is really not that much additional work.
Hey Transhuman -- Doubleclick cookies and the like are reported by Spyware apps, but aren't really spyware. If you look in Firefox, you'll have all the same cookies.
Now imagine a world where Microsoft requires a locked TCPA chip to boot a future version of Windows.
You mean like Apple is doing? In both cases the TCPA chip is a copy-protection dongle and does nothing to prevent you from running an alternate OS.
Are people angry at this? And if so, how has it hurt Microsoft? I think history has shown that treating the entire computer as a dongle had a far greater affect on consumers than copy-protection.
The hardware chip is doing public-private key encryption. If you can mimic this, there's much more profitable endevors available to you than getting OS X running on your Dell.
And if not outright outlaw Linux, at least make sure only generic whitebox motherboards from Taiwan run it. The Dell and HPs will all be locked to the copy of Windows married to their TCPA module during manufacturing.
This silly conspiracy theory is getting tiring. Why would Dell & HP prevent paying customers from running Linux or DOS or whatever the fuck they wanted to run? Both companies sell Linux and brag about how much money it makes them.
Their vendors would just build the DRM checks into the HD-DVD software or whatever they were trying to protect. Linux just wouldn't get those features, not be "outlawed" or whatever.
Apple already doesn't treat customers like scum the way Microsoft does
Assuming you are talking about Windows copy-protection... Apple has turned a blind eye to OS piracy, because with such a small installed base, it's better for everyone to have everyone on a reasonably current OS.
But recently Apple announced huge revenue from Tiger upgrades. MacOS is becoming a profit center in it's own right and not just paying it's way as it did in the past. As hardware revenues inevitably decline, Apple will look more and more to software like OS X and iLife to keep the Mac profitable. And eventually they will have to consider copy-protection to protect this revenue stream.
Java just makes you do it in a java style. I suppose this is good for someone that can't manage their own succesful style.
Well, I don't know if you were expecting an intelligent response, but yeah, that's kind of the point of Java -- there's One Way To Do Things and patterns and methodologies are standardized. That makes programmers cheaper and more replacable from a management standpoint.
For most business programming, Capital don't want style, especially unique style.
Hmm. My interpretation is that this only means that Apple hasn't finished their TPM implementation yet. The trusted boot features of TPM could prevent replacing the kernel code.
Also, it would be rather embarassing if production OS X/Intel was running partially in emulation. Shades of System 7.
I think there's merit to this theory -- A few years ago, Microsoft seemed to be shitting bricks over a potential AOL/TimeWarner or Sony closed media box. (In fact that was the main conspiracy theory explaining the XBox.)
Movie Executives (a group which Steve Jobs is a member of) are always harping on how "insecure" the PC platform is. This puts MS and Intel in a position of appeasement, where they invent bullshit technologies like encrypted monitor cables so that Hollywood will grace them with HD-DVD software.
Positioning the Mac as a "secure PC" could be a huge tactical advantage for Apple. But ultimately a temporary one, because Apple is relying on commodity parts that were originally designed for the WinTel market.
The move to Intel is all about controlling consumers.
Well, Apple has always had a very firm grip on it's userbase. The move to Intel is more about the future economics of PC hardware (remember Bill Gates' free PC prediction?), because Apple had every bit of control needed on PowerPC.
I expected Apple to hold off on anything that looks like TC until Microsoft could release it first.
It makes a lot more sense for Apple than Microsoft: Total control over hardware & software, no legacy installed base, significant revenue from media rather than just software.
Microsoft will support TPM (or any other chip that people put on mobos), but they don't really have any business reasons to push it. If their hardware customers don't use it, it doesn't really affect them.
You also have to remember that Gecko started development before there was any such thing as "strict" HTML.
Plus, the fact that XHTML is easier to implement is exactly why it shouldn't be done first. Consider the <li> tag. In traditional HTML, it didn't need to be closed. So it would be "easier" and most compatible for your renderer to not require </li> and only check for it when the spec requires that you be strict. If you assume that </li> will be present, you end up writing two different logic paths (for with and without).
Not really. MS showed no sign of slowing IE down
Huh? The thing is ancient, recieved a slapdash, contractually-obligated upgrade for OS X, and that's it. It was widely rumored that MS had given up on IE/Mac as early as 2000. When software goes for 5 years without a major upgrade, it a bit of a clue that nobody's working on it.
MS is Apple's largest ISV, and most evidence indicates that Mac Unit works fairly closely with Apple.
MS was fairly tolerant of Apple while they didn't represent a threat
For the most part, Apple just isn't a threat to MS's business. I'm sorry if the truth hurts.
Quirksmode-first is a horribly stupid decision from a software design standpoint
The biggest problem in software design is designing something that people actually want to use. A web browser that can't browse the web would be a problem.
Also, one could consider XHTML as mostly a subset of Quirky HTML. so I could see why browsers treat it as the special case, rather than visa-versa.
Don't you think it's a lot more likely that Apple didn't start on Safari until after Microsoft told them they were giving up on IE/Mac?
Regardless of how it could work, the facts are serving application/xhtml+xml has a diminished user experience for Mozilla/Firefox users. http://www.mozilla.org/docs/web-developer/faq.html #accept
... I think the bottom line in our discussion is that while W3C specs are nice, it's important to understand realwold behavior.
And while a strict browser could be faster, Mozilla and every other major browser are designed for quirky HTML first, and XHTML second. That means it's likely largely the same rendering path, with additional code on top to enforce XHTML rules. XHTML may have some great machine-parsing applications, but as of right now, I'm not sure if Web Browsers are one of them.
As for whether a DOCTYPE is worth serving ugly pages to 70-80% of your audience