Calling Jonathan Schwartz…

Matt Mullenweg (the guy who started Automattic, which produces WordPress, which runs this blog and many others) tells Jonathan Schwartz that Sun Microsystems isn’t there for startups.

It’ll be interesting to see whether Sun can improve its approach to startups. LAMP is sure getting traction — I’ve stopped asking entrepreneurs what infrastructure they are using since the answer was so consistently LAMP.

Sun has done a great job of stopping the bleeding and getting some interesting new products out (Black Box, for instance) but for Sun to dramatically increase its relevance it needs to convince Steve Jobs to put Java on his new products and needs to build something that really shakes up the Web 2.0 industry.

That’s why I asked Jonathan what Sun’s iPhone was. I notice Jonathan didn’t have a good answer to that question.

Oh, and someone asked why would Steve Jobs put Java on the iPhone? My answer: it’s dramatically easier to write applications for a runtime-compiler system like Java or .NET. Because Steve Jobs refuses to go with Java (putting .NET on iPhone doesn’t make sense for Apple’s business) it means that developers on Apple’s iPhone will need to work a lot harder to get apps out. Translation: fewer apps.

I remember back in 1989 that Apple was six years ahead of the industry with its Macintosh II. Remember, it took until late 1995 until Microsoft was able to match the innovations Apple shipped in 1989.

So, why did Apple only end up with a small sliver of market share? Developers, developers, developers.

Oh, sorry, I’m channelling Steve Ballmer again.

Aside: I’m posting this from the Golden Gate bridge. I love Verizon Wireless!!!

Comments

  1. I am doing a startup right now, and we spent about 2 months fighting over which language we were going to use: Java | PHP | Ruby on Rails. In the end, we chose Java, simply because it was a best fit for the product we are building. That being said, I still don’t like the whole “Java” mentality — I think it crushes a lot of creativity. Just listen to some of Paul Graham’s speeches on this.

    Note to your “aside”: have you ever been to Baker Beach (right near Golden Gate Bridge). It’s got to be one of the nicest beaches in Northern California, especially if you “sneak in” after it’s closed (7/8pm is when I think it closes). It’s always fun when I go there.

  2. I am doing a startup right now, and we spent about 2 months fighting over which language we were going to use: Java | PHP | Ruby on Rails. In the end, we chose Java, simply because it was a best fit for the product we are building. That being said, I still don’t like the whole “Java” mentality — I think it crushes a lot of creativity. Just listen to some of Paul Graham’s speeches on this.

    Note to your “aside”: have you ever been to Baker Beach (right near Golden Gate Bridge). It’s got to be one of the nicest beaches in Northern California, especially if you “sneak in” after it’s closed (7/8pm is when I think it closes). It’s always fun when I go there.

  3. Robert,

    “Developers, developers, developers” is a good short-hand for why the Mac had such a small market share. But it’s not the root cause.

    The game was over long before 1995. It wasn’t Windows that beat Mac in the marketplace, it was DOS! Mac couldn’t even beat MS-DOS.

    It’s one of those classic chicken-and-egg problems. Developers often found it hard to fund Mac apps because the installed base was so small. It’s circular.

    Here are some of the other factors:

    The IBM PC, which came out in 1981 with MS-DOS, established itself as a useful business machine as early as 1983 with the introduction of the IBM-XT (with a 10 MB hard drive) and its first killer app, Lotus 1-2-3.

    When Mac came out in 1984. It was pretty, but it was more a proof of concept than a computer you could actually use in business. No hard drive. 128K RAM, Very few apps.

    Mac didn’t begin to be competitive until 1985 when it became possible to add a hard disk drive, Microsoft released Word and Excel for Mac, the LaserWriter and Aldous Pagemaker appeared.

    The PC side of the world was then well-established with a much broader eco-system and lower prices. As far as general business, it was game over.

    In practical terms, the Mac wasn’t ahead of the PC, it trailed badly.

    Mac got interesting again in 1987, but by then it was niched into the graphics and creative world.

    I agree with you that the fact that Mac was harder to develop for was an important factor, but if the basic hardware had been more competitive earlier, there would have been a payoff that would have justified the higher development costs.

    Steve’s bet on the iPhone is that it will have enough functionality to appeal to a mass audience out of the gate and that, unlike a computer world, 3rd party apps will not be required to seal that appeal.

    That’s going to turn some people off. But others won’t see a problem. If Steve is right then he gets to control all the add-on apps. (He does love control, doesn’t he.) If he’s wrong, he can always change the software and open up the iPhone to development.

  4. Robert,

    “Developers, developers, developers” is a good short-hand for why the Mac had such a small market share. But it’s not the root cause.

    The game was over long before 1995. It wasn’t Windows that beat Mac in the marketplace, it was DOS! Mac couldn’t even beat MS-DOS.

    It’s one of those classic chicken-and-egg problems. Developers often found it hard to fund Mac apps because the installed base was so small. It’s circular.

    Here are some of the other factors:

    The IBM PC, which came out in 1981 with MS-DOS, established itself as a useful business machine as early as 1983 with the introduction of the IBM-XT (with a 10 MB hard drive) and its first killer app, Lotus 1-2-3.

    When Mac came out in 1984. It was pretty, but it was more a proof of concept than a computer you could actually use in business. No hard drive. 128K RAM, Very few apps.

    Mac didn’t begin to be competitive until 1985 when it became possible to add a hard disk drive, Microsoft released Word and Excel for Mac, the LaserWriter and Aldous Pagemaker appeared.

    The PC side of the world was then well-established with a much broader eco-system and lower prices. As far as general business, it was game over.

    In practical terms, the Mac wasn’t ahead of the PC, it trailed badly.

    Mac got interesting again in 1987, but by then it was niched into the graphics and creative world.

    I agree with you that the fact that Mac was harder to develop for was an important factor, but if the basic hardware had been more competitive earlier, there would have been a payoff that would have justified the higher development costs.

    Steve’s bet on the iPhone is that it will have enough functionality to appeal to a mass audience out of the gate and that, unlike a computer world, 3rd party apps will not be required to seal that appeal.

    That’s going to turn some people off. But others won’t see a problem. If Steve is right then he gets to control all the add-on apps. (He does love control, doesn’t he.) If he’s wrong, he can always change the software and open up the iPhone to development.

  5. “…developers on Apple’s iPhone will need to work a lot harder to get apps out.”

    I actually believe that the opposite of that statement is true. (I realize that iPhone is currently closed to third-parties. But, let’s assume that Apple opens it up eventually.)

    Ask any developer who’s taken the time to learn and understand Objective-C, and they’ll tell you how easy it will be to develop for the iPhone. The learning curve is gradual, so the bootstrapping time can be safely disregarded. Once a developer gets into the “flow” of Objective-C, it is pure bliss.

    Objective-C is a message-passing based language. It derives that behavior from Smalltalk, as does Ruby. These languages presently have a lot of momentum, as developers are rediscovering the benefits they present.

    Java, however, while widely used, seems to be falling out of favor among leading-edge developers. Ask these people, and often they will site the *pain* of having to develop in Java.

    I actually suspect Apple is aware of this. If iPhone follows in iPod’s footsteps to become a phenomenal success, and Apple opens up the platform, Objective-C will be exposed to a huge audience. Percentage wise, it will dwarf other APIs, including Java *and* Win32.

    The iPhone may just be Apple’s Trojan horse, that once given entry, makes Apple the dominant provider of development platforms.

  6. “…developers on Apple’s iPhone will need to work a lot harder to get apps out.”

    I actually believe that the opposite of that statement is true. (I realize that iPhone is currently closed to third-parties. But, let’s assume that Apple opens it up eventually.)

    Ask any developer who’s taken the time to learn and understand Objective-C, and they’ll tell you how easy it will be to develop for the iPhone. The learning curve is gradual, so the bootstrapping time can be safely disregarded. Once a developer gets into the “flow” of Objective-C, it is pure bliss.

    Objective-C is a message-passing based language. It derives that behavior from Smalltalk, as does Ruby. These languages presently have a lot of momentum, as developers are rediscovering the benefits they present.

    Java, however, while widely used, seems to be falling out of favor among leading-edge developers. Ask these people, and often they will site the *pain* of having to develop in Java.

    I actually suspect Apple is aware of this. If iPhone follows in iPod’s footsteps to become a phenomenal success, and Apple opens up the platform, Objective-C will be exposed to a huge audience. Percentage wise, it will dwarf other APIs, including Java *and* Win32.

    The iPhone may just be Apple’s Trojan horse, that once given entry, makes Apple the dominant provider of development platforms.

  7. In my opinion, the root cause of Apple failure was not the lack of developers. There were fanatics both from developer and user site. They then got frustated of Apple arrogance.

    Apple tried to fix their mistakes years later by migrating its processor to Intel. Something that still in question whether it’s profitable or not. Or in favor of its fans or admirers.

    Apple will no longer be remembered as computer maker or innovator. It is its gadgets that make Apple survives and makes profits.

  8. In my opinion, the root cause of Apple failure was not the lack of developers. There were fanatics both from developer and user site. They then got frustated of Apple arrogance.

    Apple tried to fix their mistakes years later by migrating its processor to Intel. Something that still in question whether it’s profitable or not. Or in favor of its fans or admirers.

    Apple will no longer be remembered as computer maker or innovator. It is its gadgets that make Apple survives and makes profits.

  9. Robert,

    My company hired me to work on a huge Java based server; the original developers swore that it was wonderful and would fix all their problems. After 3 months, I threw the whole mess (50k lines of java) out and replaced it with about 10k lines of C++ – runs orders of magnitude faster and without any memory leaks. (Yes, you *can* and *do* leak memory in Java – typically with the result that the entire server goes down, rather than just paging…)

    Java is not a panacea, and can actually be an impediment to delivering products. Any language that requires specialized tools (beyond a compiler and loader) is going to end up costing more to develop on. At least C is not yet owned by a single entity. (And objective C is a trivially simple extension to the language…)

  10. Robert,

    My company hired me to work on a huge Java based server; the original developers swore that it was wonderful and would fix all their problems. After 3 months, I threw the whole mess (50k lines of java) out and replaced it with about 10k lines of C++ – runs orders of magnitude faster and without any memory leaks. (Yes, you *can* and *do* leak memory in Java – typically with the result that the entire server goes down, rather than just paging…)

    Java is not a panacea, and can actually be an impediment to delivering products. Any language that requires specialized tools (beyond a compiler and loader) is going to end up costing more to develop on. At least C is not yet owned by a single entity. (And objective C is a trivially simple extension to the language…)

  11. As far as I can understand, the support for Java in OS X is no less than in any other operating system. It should therefore be possible to run Java apps on the iPhone, if Apple want to. OS X on the other hand is mostly written in Objective-C and C, using the Next Step framework. The reason for this is probably long and historic and has to do with Jobs’ tour at Next Computer. The telco business, on the other hand, has always been a walled garden, which are a blessing and a curse at the same time. If you come from the computer world, you don’t always understand that, but it has a lot to do with reliability, communication in real-time, and how business models work in that world.

  12. As far as I can understand, the support for Java in OS X is no less than in any other operating system. It should therefore be possible to run Java apps on the iPhone, if Apple want to. OS X on the other hand is mostly written in Objective-C and C, using the Next Step framework. The reason for this is probably long and historic and has to do with Jobs’ tour at Next Computer. The telco business, on the other hand, has always been a walled garden, which are a blessing and a curse at the same time. If you come from the computer world, you don’t always understand that, but it has a lot to do with reliability, communication in real-time, and how business models work in that world.

  13. Jan – it’s pretty unlikely that iPhone is running Mac OS X. Rather, it’s more likely to be running a new embedded OS that Apple is calling “OS X”. So, it’s not a matter of saying – “if it runs under Mac OS X, it will run on iPhone.”

    It would be unlikely that Apple would run Java SE on this device (the iPhone would be *truly* amazing if they did that). Rather, the decision for Apple is more likely to be whether to run a flavour of Java ME. The CDC (as opposed to CLDC) flavour would make most sense for iPhone.

  14. Jan – it’s pretty unlikely that iPhone is running Mac OS X. Rather, it’s more likely to be running a new embedded OS that Apple is calling “OS X”. So, it’s not a matter of saying – “if it runs under Mac OS X, it will run on iPhone.”

    It would be unlikely that Apple would run Java SE on this device (the iPhone would be *truly* amazing if they did that). Rather, the decision for Apple is more likely to be whether to run a flavour of Java ME. The CDC (as opposed to CLDC) flavour would make most sense for iPhone.

  15. “Oh, and someone asked why would Steve Jobs put Java on the iPhone? My answer: it’s dramatically easier to write applications for a runtime-compiler system like Java or .NET. Because Steve Jobs refuses to go with Java (putting .NET on iPhone doesn’t make sense for Apple’s business) it means that developers on Apple’s iPhone will need to work a lot harder to get apps out. Translation: fewer apps.”

    How does .Net get on the iPhone? Scoble, you don’t work for Microsoft anymore, no need for the evangelism of a product option that is not even an option. You just make yourself sound like you have no freaking clue.

    As for Java: how is it easier? Few will say it’s easier. It may be more widespread, but is that an issue? According to you and everyone else, everyone is crawling up Apple’s @ss to get on the iPhone. Why would Apple need to choose the Least Common Denominator option?

    And then how do you address the UI? Java apps would kill the very thing that you and others are allegedly so excited about. Yay, I get this great new iPhone and made it look like a WinMobile crap device!

    “In my opinion, the root cause of Apple failure was not the lack of developers. There were fanatics both from developer and user site. They then got frustated of Apple arrogance.”

    That’s hardly the case. Word, Excel, PowerPoint, all Adobe products started on the Apple products. It wasn’t Apple’s arrogance, it was their lack of marketshare, plain and simple.

  16. “Oh, and someone asked why would Steve Jobs put Java on the iPhone? My answer: it’s dramatically easier to write applications for a runtime-compiler system like Java or .NET. Because Steve Jobs refuses to go with Java (putting .NET on iPhone doesn’t make sense for Apple’s business) it means that developers on Apple’s iPhone will need to work a lot harder to get apps out. Translation: fewer apps.”

    How does .Net get on the iPhone? Scoble, you don’t work for Microsoft anymore, no need for the evangelism of a product option that is not even an option. You just make yourself sound like you have no freaking clue.

    As for Java: how is it easier? Few will say it’s easier. It may be more widespread, but is that an issue? According to you and everyone else, everyone is crawling up Apple’s @ss to get on the iPhone. Why would Apple need to choose the Least Common Denominator option?

    And then how do you address the UI? Java apps would kill the very thing that you and others are allegedly so excited about. Yay, I get this great new iPhone and made it look like a WinMobile crap device!

    “In my opinion, the root cause of Apple failure was not the lack of developers. There were fanatics both from developer and user site. They then got frustated of Apple arrogance.”

    That’s hardly the case. Word, Excel, PowerPoint, all Adobe products started on the Apple products. It wasn’t Apple’s arrogance, it was their lack of marketshare, plain and simple.

  17. Some interesting comments here….

    Exactly what is it that makes OS X “OS X”? Is it FreeBSD? Cocoa? If it is, than whatever the differences in the OS embedded in iPhone, it is OS X.

    Back on topic, what makes the iPhone an “iPhone”? Is it the UI? The experience? If so, than name me any Java app currently running on OS X that duplicates the experience of using a Cocoa app.

    Robert, there’s good reason for the lack of Java on the iPhone. It lacks that experience factor that Apple does so well in consumer products.

    You’ve already mentioned how the touchscreen will prevent people from using it under a table in business meetings. That should have clued you in to something you just don’t get… the target audience here isn’t business.

  18. Some interesting comments here….

    Exactly what is it that makes OS X “OS X”? Is it FreeBSD? Cocoa? If it is, than whatever the differences in the OS embedded in iPhone, it is OS X.

    Back on topic, what makes the iPhone an “iPhone”? Is it the UI? The experience? If so, than name me any Java app currently running on OS X that duplicates the experience of using a Cocoa app.

    Robert, there’s good reason for the lack of Java on the iPhone. It lacks that experience factor that Apple does so well in consumer products.

    You’ve already mentioned how the touchscreen will prevent people from using it under a table in business meetings. That should have clued you in to something you just don’t get… the target audience here isn’t business.

  19. Excellent post, Michael Markman. It wasn’t Apple’s arrogance or lack of openness. Even if it was more difficult to program for (or if Apple was more arrogant), developers prefered and more often introduced truly revolutionary software on Apple products. But Apple, in transitioning from the Apple line when PCs were more educational tools to the Mac when computers became personal and business tools, had already chosen to be a niche: they focused on the personal and creative. Business could be 10 years behind (in terms of UI and some other factors) because of the application needs were more basic, but broader, and less user-friendly (people were trained in the business setting to use specialized apps). This created a larger ecosystem and userbase, and it took many decades for personal use to catch up to business use as a market. In fact, it still hasn’t fully happened. But it may finally be happening which is why Sun, despite improved financials, is still in trouble and Microsoft can’t really pull themselves properly together in key markets using their long-term, tried-and-true strategies.

  20. Excellent post, Michael Markman. It wasn’t Apple’s arrogance or lack of openness. Even if it was more difficult to program for (or if Apple was more arrogant), developers prefered and more often introduced truly revolutionary software on Apple products. But Apple, in transitioning from the Apple line when PCs were more educational tools to the Mac when computers became personal and business tools, had already chosen to be a niche: they focused on the personal and creative. Business could be 10 years behind (in terms of UI and some other factors) because of the application needs were more basic, but broader, and less user-friendly (people were trained in the business setting to use specialized apps). This created a larger ecosystem and userbase, and it took many decades for personal use to catch up to business use as a market. In fact, it still hasn’t fully happened. But it may finally be happening which is why Sun, despite improved financials, is still in trouble and Microsoft can’t really pull themselves properly together in key markets using their long-term, tried-and-true strategies.

  21. “for Sun to dramatically increase its relevance it needs to convince Steve Jobs to put Java on his new products”

    That doesn’t make any sense. Java on the iPhone (it’s already on all Macs) won’t increase its relevance. You’re talking about a product that even Jobs is only hoping can capture a tiny share of the market.

    Java is huge for mobile development. Why do you think Google’s mobile apps (Maps and GMail) are both in Java? What about the Opera mini-browser? It’s in Java. The only phones that don’t include a J2ME runtime are BREW phones. Even the Windows Mobile phones have it.

    That’s J2ME. Most of the post is about Java for startups, so I presume that means Java for web development. The reason that most startups these days go for LAMP is because it is much cheaper. Not cheaper as in hardware/software costs cheaper, but cheaper because the developers are much cheaper. There is heavy competition for Java developers from large companies (anybody from Google to Chevron, you name it) thus the salaries of Java developers are much higher. If you’re a startup, you can afford five PHP guys or two-three Java guys. Ruby is popular among Java developers, thus the cost of Ruby developers tends to be higher than PHP devs, especially since there are fewer of them around. It’s the higher demand for Java developers that causes cash-strapped startups to pursue LAMP. Or at least pursue the P in LAMP. The LAM part of LAMP has nothing to do with Java.

  22. “for Sun to dramatically increase its relevance it needs to convince Steve Jobs to put Java on his new products”

    That doesn’t make any sense. Java on the iPhone (it’s already on all Macs) won’t increase its relevance. You’re talking about a product that even Jobs is only hoping can capture a tiny share of the market.

    Java is huge for mobile development. Why do you think Google’s mobile apps (Maps and GMail) are both in Java? What about the Opera mini-browser? It’s in Java. The only phones that don’t include a J2ME runtime are BREW phones. Even the Windows Mobile phones have it.

    That’s J2ME. Most of the post is about Java for startups, so I presume that means Java for web development. The reason that most startups these days go for LAMP is because it is much cheaper. Not cheaper as in hardware/software costs cheaper, but cheaper because the developers are much cheaper. There is heavy competition for Java developers from large companies (anybody from Google to Chevron, you name it) thus the salaries of Java developers are much higher. If you’re a startup, you can afford five PHP guys or two-three Java guys. Ruby is popular among Java developers, thus the cost of Ruby developers tends to be higher than PHP devs, especially since there are fewer of them around. It’s the higher demand for Java developers that causes cash-strapped startups to pursue LAMP. Or at least pursue the P in LAMP. The LAM part of LAMP has nothing to do with Java.

  23. Michael, are you confusing Java with Javascript (Ajax)?

    Scoble, the Java runtime environment is big (relatively) has a poor default UI, and lends itself to programming ugly GUIs. Not really a good choice for Apple’s user-experience focus.

    Popularity of LAMP? Not just programmer availability, but the fact that some of the most common, easy to configure out of the box applications run on it without much (any?) tweaking. WordPress. Phpbb. etc etc. LAMP is the new DOS. LAMP is the software expression of the engineering goal of “reducing it to a problem we’ve solved before.”

    If I can already run 90% of my business critical apps with LAMP, on commodity hardware, why should I be interested in tying myself to someone’s proprietary stuff to run my 10% of secret-sauce apps? (And pay licensing fees, etc.) I was kinda hoping that this question would get answered here in the comments.

    The future of app development on the iphone? Hope that it does use some kind of LAMP* stack on it. Then you can run your app locally and use the webbrowser as the UI layer. Users don’t mind the crappiness of browser UI, they’re used to it, and don’t associate it with Apple. (The other benefit is that UI development and testing is pretty transparent in browsers and doesn’t require actual programming knowlege. Just try that with Java.)

    *okay, MAMP. You get the idea. And yeah, a whole MAMP stack would be kinda big, security issues, user access to the source, blah blah blah.

  24. Michael, are you confusing Java with Javascript (Ajax)?

    Scoble, the Java runtime environment is big (relatively) has a poor default UI, and lends itself to programming ugly GUIs. Not really a good choice for Apple’s user-experience focus.

    Popularity of LAMP? Not just programmer availability, but the fact that some of the most common, easy to configure out of the box applications run on it without much (any?) tweaking. WordPress. Phpbb. etc etc. LAMP is the new DOS. LAMP is the software expression of the engineering goal of “reducing it to a problem we’ve solved before.”

    If I can already run 90% of my business critical apps with LAMP, on commodity hardware, why should I be interested in tying myself to someone’s proprietary stuff to run my 10% of secret-sauce apps? (And pay licensing fees, etc.) I was kinda hoping that this question would get answered here in the comments.

    The future of app development on the iphone? Hope that it does use some kind of LAMP* stack on it. Then you can run your app locally and use the webbrowser as the UI layer. Users don’t mind the crappiness of browser UI, they’re used to it, and don’t associate it with Apple. (The other benefit is that UI development and testing is pretty transparent in browsers and doesn’t require actual programming knowlege. Just try that with Java.)

    *okay, MAMP. You get the idea. And yeah, a whole MAMP stack would be kinda big, security issues, user access to the source, blah blah blah.

  25. Hey, Rob, since Google sucks so much these days (forum-ized newsgroups and weird aggregation pages), could you hyperlink your acronyms? It took 7 different page hits to find out what “LAMP” actually stands for (Linux, Apache, MySQL, PHP&Friends); it wasn’t even in the Wiki for Sun.

  26. Hey, Rob, since Google sucks so much these days (forum-ized newsgroups and weird aggregation pages), could you hyperlink your acronyms? It took 7 different page hits to find out what “LAMP” actually stands for (Linux, Apache, MySQL, PHP&Friends); it wasn’t even in the Wiki for Sun.

  27. Keith, why should anyone need to account for your deficiencies? It’s a commonplace acronym in the industry and the simple search for “LAMP acronym” gives you the answer first hit. Don’t know why you had such trouble or were wandering around the wiki for Sun.

  28. Keith, why should anyone need to account for your deficiencies? It’s a commonplace acronym in the industry and the simple search for “LAMP acronym” gives you the answer first hit. Don’t know why you had such trouble or were wandering around the wiki for Sun.

  29. What did you mean by “Sun’s iPhone”? Do you mean: “Not to be released in six months, overhyped, expensive, limited number of apps, and already under lawsuit attack”?

    I think Sun has gotten away from preannouncing products six months in advance.

    My personal guess is Sun’s iPhone will be either the Honeycomb storage system, or perhaps a follow-on NAS box using Solaris and ZFS, or it will be the Niagara 2 systems.

    At least Sun is being asked this. What is IBM’s, Dell’s or HP’s “iPhone”?

  30. What did you mean by “Sun’s iPhone”? Do you mean: “Not to be released in six months, overhyped, expensive, limited number of apps, and already under lawsuit attack”?

    I think Sun has gotten away from preannouncing products six months in advance.

    My personal guess is Sun’s iPhone will be either the Honeycomb storage system, or perhaps a follow-on NAS box using Solaris and ZFS, or it will be the Niagara 2 systems.

    At least Sun is being asked this. What is IBM’s, Dell’s or HP’s “iPhone”?

  31. Sun isn’t relevant anymore because they are following in the footsteps of SGI — both are relics of an era that relied on CIO/CTO fear-uncertainty-doubt to force large companies into thinking they needed this type of hardware to do anything relevant. Now you’ve got Google running thousands of the cheapest boxes they could find and it all works because of the custom software file system they put together that holds the Internet in RAM. Expensive, proprietary *hardware* is absolutely worthless for a server environment because it is now the software that can make or break your server architecture. Sun created its business by selling the whole widget (expensive Sun servers + Solaris) and now that half that widget is irrelevant they have to reinvent themselves or they’re already halfway in the grave.

    Java has no business on the iPhone because the device is meant to be the future of mobile devices, not the past. I can write a Dashboard Widget for Mac OS X that calls native Objective-C code straight from the JavaScript, enabling a world of functional possibilities, all without the need for a middleware compiler. Java is a slow, bulky, resource hog that has ancient UI hooks built throughout. Try running Eclipse for Mac OS X and then switch over to Apple xCode and tell me the difference isn’t night and day.

    The fundamental problem with Java is that it’s built for cross-platform compatibility and easy development, for developers and devices of the lowest common denominator. Jobs’ iPhone is built to be a world-class interface linked to smartly-built and intuitive applications that no one in the mobile world has gotten “right” yet. The iPhone is all about elegance, minimalism, and beauty, and Java applications are the antithesis of those three.

  32. Sun isn’t relevant anymore because they are following in the footsteps of SGI — both are relics of an era that relied on CIO/CTO fear-uncertainty-doubt to force large companies into thinking they needed this type of hardware to do anything relevant. Now you’ve got Google running thousands of the cheapest boxes they could find and it all works because of the custom software file system they put together that holds the Internet in RAM. Expensive, proprietary *hardware* is absolutely worthless for a server environment because it is now the software that can make or break your server architecture. Sun created its business by selling the whole widget (expensive Sun servers + Solaris) and now that half that widget is irrelevant they have to reinvent themselves or they’re already halfway in the grave.

    Java has no business on the iPhone because the device is meant to be the future of mobile devices, not the past. I can write a Dashboard Widget for Mac OS X that calls native Objective-C code straight from the JavaScript, enabling a world of functional possibilities, all without the need for a middleware compiler. Java is a slow, bulky, resource hog that has ancient UI hooks built throughout. Try running Eclipse for Mac OS X and then switch over to Apple xCode and tell me the difference isn’t night and day.

    The fundamental problem with Java is that it’s built for cross-platform compatibility and easy development, for developers and devices of the lowest common denominator. Jobs’ iPhone is built to be a world-class interface linked to smartly-built and intuitive applications that no one in the mobile world has gotten “right” yet. The iPhone is all about elegance, minimalism, and beauty, and Java applications are the antithesis of those three.

  33. This is all fairly irrelevant right now… not just because the iPhone is vaporware… but because Apple has indicated that the iPhone is a closed platform that will not be open to third-party developers. All applications will eiher be developed by Apple, or under strict control of Apple. Maybe someone else will have a cool, elegant phone with a world-class interface that will actually be an open platform. Or maybe Apple will have an epiphany and open things up.

    PS: if you think MS has been harsh to developers, you’ve never dealt with Apple.

    PPS: here’s a little limerick I wrote for the iPhone introduction… http://fredtime.com/?p=6

  34. This is all fairly irrelevant right now… not just because the iPhone is vaporware… but because Apple has indicated that the iPhone is a closed platform that will not be open to third-party developers. All applications will eiher be developed by Apple, or under strict control of Apple. Maybe someone else will have a cool, elegant phone with a world-class interface that will actually be an open platform. Or maybe Apple will have an epiphany and open things up.

    PS: if you think MS has been harsh to developers, you’ve never dealt with Apple.

    PPS: here’s a little limerick I wrote for the iPhone introduction… http://fredtime.com/?p=6

  35. @13. Thanks Goebbels. Apple did get niched, but I disagree that they chose to do so.

    The current “I’m a Mac. I’m a PC.” campaign, is the first campaign to actually celebrate the niche. The new strategy is to position PC as good (only) for business while Mac is good for creativity and individuals.

    But when I was at Apple (1986-1985) we did everything possible to climbe out of the niche and gain broad acceptance in business: we hired sales and marketing folks from IBM, Xerox, AT&T etc; brought in a guy who ran a Fortune 500 IT department and made him president Apple USA; made huge investments in connectivity and networking; even forbade the folks who made ads and brochures to use anything but business screen shots in the marketing materials. During that time, they ran one “switch” campaign after another. But until the current campaign, it was always some flavor of “Mac is better than a PC.”

    The current campaign has evolved to say “Mac is different from a PC.” It even goes so far as to cast PC (John Hodgman) as a basicaly likeable character, while Mac (Justin Long) is a pretty irritating twerp.

  36. @13. Thanks Goebbels. Apple did get niched, but I disagree that they chose to do so.

    The current “I’m a Mac. I’m a PC.” campaign, is the first campaign to actually celebrate the niche. The new strategy is to position PC as good (only) for business while Mac is good for creativity and individuals.

    But when I was at Apple (1986-1985) we did everything possible to climbe out of the niche and gain broad acceptance in business: we hired sales and marketing folks from IBM, Xerox, AT&T etc; brought in a guy who ran a Fortune 500 IT department and made him president Apple USA; made huge investments in connectivity and networking; even forbade the folks who made ads and brochures to use anything but business screen shots in the marketing materials. During that time, they ran one “switch” campaign after another. But until the current campaign, it was always some flavor of “Mac is better than a PC.”

    The current campaign has evolved to say “Mac is different from a PC.” It even goes so far as to cast PC (John Hodgman) as a basicaly likeable character, while Mac (Justin Long) is a pretty irritating twerp.

  37. I’m actually really surprised that people keep harping on the whole ‘iPhone doesn’t support java’ thing. For me, the big development platform for the iPhone is Widgets. The iPhone comes with safari, supports widgets and is running osx (or some derivation of it). Thus, I would imagine that just about every widget that currently runs on OSX will run on the iPhone. That means that developers will be able to develop widget based applications now, for OSX, that will migrate easily to the new platform when it comes out and will work on the existing OS.

    Desktop app and mobile app from the same code base with no migration and no need for emulation software at dev time? That sounds pretty cool to me.

  38. I’m actually really surprised that people keep harping on the whole ‘iPhone doesn’t support java’ thing. For me, the big development platform for the iPhone is Widgets. The iPhone comes with safari, supports widgets and is running osx (or some derivation of it). Thus, I would imagine that just about every widget that currently runs on OSX will run on the iPhone. That means that developers will be able to develop widget based applications now, for OSX, that will migrate easily to the new platform when it comes out and will work on the existing OS.

    Desktop app and mobile app from the same code base with no migration and no need for emulation software at dev time? That sounds pretty cool to me.

  39. Sure, Michael, I agree that they continued to target the sector. But I think there were already the seeds for seeing the market differently in the timespan you are talking about. Of course, at the time you are talking about, depending on your business, Apple did have a compelling small office ecosystem. I suppose what I’m realling talking about came a couple of years later as DTP really started to take off. (And, again, I’m not suggesting that from a marketing and sales perspective they weren’t attacking a larger market… Simple that it was clear technologically and strategically they were serving a smaller but still profitable market.)

    The point being: it wasn’t lack of developers or quality applications that limited Apple for two decades. It was the fact that the PC industry, from the business market on down, grew faster by commoditizing hardware.

  40. Sure, Michael, I agree that they continued to target the sector. But I think there were already the seeds for seeing the market differently in the timespan you are talking about. Of course, at the time you are talking about, depending on your business, Apple did have a compelling small office ecosystem. I suppose what I’m realling talking about came a couple of years later as DTP really started to take off. (And, again, I’m not suggesting that from a marketing and sales perspective they weren’t attacking a larger market… Simple that it was clear technologically and strategically they were serving a smaller but still profitable market.)

    The point being: it wasn’t lack of developers or quality applications that limited Apple for two decades. It was the fact that the PC industry, from the business market on down, grew faster by commoditizing hardware.

  41. Dave: Steve Jobs already said that all widgets that run on OSX won’t run on iPhone. So, what does the API really look like on the iPhone? What percentage of desktop OSX is on the iPhone? Desktop OSX widgets should be able to call into Java libraries, no?

    Simon: PodTech’s “iPhone?” You should watch 1938Media. Or, watch for what Steve Gillmor and Jason Calacanis are up to. Or follow Eric Rice.

    Lots of surprises and thrills ahead.

  42. Dave: Steve Jobs already said that all widgets that run on OSX won’t run on iPhone. So, what does the API really look like on the iPhone? What percentage of desktop OSX is on the iPhone? Desktop OSX widgets should be able to call into Java libraries, no?

    Simon: PodTech’s “iPhone?” You should watch 1938Media. Or, watch for what Steve Gillmor and Jason Calacanis are up to. Or follow Eric Rice.

    Lots of surprises and thrills ahead.

  43. When you don’t just write for the tech savvy person interested in tech you should explain what LAMP is with first use of the acronym.

    It’s common courtesy to do so, particularly when the meaning of such an acronym has a lot to do with what the writer’s talking about.

    A reader shouldn’t have to look up in Wikipedia or elsewhere what LAMP is.

    This blog isn’t just for people like Goebbels, after all, Robert’s family reads the blog. How many of them, other than me, do you think know what LAMP means, other than it’s a device that illuminates a space, usually while sitting on a DESK or TABLE.

  44. Scoble, what if we know of everything you mentioned (1938Media, Steve Gillmor, Jason Calacanis, Eric Rice), and we are still wondering what PodTech’s “iPhone” is? I don’t see anything that’s going to save you yet.

    You still haven’t explained why Java is superior to Apple’s own development tools.

  45. When you don’t just write for the tech savvy person interested in tech you should explain what LAMP is with first use of the acronym.

    It’s common courtesy to do so, particularly when the meaning of such an acronym has a lot to do with what the writer’s talking about.

    A reader shouldn’t have to look up in Wikipedia or elsewhere what LAMP is.

    This blog isn’t just for people like Goebbels, after all, Robert’s family reads the blog. How many of them, other than me, do you think know what LAMP means, other than it’s a device that illuminates a space, usually while sitting on a DESK or TABLE.

  46. Scoble, what if we know of everything you mentioned (1938Media, Steve Gillmor, Jason Calacanis, Eric Rice), and we are still wondering what PodTech’s “iPhone” is? I don’t see anything that’s going to save you yet.

    You still haven’t explained why Java is superior to Apple’s own development tools.

  47. Did Steve Jobs tell you about his iPhone before he was ready to? No.

    Why is Java superior? Developers tell me they don’t need to worry as much about managing memory as they do with non garbage collected languages and platforms. They also like being able to reuse their algorithms on many different platforms, not just one with 5% market share.

  48. Did Steve Jobs tell you about his iPhone before he was ready to? No.

    Why is Java superior? Developers tell me they don’t need to worry as much about managing memory as they do with non garbage collected languages and platforms. They also like being able to reuse their algorithms on many different platforms, not just one with 5% market share.

  49. I didn’t ask Jobs to tell me about the iPhone before he released it. You DID ask Jonathan Schwartz what his iPhone is. See the difference?

    Apple now has garbage collection. Catch up with the times.

    So now we are down to just reusing code. Do you want iPhone apps that do not have the native UI and no means currently to access the UI through gestures?

    Do you really think that interest in the iPhone as a platform is limited by the ability to re-use existing Java apps? Or do you think developers would be willing to adopt new methods to get onboard?

    (If you provide rational answers to those questions, I can ask again: why is Java superior to Apple’s own APIs.)

  50. I didn’t ask Jobs to tell me about the iPhone before he released it. You DID ask Jonathan Schwartz what his iPhone is. See the difference?

    Apple now has garbage collection. Catch up with the times.

    So now we are down to just reusing code. Do you want iPhone apps that do not have the native UI and no means currently to access the UI through gestures?

    Do you really think that interest in the iPhone as a platform is limited by the ability to re-use existing Java apps? Or do you think developers would be willing to adopt new methods to get onboard?

    (If you provide rational answers to those questions, I can ask again: why is Java superior to Apple’s own APIs.)

  51. Sorry for an off topic question, but I’ve been subscribed to Robert’s RSS feed via Sage in Firefox for a while now (love it!). Recently, all the videos he posts start playing simultaneously from within Sage. How can I stop that?

    Thanks

  52. Sorry for an off topic question, but I’ve been subscribed to Robert’s RSS feed via Sage in Firefox for a while now (love it!). Recently, all the videos he posts start playing simultaneously from within Sage. How can I stop that?

    Thanks

  53. Scoble, just by asking the question about Java on the iPhone shows a true misunderstanding about Apple. They don’t do lowest common denominator technology and in this context, that’s what Java is.

    Anyone developing for this device knows that whatever they develop will be specific to the iPhone. But that’s okay, because they’ll be able to make apps that could not exist on any other device anyway. Not to mention, this is the first of a family of devices and Apple is going to sell many millions of them. Just like the iPod.

    People seem to forget that Apple has been making developer tools since Woz created Integer BASIC. They already have great technology (QuickTime, Core Animation, Quartz, etc.) so there’s no need to run someone else’s stuff.

    As someone mentioned, Objective C 2.0 now has garbage collection and some other new features. Existing Mac OS X developers will feel right at home. And developers that know C/C++ can learn Objective C in a day or so.

    The opportunity will be great, once Apple announces how developers can join in, as there will likely be lots of different devices running OS X. It won’t be a free for all–I don’t hear the folks participating in the $1 billion add-ons market for the iPod complaining.

  54. Scoble, just by asking the question about Java on the iPhone shows a true misunderstanding about Apple. They don’t do lowest common denominator technology and in this context, that’s what Java is.

    Anyone developing for this device knows that whatever they develop will be specific to the iPhone. But that’s okay, because they’ll be able to make apps that could not exist on any other device anyway. Not to mention, this is the first of a family of devices and Apple is going to sell many millions of them. Just like the iPod.

    People seem to forget that Apple has been making developer tools since Woz created Integer BASIC. They already have great technology (QuickTime, Core Animation, Quartz, etc.) so there’s no need to run someone else’s stuff.

    As someone mentioned, Objective C 2.0 now has garbage collection and some other new features. Existing Mac OS X developers will feel right at home. And developers that know C/C++ can learn Objective C in a day or so.

    The opportunity will be great, once Apple announces how developers can join in, as there will likely be lots of different devices running OS X. It won’t be a free for all–I don’t hear the folks participating in the $1 billion add-ons market for the iPod complaining.

  55. Remember (or, if you’re too young, research) the early history of Apple’s relationship with developers…they sued them out of business if they didn’t do exactly as they were told. They don’t act much like they want to be a development platform – they seem to only want a small cadre of “Apple Approved” apps. Microsoft isn’t the only “Be quiet – we know what’s best for you” major player.

  56. Remember (or, if you’re too young, research) the early history of Apple’s relationship with developers…they sued them out of business if they didn’t do exactly as they were told. They don’t act much like they want to be a development platform – they seem to only want a small cadre of “Apple Approved” apps. Microsoft isn’t the only “Be quiet – we know what’s best for you” major player.

  57. “Oh, and someone asked why would Steve Jobs put Java on the iPhone? My answer: it’s dramatically easier to write applications for a runtime-compiler system like Java or .NET. ”

    Actually, the advantage of a virtual machine is portability. Java and C# just happen to be high-level languages with garbage collection and ship with a large collection of libraries. It’s these features that allow developers build apps quicker.

    Cocoa was actually born as the OPENSTEP API – a joint project between Sun and NeXT. When used with Interface Builder, it’s very easy to put together a Cocoa application with just a few (or even zero) lines of code. Before being ported to Power PC by Apple, It ran on Sparc, 68k, Intel, and even Windows NT. This allows Cocoa applications to be ported to other platforms, such as the ARM processor on the iPhone. with little difficulty. Essentially, Cocoa is the “VM” that powers both Mac OS X and the iPhone.

    Objective-C is very similar to Java and C# in that it’s a high-level, object oriented language based on a C like syntax. By the time the iPhone ships, Objective-C 2.0 will be released with optional garbage collection and support for class properties. Developers coming from a background in Java (as I did) or C# should find it relatively easy to move to Objective-C and Cocoa.

    As for it being difficult to develop on the Mac, Mac OS X really isn’t Mac OS anymore – it’s essentially OPENSTEP version 7 in Mac OS clothing. If opened to third-party developers by Apple, the iPhone has the potential to expose significantly more developers to Cocoa and Objective-C which, in turn, could create more applications on Mac OS X.

  58. “Oh, and someone asked why would Steve Jobs put Java on the iPhone? My answer: it’s dramatically easier to write applications for a runtime-compiler system like Java or .NET. ”

    Actually, the advantage of a virtual machine is portability. Java and C# just happen to be high-level languages with garbage collection and ship with a large collection of libraries. It’s these features that allow developers build apps quicker.

    Cocoa was actually born as the OPENSTEP API – a joint project between Sun and NeXT. When used with Interface Builder, it’s very easy to put together a Cocoa application with just a few (or even zero) lines of code. Before being ported to Power PC by Apple, It ran on Sparc, 68k, Intel, and even Windows NT. This allows Cocoa applications to be ported to other platforms, such as the ARM processor on the iPhone. with little difficulty. Essentially, Cocoa is the “VM” that powers both Mac OS X and the iPhone.

    Objective-C is very similar to Java and C# in that it’s a high-level, object oriented language based on a C like syntax. By the time the iPhone ships, Objective-C 2.0 will be released with optional garbage collection and support for class properties. Developers coming from a background in Java (as I did) or C# should find it relatively easy to move to Objective-C and Cocoa.

    As for it being difficult to develop on the Mac, Mac OS X really isn’t Mac OS anymore – it’s essentially OPENSTEP version 7 in Mac OS clothing. If opened to third-party developers by Apple, the iPhone has the potential to expose significantly more developers to Cocoa and Objective-C which, in turn, could create more applications on Mac OS X.

  59. “Cocoa was actually born as the OPENSTEP API – a joint project between Sun and NeXT.”

    No. Before Openstep existed, it was Nextsep and it began life at NeXT on the black cube.

    Once NeXT decided to get out of the hardware business, it ported Nextstep (renaming it OpenStep) to Intel, etc. Later came a runtime environment that allowed OpenStep applications to run unmodified on Sparc and Windows NT.

  60. “Cocoa was actually born as the OPENSTEP API – a joint project between Sun and NeXT.”

    No. Before Openstep existed, it was Nextsep and it began life at NeXT on the black cube.

    Once NeXT decided to get out of the hardware business, it ported Nextstep (renaming it OpenStep) to Intel, etc. Later came a runtime environment that allowed OpenStep applications to run unmodified on Sparc and Windows NT.

  61. While it’s true that Cocoa is firmly based on technology which originally debuted on the NeXT cube, much of Cocoa’s platform independence is due the OpenStep specification created by NeXT and Sun.

    [Having said that, I'd still love to get my hands on a black NeXT cube to go along with my blue BeBox development system from Be, Inc.]

    OpenStep was a significant in that it separated the NeXTStep API from the core OS and refactored it into separate Foundation and AppKit frameworks. It also added important features, such as retain/release memory management and low-level objects for basic data types and conversion. This resulted in a relatively ‘endian-free’ API that could run on nearly any platform, including Solaris running on Sun’s Sparc CPUs.

    Since Mac OS X is based on NeXT’s implementation of OpenStep, Apple was able to bring Cocoa to the iPhone’s ARM processor with relative ease. Non-Cocoa based technologies, such as Java, would require more effort to port and a significant increase increase the iPhone’s OS footprint.

    Before losing interest in OpenStep, Sun purchased a major NeXT development company, Lighthouse Design, for it’s internal OpenStep development group in 1996. Jonathan Schwartz was the CEO and cofounder of Lighthouse and joined Sun during the acquisition. Lighthouse was moved into the Javasoft group when Sun shifted it’s desktop application development effort to Java.

  62. While it’s true that Cocoa is firmly based on technology which originally debuted on the NeXT cube, much of Cocoa’s platform independence is due the OpenStep specification created by NeXT and Sun.

    [Having said that, I'd still love to get my hands on a black NeXT cube to go along with my blue BeBox development system from Be, Inc.]

    OpenStep was a significant in that it separated the NeXTStep API from the core OS and refactored it into separate Foundation and AppKit frameworks. It also added important features, such as retain/release memory management and low-level objects for basic data types and conversion. This resulted in a relatively ‘endian-free’ API that could run on nearly any platform, including Solaris running on Sun’s Sparc CPUs.

    Since Mac OS X is based on NeXT’s implementation of OpenStep, Apple was able to bring Cocoa to the iPhone’s ARM processor with relative ease. Non-Cocoa based technologies, such as Java, would require more effort to port and a significant increase increase the iPhone’s OS footprint.

    Before losing interest in OpenStep, Sun purchased a major NeXT development company, Lighthouse Design, for it’s internal OpenStep development group in 1996. Jonathan Schwartz was the CEO and cofounder of Lighthouse and joined Sun during the acquisition. Lighthouse was moved into the Javasoft group when Sun shifted it’s desktop application development effort to Java.