I thought we were talking about mobile platforms where IE has next to no influence. HTML/CSS/Javascript are coming together and is the major web development platform of the future.
What part of "no common runtime for developers to develop and test against" do you not understand?
Quote:
What are you talking about. Both the iPhone and Android fully support apps built in HTML/CSS/Javascript.
Each with their individual quirks. No two browsers have a 100% identical runtime.
Quote:
On Apple's website they highlight over 4000 apps built in javascript, all of Google's in-house apps are built in javascript.
I think everyone will agree that the App Store for mobile apps written in Cocoa is what makes the iPhone platform so compelling. It didn't catch developers' imagination with 3rd party apps were web only and native apps were restricted.
And don't think for a second that Google doesn't struggle with supporting multiple browsers, including IE 6, which is the retarded stepchild of the web that simply won't go away.
Quote:
How do you define "took off"? I don't see the best most innovative developers running to J2ME, and don't see the best most innovative apps being created in J2ME.
Until the iPhone came out, J2ME was the platform of choice for mobile developers. Every service provider supported it and J2ME was a cash cow. Companies like Ericsson are still heavily invested in it.
So....Oracle goes from being a database vendor to a database + middleware + ERP vendor and is now offering the full stack of client + middleware + server + database + + operating system + hardware, and you think they're going to fail?
Don't get me wrong, they're absolutely brutal. What they did to WebLogic customers wasn't pretty. However, these guys know how to make a buck.
"The more you tighten your grip, Ellison, the more star systems will slip through your fingers."
No, it wasn't pretty. We used to have a comp'd dev license for prototype development which Oracle nix'd. So now we do JBoss. Who lost?
But the xerox comment was more directed at Sun than Oracle. You know...the company that is now going to be owned by Oracle?
But no, Oracle won't do anything better with the technology. It's just not that sorta company.
Besides...did you listen to Ellison at Java One? Who are you going to bet on? Google and Sergey or Oracle and Larry? JavaFX vs Ajax/Gears/GWT/HTML5?
He's gonna get serious with netbooks and phones? Really? Against Android and Chrome?
First of all, Java is the world's most popular programming language and the preferred platform for enterprise development. Sun makes a tidy profit from its Java division, so your comment about Java being unprofitable is just nonsense.
Java is most popular but look at the slope.
Quote:
Sun lost for different reasons.
Sun lost due to missteps at two key points.
Charging too much for Solaris X86 and allowing Windows NT Server entry into the low end server market. If someone is going to take share from you, it's better that the someone is you. Instead of wide deployment of Solaris X86 on commodity Dell servers for cheap, it was Windows NT Server 4.0 that took massive share away from low end Sparc boxes.
Dropping OpenStep in favor of Java. Had they kept OpenStep as thier desktop then it would have been Solaris/OpenStep vs HPUX/CDE and AIX/CDE. No contest.
AND they would have had a desktop unix that didn't suck. Which Apple has with OSX and Sun never did.
Sun was a HARDWARE company and should have remembered that. Cross platform Java may have made them money, but nothing near to the money they didn't make had they managed to dominate the Unix market and pushed unix to the desktop.
Android is very open, which is a plus for kids in dorm rooms. But the iPhone hardware is so much more appealing, as is the OS, so the kids want to develop for iPhone more.
As for us non-kids, I would say that I'll only develop for Android if want I need to achieve is unreasonably difficult on the iPhone. If you're trying to gain support, get financing, or impress people, there's nothing like an iPhone demo.
Not sure I see your point. The only other language that's even close is C.
Quote:
Sun lost due to missteps at two key points.
Charging too much for Solaris X86 and allowing Windows NT Server entry into the low end server market. If someone is going to take share from you, it's better that the someone is you. Instead of wide deployment of Solaris X86 on commodity Dell servers for cheap, it was Windows NT Server 4.0 that took massive share away from low end Sparc boxes.
Well, actually, it's Linux that hurt them more than NT. That's their real competitor, and prompted them to port Solaris to x86 and open source it as OpenSolaris. OpenSolaris is really coming along nicely, and has a lot of customers. The problem is, they should have started that initiative at least 5 years before they did.
Quote:
Dropping OpenStep in favor of Java. Had they kept OpenStep as thier desktop then it would have been Solaris/OpenStep vs HPUX/CDE and AIX/CDE. No contest.
Not sure I follow you. I thought we were talking about Sun's business prowess, not its technological prowess. How is supporting OpenStep a business decision? From what you're telling me, it was more a question of picking the best technology.
Quote:
AND they would have had a desktop unix that didn't suck. Which Apple has with OSX and Sun never did.
I thought you said Sun was in the business of selling hardware, not software. Besides, OpenStep would have failed even with Sun, and NeXT would have still been taken over by Apple. NeXT's products lost credibility once Apple took over the company. One important example is WebObjects, which was miles ahead technologically and supported by many large corporations. Once Apple took over, it lost to Microsoft's inferior IIS/ASP framework.
Quote:
Sun was a HARDWARE company and should have remembered that. Cross platform Java may have made them money, but nothing near to the money they didn't make had they managed to dominate the Unix market and pushed unix to the desktop.
It would still have been a wash. IBM would have still rallied behind Linux and if the Unix desktop market existed they would have found a way to compete.
Ultimately, Java succeeded on the server. OpenStep was purely a GUI platform, and Objective-C never took off as a server-side language.
I read all these comments on AppleInsider about how Android will "suck", "fail", or both, and that the only mobile platform that has a snowball's chance in hell of succeeding will be the iPhone. It's as though Apple will be able to single-handedly wipe out the following platforms:
1) Blackberry
2) Nokia/Symbian
3) Windows Mobile + Hardware partners
4) Google Android + Hardware partners
5) J2ME/JavaFX + Hardware partners
Still, I'm going to bet that this time next year, one of (if not both) Windows Mobile and Symbian drop out of contention in the mobile market, since the space is getting quite crowded. Android is free. Windows Mobile isn't, and its UI is worse.
But let's focus on your "laughing stock" comment. Please justify how Android is more of a "laughing stock" than the following platforms:
1) Windows Mobile (IMO, has 0 usability beyond making bullet points for PHBs)
2) Symbian (arguably even worse than Windows Mobile
3) J2ME (which even I'll admit, while it had business backing from the get-go, is far from the most pleasant user experience)
4) Blackberry OS (while business's love it, it scores very low in user friendliness)
I haven't personally used an Android phones, but I've seen demos and it ranks right up there in usability and is a solid #2 to the iPhone, which I'll admit has by a safe margin the best user interface of all.
I suggest you actually try an Android phone, and try out some of the Windows Mobile/Symbian/Blackberry phones before declaring it a "laughing stock".
What part of "no common runtime for developers to develop and test against" do you not understand?
I agreed that it's currently rough around the edges. But there is a great deal more money and effort into making it a common runtime, than is going into java.
Quote:
Each with their individual quirks. No two browsers have a 100% identical runtime.
This is the nature of developing open source vs software development from one company. This won't be the situation for long.
Quote:
I think everyone will agree that the App Store for mobile apps written in Cocoa is what makes the iPhone platform so compelling. It didn't catch developers' imagination with 3rd party apps were web only and native apps were restricted.
This is true. But native apps don't and cannot cover everything. There are a great deal of services that are web based. Most all services from Google, Yahoo, AOL are all web based.
Quote:
And don't think for a second that Google doesn't struggle with supporting multiple browsers, including IE 6, which is the retarded stepchild of the web that simply won't go away.
IE 6 has no influence on mobile devices.
Quote:
Until the iPhone came out, J2ME was the platform of choice for mobile developers. Every service provider supported it and J2ME was a cash cow. Companies like Ericsson are still heavily invested in it.
Looking at the quality of those apps and how widely they were used I'm not sure that is something to brag about.
"The more you tighten your grip, Ellison, the more star systems will slip through your fingers."
No, it wasn't pretty. We used to have a comp'd dev license for prototype development which Oracle nix'd. So now we do JBoss. Who lost?
I still see tons of job postings in my area that ask for WebLogic. The company I currently work for is still using it for several of their projects, and a couple of the managers were bitching that Oracle keeps trying to sign them up for more products/services and charge them more. From what I understand, Oracle made a calculated decision to lose, say, 20% of WebLogic customers and fleece the rest. Thus far, I see no indication that the other 80% have switched. So to answer your question, Oracle's customers seem to be losing so far.
Quote:
But the xerox comment was more directed at Sun than Oracle. You know...the company that is now going to be owned by Oracle?
Java is a successful technology and a successful business. Granted, it's not a runaway success but by no means a failure as a business. Sun made other mistakes that did them in, most of them when Scott McNealy was running the show.
Quote:
But no, Oracle won't do anything better with the technology. It's just not that sorta company.
So... technologies like BTFS don't constitute "doing better with technology"?
Quote:
Besides...did you listen to Ellison at Java One? Who are you going to bet on? Google and Sergey or Oracle and Larry? JavaFX vs Ajax/Gears/GWT/HTML5?
JavaFX is a technology with a lot of promise, but still a very young technology. Sun has learned from the mistakes they made with Java on the client side (Swing, Applets, J2ME), and you can tell in the design decisions made with JavaFX. What's more, Sun has totally ignored backward compatibility consideration, as each new release of JavaFX breaks the next. Oracle can invest the resources in making this technology work.
Pure AJAX just doesn't work. There's too much configuration and boilerplate code and very little tooling. GWT is another story, but it's hampered by the fact that it has to support IE 6 and other strange browsers. Gears for local storage is well and all, but is still a bit of a hack when compared to true client-side applications. HTML 5 isn't even a complete spec yet, and each browser treats it differently. Obviously, IE 6/7/8 won't support it at all, which creates another set of problems.
On the other hand, JavaFX represents a single runtime to code against. It leverages the JVM and the rich Java API and class libraries. What's more, it's the same API (with a few differences) between mobile and desktop. The potential is there to blow the lid off Flash/Flex, because Adobe has different versions of Flash for desktop and mobile. You can write the same app for JavaFX desktop and mobile.
And, oh yeah, contrary to popular belief, the Java runtime is available on most Windows, Linux and OS X desktops. It's not as ubiquitous as Flash, but it's not that far off.
Quote:
He's gonna get serious with netbooks and phones? Really? Against Android and Chrome?
JavaFX can in theory run on top of both platforms.
I read all these comments on AppleInsider about how Android will "suck", "fail", or both, and that the only mobile platform that has a snowball's chance in hell of succeeding will be the iPhone. It's as though Apple will be able to single-handedly wipe out the following platforms:
1) Blackberry
2) Nokia/Symbian
3) Windows Mobile + Hardware partners
4) Google Android + Hardware partners
5) J2ME/JavaFX + Hardware partners
Still, I'm going to bet that this time next year, one of (if not both) Windows Mobile and Symbian drop out of contention in the mobile market, since the space is getting quite crowded. Android is free. Windows Mobile isn't, and its UI is worse.
But let's focus on your "laughing stock" comment. Please justify how Android is more of a "laughing stock" than the following platforms:
1) Windows Mobile (IMO, has 0 usability beyond making bullet points for PHBs)
2) Symbian (arguably even worse than Windows Mobile
3) J2ME (which even I'll admit, while it had business backing from the get-go, is far from the most pleasant user experience)
4) Blackberry OS (while business's love it, it scores very low in user friendliness)
I haven't personally used an Android phones, but I've seen demos and it ranks right up there in usability and is a solid #2 to the iPhone, which I'll admit has by a safe margin the best user interface of all.
I suggest you actually try an Android phone, and try out some of the Windows Mobile/Symbian/Blackberry phones before declaring it a "laughing stock".
Humm lets see... the iphone, the 2007 invention of the year, has sold 22 million copies worldwide and over 1 billion apps have been downloaded from the app store.
The android has sold a copy to Sergey, one to larry, and the CEO also has one because it would be inappropriate for him to show up to board meetings without one. Android was created by Sergey because he was getting too bored on the "bang-plane".
Need I go on? LOL!!!
This whole thread is absolutely ridiculous.
The only thing that can stand up to the iphone is blackberry.
Humm lets see... the iphone, the 2007 invention of the year, has sold 22 million copies worldwide and over 1 billion apps have been downloaded from the app store.
I never disagreed with the fact that the iPhone is a runaway success. The subject is your statement that Android is a "laughing stock". This part of your post is a non sequitur.
Quote:
The android has sold a copy to Sergey, one to larry, and the CEO also has one because it would be inappropriate for him to show up to board meetings without one. Android was created by Sergey because he was getting too bored on the "bang-plane".
I suppose you have proof to back up your statement that Android was "just for fun". Of course you.
Here again is the list of partners in Android's Open Handset Alliance:
Apparently you do, because you're making your case based on hyperbole and ad hominem statements. In other words, you've failed to leave out logical reasoning and facts so far.
Quote:
This whole thread is absolutely ridiculous.
Since you started it by stating that Android is a "laughing stock", I wholeheartedly agree.
Quote:
The only thing that can stand up to the iphone is blackberry.
Since four Android phones are being sold thus far, with many more on the way before the end of the year, I wholeheartedly disagree.
Not sure I see your point. The only other language that's even close is C.
The slope is downward not upward.
Quote:
Well, actually, it's Linux that hurt them more than NT. That's their real competitor, and prompted them to port Solaris to x86 and open source it as OpenSolaris. OpenSolaris is really coming along nicely, and has a lot of customers. The problem is, they should have started that initiative at least 5 years before they did.
No. In 1997 NT server market share was 36%. Microsoft took the low end server market and workstation markets that unix makers used to own.
By the time linux became a real server OS (as in after IBM dumped an assload of IP into Linux) Microsoft had already taken a large chunk out of the mid-range server markets too with 4-way x86 boxes competing against Sparcs, PA-RISC, etc...and winning.
Sure, you can look at companies like Google and see linux as dominant...but Exchange, windows file shares and Sharepoint are what's deployed by IT for their enterprise services on those low and mid-grade servers. Shops that used to be Solaris or HPUX or AIX or IRIX or Digital Unix.
If linux killed Solaris, the pray tell why folks aren't using sendmail but Exchange? Why don't folks nfs rather than smb? Why are folks authenticated via active directory vs nds/fds?
Windows won. Unix lost. It wasn't linux that killed Solaris although it sure hastened its death. The fatal shot was fired by MS.
Linux just took away what meager share Sun still had in the lower end of the unix market. IBM did what it did because they cared more about selling services than selling servers with AIX. Solaris was beating AIX anyway so by going Linux they hurt HP and Sun far more than themselves.
Quote:
Not sure I follow you. I thought we were talking about Sun's business prowess, not its technological prowess. How is supporting OpenStep a business decision? From what you're telling me, it was more a question of picking the best technology.
It was a business decision on what to focus on. The desktop or a new language. One is a user centric focus. The other a developer one.
One gives you desktop penetration and the ability to fight on MS's incursion into the server market and defend your critical workstation market hardware sales. The other, not so much.
Quote:
I thought you said Sun was in the business of selling hardware, not software.
That's the whole point. OpenStep is a competitive advantage for selling Sun equipment running Solaris. Even if that hardware is commodity x86 hardware with a "Sun Tax" (vs "Apple Tax").
Java is not given it runs just as well on any JVM living anywhere.
Quote:
Besides, OpenStep would have failed even with Sun, and NeXT would have still been taken over by Apple.
NEXTSTEP and OpenStep were MILES ahead of CDE and any other unix desktop on the market. It would have taken share from HPUX, AIX, IRIX. It was also better than either MacOS or Windows. It would not have failed.
Sun should have bought NeXT and pulled Jobs in. After black Tuesday Jobs would have sold if Sun offered Jobs a free hand for the workstation market.
Jobs and the NeXT guys did for unix what Jobs and the Pixar guys did for animation.
There are millions of unix users today who've never seen a command line. There are millions of unix users today thrilled with their desktop experience.
It sure as hell wasn't because of either Java or Linux.
Quote:
NeXT's products lost credibility once Apple took over the company. One important example is WebObjects, which was miles ahead technologically and supported by many large corporations. Once Apple took over, it lost to Microsoft's inferior IIS/ASP framework.
More like once Jobs took over Apple, he focused on the consumer market which has always been his focus. He was precluded from competing in that market and had to do the workstation market. That WebObjects fell by the wayside was to be expected.
OpenStep was never abandoned by Jobs. It's what OSX is based on.
Quote:
It would still have been a wash. IBM would have still rallied behind Linux and if the Unix desktop market existed they would have found a way to compete.
The Unix desktop market exists today. It's called OSX. It could have existed in 1998...called OpenStep on Solaris x86 at NT pricing.
And IBM wouldn't have anything to counter that.
Quote:
Ultimately, Java succeeded on the server. OpenStep was purely a GUI platform, and Objective-C never took off as a server-side language.
Ultimately, Java failed to save Sun. OpenStep became the dominant unix desktop UI, and Objective-C took off on the iPhone and there are many thousands of new Objective-C developers today as a consequence. More importantly, Apple has billions in the bank while Sun is dead. OpenStep made Apple far more money than Sun has ever made from Java.
Because Java killed OpenStep interest at Sun, effectively Java killed Sun.
I was a NEXTSTEP and SunOS/Solaris developer in the early 90s. It was a crucial moment for Sun and they went down the wrong path. That path was Java.
Actually, the other thing you seem to forget is that Google's use of Java isn't standard, either in GWT or in Android or in ChromeOS.
So no common runtime...Android isn't J2ME. ChromeOS won't be J2SE. And Android's NDK allows C/C++. Heh...and maybe in the future they'll add ObjC.
Dalvik isn't you dad's JVM.
It's standard in the sense that there aren't multiple implementations of Android to code against, as there are multiple browsers to code against. This means if I write an Android app, I only have to test it once, not multiple times as I would for a HTML/CSS/JavaScript app for each browser.
Sun's JVM supports multiple languages, and this is not a weakness. It allows developers a choice of languages to use to code against the JVM. Java contains JNI (Java Native Interface), which Google could re-implement in Android's JVM.
BTW, you're on crack if you think anybody except Apple and maybe a couple of obscure open source projects (ex Étoile) is ever going to use Objective-C. It's a language from the 80's with a bizarre syntax. Apple has obviously done amazing things with it, but the only reason it has any traction outside the company is because it's the only way to code iPhone apps currently. Apple would be wise to enhance multiple language support for Cocoa.
It's standard in the sense that there aren't multiple implementations of Android to code against, as there are multiple browsers to code against. This means if I write an Android app, I only have to test it once, not multiple times as I would for a HTML/CSS/JavaScript app for each browser.
Sure, but this is the same as saying .NET or Silverlight is standard. Well...yeah, for MS based platforms (ignore Mono and Moonlight for the moment).
Quote:
Sun's JVM supports multiple languages, and this is not a weakness. It allows developers a choice of languages to use to code against the JVM. Java contains JNI (Java Native Interface), which Google could re-implement in Android's JVM.
BTW, you're on crack if you think anybody except Apple and maybe a couple of obscure open source projects (ex Étoile) is ever going to use Objective-C. It's a language from the 80's with a bizarre syntax. Apple has obviously done amazing things with it, but the only reason it has any traction outside the company is because it's the only way to code iPhone apps currently. Apple would be wise to enhance multiple language support for Cocoa.
Heh, I could say the same about being on crack if you think the multi-language support in JVMs will ever be any more significant than ObjC.
Sure, but this is the same as saying .NET or Silverlight is standard. Well...yeah, for MS based platforms (ignore Mono and Moonlight for the moment).
Yes. .NET is a standard according to the point that I was making.. Developers can write to the .NET runtime and expect consistent performance on all installations. Aside from Mono, there's only one .NET runtime.
Quote:
Heh, I could say the same about being on crack if you think the multi-language support in JVMs will ever be any more significant than ObjC.
They'll be just as significant as the JVM as a whole is, which is a hell of a lot more significant than Objective-C is for enterprise development.
Quote:
Way to go ignoring all the other points though.
I'm responding to your points in this particular post, not your other post about the history of OpenSTEP.
True, but the JVM has more than one language. The Java language has been in legacy mode since Java 5. Groovy, Scala, JPython, JRuby, and JavaFX are picking up much of the ground that Java is leaving behind. Scala in particular is interesting, since it has all the advantages of Java with more compact syntax and modern features such as closures and the ability to use domain specific languages such as XML and SQL.
Quote:
No. In 1997 NT server market share was 36%. Microsoft took the low end server market and workstation markets that unix makers used to own.
By the time linux became a real server OS (as in after IBM dumped an assload of IP into Linux) Microsoft had already taken a large chunk out of the mid-range server markets too with 4-way x86 boxes competing against Sparcs, PA-RISC, etc...and winning.
Sure, you can look at companies like Google and see linux as dominant...but Exchange, windows file shares and Sharepoint are what's deployed by IT for their enterprise services on those low and mid-grade servers. Shops that used to be Solaris or HPUX or AIX or IRIX or Digital Unix.
If linux killed Solaris, the pray tell why folks aren't using sendmail but Exchange? Why don't folks nfs rather than smb? Why are folks authenticated via active directory vs nds/fds?
Windows won. Unix lost. It wasn't linux that killed Solaris although it sure hastened its death. The fatal shot was fired by MS.
Linux just took away what meager share Sun still had in the lower end of the unix market. IBM did what it did because they cared more about selling services than selling servers with AIX. Solaris was beating AIX anyway so by going Linux they hurt HP and Sun far more than themselves.
Ah...screw it! I have to admit defeat on this one.
Quote:
It was a business decision on what to focus on. The desktop or a new language. One is a user centric focus. The other a developer one.
One gives you desktop penetration and the ability to fight on MS's incursion into the server market and defend your critical workstation market hardware sales. The other, not so much.
So, are you saying that Sun should have totally ignored Java when it came out of their R&D lab? Java still has a lot of tech that Objective-C doesn't have. There's a reason Apple switched WebObjects from Objective-C to Java, and I doubt that it was political.
Quote:
That's the whole point. OpenStep is a competitive advantage for selling Sun equipment running Solaris. Even if that hardware is commodity x86 hardware with a "Sun Tax" (vs "Apple Tax").
Java is not given it runs just as well on any JVM living anywhere.
NEXTSTEP and OpenStep were MILES ahead of CDE and any other unix desktop on the market. It would have taken share from HPUX, AIX, IRIX. It was also better than either MacOS or Windows. It would not have failed.
Unless Sun's marketing really screwed up, which even I would admit is a possibility
Quote:
Sun should have bought NeXT and pulled Jobs in. After black Tuesday Jobs would have sold if Sun offered Jobs a free hand for the workstation market.
Jobs and the NeXT guys did for unix what Jobs and the Pixar guys did for animation.
There are millions of unix users today who've never seen a command line. There are millions of unix users today thrilled with their desktop experience.
It sure as hell wasn't because of either Java or Linux.
From what I hear, Apple was dead set on buying Be before they changed their mind at the last minute, and that was just pure chance. Nobody apparently wanted NeXT as nobody really saw the point of the technology.
Quote:
More like once Jobs took over Apple, he focused on the consumer market which has always been his focus. He was precluded from competing in that market and had to do the workstation market. That WebObjects fell by the wayside was to be expected.
OpenStep was never abandoned by Jobs. It's what OSX is based on.
It was abandoned as a standalone product sold/distributed to other companies. That was my point.
Quote:
The Unix desktop market exists today. It's called OSX. It could have existed in 1998...called OpenStep on Solaris x86 at NT pricing.
And IBM wouldn't have anything to counter that.
Ultimately, Java failed to save Sun. OpenStep became the dominant unix desktop UI, and Objective-C took off on the iPhone and there are many thousands of new Objective-C developers today as a consequence. More importantly, Apple has billions in the bank while Sun is dead. OpenStep made Apple far more money than Sun has ever made from Java.
Because Java killed OpenStep interest at Sun, effectively Java killed Sun.
I was a NEXTSTEP and SunOS/Solaris developer in the early 90s. It was a crucial moment for Sun and they went down the wrong path. That path was Java.
There's one inconsistency in this story though. NeXT allowed an open source implementation: GNUStep. It's what WindowMaker and some other window managers use. Sure, it was GPL'd, but in theory any Unix vendor could have taken that tech and built a GUI out of it. Nobody did.
True, but the JVM has more than one language. The Java language has been in legacy mode since Java 5. Groovy, Scala, JPython, JRuby, and JavaFX are picking up much of the ground that Java is leaving behind. Scala in particular is interesting, since it has all the advantages of Java with more compact syntax and modern features such as closures and the ability to use domain specific languages such as XML and SQL.
Actually, Java 6 with finally a new LAF seems finally be moving forward a bit.
JavaFX is just bizzaro to me given how well WPF was designed by MS. When your avowed architecture is MVC and you add a new View technology, you kinda want Java to be able to cover the Model and Controller parts since it does it fairly well and there's a huge assload of legacy Java Model/Controller code tied to crappy Swing Views..
But you can't...because Java can't call into JavaFX directly except through hacks. Something that has been broken since 1.0. How on earth do consider something 1.0 when you can't do that? Beats me.
This has been the Sun Java story all along. Incredibly brilliant engineering crippled by an almost total lack of attention to what doesn't make life suck for the average developer.
Something that both Apple and MS does a much better job at IMHO.
Quote:
Ah...screw it! I have to admit defeat on this one.
Yah, I am (or was since Sun is no more) a big Sun fan. If I bitch about Sun it's because they had the potential for greatness. Or at least lasting greatness. Or at least an ending not quite this pathetic.
Sun is kinda like how Apple would have ended up without Jobs coming back.
Quote:
So, are you saying that Sun should have totally ignored Java when it came out of their R&D lab? Java still has a lot of tech that Objective-C doesn't have. There's a reason Apple switched WebObjects from Objective-C to Java, and I doubt that it was political.
Totally ignored? Heck no. Build the corporate strategy around it? Heck no. VB, .NET, C#, ObjC are all simply a means to an end for MS and Apple. Even C/C++ via gcc is simply a means to and end for the FSF.
Java could have been another great means to an end. The end being selling Sun hardware and not going out of business. But that requires also attention to the user, not to a language technologies for developers who are most likely on Windows.
Quote:
Unless Sun's marketing really screwed up, which even I would admit is a possibility
Yah, sure.
Quote:
It was abandoned as a standalone product sold/distributed to other companies. That was my point.
Sure, you don't give your competitors your competitive advantage if you have the resources to leverage it yourself. Licensing NEXTSTEP happened only because NeXT couldn't sell its own hardware in qty.
Quote:
There's one inconsistency in this story though. NeXT allowed an open source implementation: GNUStep. It's what WindowMaker and some other window managers use. Sure, it was GPL'd, but in theory any Unix vendor could have taken that tech and built a GUI out of it. Nobody did.
The value of an API is far lower than the amount of software required to implement it. Sun purchased a lot of code from NeXT. I'm not certain what happened to it but it sure didn't get open sourced. I think I remember Apple buying the licenses back or Sun letting lapse or something.
The GNUStep stuff is half done at best, isn't a desktop (but a "window manager") and frankly the GPL aspects kills any incentive for commercial companies to invest in it.
It's different when you GPL your own IP. You still get to use it for your own proprietary products. Adopting someone else's GPL code kills the monetary value of any IP you attach to it.
You do so because it doesn't matter to you anymore like with IBM and it's unix technology.
Comments
I thought we were talking about mobile platforms where IE has next to no influence. HTML/CSS/Javascript are coming together and is the major web development platform of the future.
What part of "no common runtime for developers to develop and test against" do you not understand?
What are you talking about. Both the iPhone and Android fully support apps built in HTML/CSS/Javascript.
Each with their individual quirks. No two browsers have a 100% identical runtime.
On Apple's website they highlight over 4000 apps built in javascript, all of Google's in-house apps are built in javascript.
I think everyone will agree that the App Store for mobile apps written in Cocoa is what makes the iPhone platform so compelling. It didn't catch developers' imagination with 3rd party apps were web only and native apps were restricted.
And don't think for a second that Google doesn't struggle with supporting multiple browsers, including IE 6, which is the retarded stepchild of the web that simply won't go away.
How do you define "took off"? I don't see the best most innovative developers running to J2ME, and don't see the best most innovative apps being created in J2ME.
Until the iPhone came out, J2ME was the platform of choice for mobile developers. Every service provider supported it and J2ME was a cash cow. Companies like Ericsson are still heavily invested in it.
So....Oracle goes from being a database vendor to a database + middleware + ERP vendor and is now offering the full stack of client + middleware + server + database + + operating system + hardware, and you think they're going to fail?
Don't get me wrong, they're absolutely brutal. What they did to WebLogic customers wasn't pretty. However, these guys know how to make a buck.
"The more you tighten your grip, Ellison, the more star systems will slip through your fingers."
No, it wasn't pretty. We used to have a comp'd dev license for prototype development which Oracle nix'd. So now we do JBoss. Who lost?
But the xerox comment was more directed at Sun than Oracle. You know...the company that is now going to be owned by Oracle?
But no, Oracle won't do anything better with the technology. It's just not that sorta company.
Besides...did you listen to Ellison at Java One? Who are you going to bet on? Google and Sergey or Oracle and Larry? JavaFX vs Ajax/Gears/GWT/HTML5?
He's gonna get serious with netbooks and phones? Really? Against Android and Chrome?
First of all, Java is the world's most popular programming language and the preferred platform for enterprise development. Sun makes a tidy profit from its Java division, so your comment about Java being unprofitable is just nonsense.
Java is most popular but look at the slope.
Sun lost for different reasons.
Sun lost due to missteps at two key points.
Charging too much for Solaris X86 and allowing Windows NT Server entry into the low end server market. If someone is going to take share from you, it's better that the someone is you. Instead of wide deployment of Solaris X86 on commodity Dell servers for cheap, it was Windows NT Server 4.0 that took massive share away from low end Sparc boxes.
Dropping OpenStep in favor of Java. Had they kept OpenStep as thier desktop then it would have been Solaris/OpenStep vs HPUX/CDE and AIX/CDE. No contest.
AND they would have had a desktop unix that didn't suck. Which Apple has with OSX and Sun never did.
Sun was a HARDWARE company and should have remembered that. Cross platform Java may have made them money, but nothing near to the money they didn't make had they managed to dominate the Unix market and pushed unix to the desktop.
Android is very open, which is a plus for kids in dorm rooms. But the iPhone hardware is so much more appealing, as is the OS, so the kids want to develop for iPhone more.
As for us non-kids, I would say that I'll only develop for Android if want I need to achieve is unreasonably difficult on the iPhone. If you're trying to gain support, get financing, or impress people, there's nothing like an iPhone demo.
Java is most popular but look at the slope.
Not sure I see your point. The only other language that's even close is C.
Sun lost due to missteps at two key points.
Charging too much for Solaris X86 and allowing Windows NT Server entry into the low end server market. If someone is going to take share from you, it's better that the someone is you. Instead of wide deployment of Solaris X86 on commodity Dell servers for cheap, it was Windows NT Server 4.0 that took massive share away from low end Sparc boxes.
Well, actually, it's Linux that hurt them more than NT. That's their real competitor, and prompted them to port Solaris to x86 and open source it as OpenSolaris. OpenSolaris is really coming along nicely, and has a lot of customers. The problem is, they should have started that initiative at least 5 years before they did.
Dropping OpenStep in favor of Java. Had they kept OpenStep as thier desktop then it would have been Solaris/OpenStep vs HPUX/CDE and AIX/CDE. No contest.
Not sure I follow you. I thought we were talking about Sun's business prowess, not its technological prowess. How is supporting OpenStep a business decision? From what you're telling me, it was more a question of picking the best technology.
AND they would have had a desktop unix that didn't suck. Which Apple has with OSX and Sun never did.
I thought you said Sun was in the business of selling hardware, not software. Besides, OpenStep would have failed even with Sun, and NeXT would have still been taken over by Apple. NeXT's products lost credibility once Apple took over the company. One important example is WebObjects, which was miles ahead technologically and supported by many large corporations. Once Apple took over, it lost to Microsoft's inferior IIS/ASP framework.
Sun was a HARDWARE company and should have remembered that. Cross platform Java may have made them money, but nothing near to the money they didn't make had they managed to dominate the Unix market and pushed unix to the desktop.
It would still have been a wash. IBM would have still rallied behind Linux and if the Unix desktop market existed they would have found a way to compete.
Ultimately, Java succeeded on the server. OpenStep was purely a GUI platform, and Objective-C never took off as a server-side language.
Android will become the laughing stock.
I read all these comments on AppleInsider about how Android will "suck", "fail", or both, and that the only mobile platform that has a snowball's chance in hell of succeeding will be the iPhone. It's as though Apple will be able to single-handedly wipe out the following platforms:
1) Blackberry
2) Nokia/Symbian
3) Windows Mobile + Hardware partners
4) Google Android + Hardware partners
5) J2ME/JavaFX + Hardware partners
Still, I'm going to bet that this time next year, one of (if not both) Windows Mobile and Symbian drop out of contention in the mobile market, since the space is getting quite crowded. Android is free. Windows Mobile isn't, and its UI is worse.
But let's focus on your "laughing stock" comment. Please justify how Android is more of a "laughing stock" than the following platforms:
1) Windows Mobile (IMO, has 0 usability beyond making bullet points for PHBs)
2) Symbian (arguably even worse than Windows Mobile
3) J2ME (which even I'll admit, while it had business backing from the get-go, is far from the most pleasant user experience)
4) Blackberry OS (while business's love it, it scores very low in user friendliness)
I haven't personally used an Android phones, but I've seen demos and it ranks right up there in usability and is a solid #2 to the iPhone, which I'll admit has by a safe margin the best user interface of all.
I suggest you actually try an Android phone, and try out some of the Windows Mobile/Symbian/Blackberry phones before declaring it a "laughing stock".
Oh yeah, and take a look at Android's partners:
http://www.openhandsetalliance.com/oha_members.html
What part of "no common runtime for developers to develop and test against" do you not understand?
I agreed that it's currently rough around the edges. But there is a great deal more money and effort into making it a common runtime, than is going into java.
Each with their individual quirks. No two browsers have a 100% identical runtime.
This is the nature of developing open source vs software development from one company. This won't be the situation for long.
I think everyone will agree that the App Store for mobile apps written in Cocoa is what makes the iPhone platform so compelling. It didn't catch developers' imagination with 3rd party apps were web only and native apps were restricted.
This is true. But native apps don't and cannot cover everything. There are a great deal of services that are web based. Most all services from Google, Yahoo, AOL are all web based.
And don't think for a second that Google doesn't struggle with supporting multiple browsers, including IE 6, which is the retarded stepchild of the web that simply won't go away.
IE 6 has no influence on mobile devices.
Until the iPhone came out, J2ME was the platform of choice for mobile developers. Every service provider supported it and J2ME was a cash cow. Companies like Ericsson are still heavily invested in it.
Looking at the quality of those apps and how widely they were used I'm not sure that is something to brag about.
"The more you tighten your grip, Ellison, the more star systems will slip through your fingers."
No, it wasn't pretty. We used to have a comp'd dev license for prototype development which Oracle nix'd. So now we do JBoss. Who lost?
I still see tons of job postings in my area that ask for WebLogic. The company I currently work for is still using it for several of their projects, and a couple of the managers were bitching that Oracle keeps trying to sign them up for more products/services and charge them more. From what I understand, Oracle made a calculated decision to lose, say, 20% of WebLogic customers and fleece the rest. Thus far, I see no indication that the other 80% have switched. So to answer your question, Oracle's customers seem to be losing so far.
But the xerox comment was more directed at Sun than Oracle. You know...the company that is now going to be owned by Oracle?
Java is a successful technology and a successful business. Granted, it's not a runaway success but by no means a failure as a business. Sun made other mistakes that did them in, most of them when Scott McNealy was running the show.
But no, Oracle won't do anything better with the technology. It's just not that sorta company.
So... technologies like BTFS don't constitute "doing better with technology"?
Besides...did you listen to Ellison at Java One? Who are you going to bet on? Google and Sergey or Oracle and Larry? JavaFX vs Ajax/Gears/GWT/HTML5?
JavaFX is a technology with a lot of promise, but still a very young technology. Sun has learned from the mistakes they made with Java on the client side (Swing, Applets, J2ME), and you can tell in the design decisions made with JavaFX. What's more, Sun has totally ignored backward compatibility consideration, as each new release of JavaFX breaks the next. Oracle can invest the resources in making this technology work.
Pure AJAX just doesn't work. There's too much configuration and boilerplate code and very little tooling. GWT is another story, but it's hampered by the fact that it has to support IE 6 and other strange browsers. Gears for local storage is well and all, but is still a bit of a hack when compared to true client-side applications. HTML 5 isn't even a complete spec yet, and each browser treats it differently. Obviously, IE 6/7/8 won't support it at all, which creates another set of problems.
On the other hand, JavaFX represents a single runtime to code against. It leverages the JVM and the rich Java API and class libraries. What's more, it's the same API (with a few differences) between mobile and desktop. The potential is there to blow the lid off Flash/Flex, because Adobe has different versions of Flash for desktop and mobile. You can write the same app for JavaFX desktop and mobile.
And, oh yeah, contrary to popular belief, the Java runtime is available on most Windows, Linux and OS X desktops. It's not as ubiquitous as Flash, but it's not that far off.
He's gonna get serious with netbooks and phones? Really? Against Android and Chrome?
JavaFX can in theory run on top of both platforms.
I read all these comments on AppleInsider about how Android will "suck", "fail", or both, and that the only mobile platform that has a snowball's chance in hell of succeeding will be the iPhone. It's as though Apple will be able to single-handedly wipe out the following platforms:
1) Blackberry
2) Nokia/Symbian
3) Windows Mobile + Hardware partners
4) Google Android + Hardware partners
5) J2ME/JavaFX + Hardware partners
Still, I'm going to bet that this time next year, one of (if not both) Windows Mobile and Symbian drop out of contention in the mobile market, since the space is getting quite crowded. Android is free. Windows Mobile isn't, and its UI is worse.
But let's focus on your "laughing stock" comment. Please justify how Android is more of a "laughing stock" than the following platforms:
1) Windows Mobile (IMO, has 0 usability beyond making bullet points for PHBs)
2) Symbian (arguably even worse than Windows Mobile
3) J2ME (which even I'll admit, while it had business backing from the get-go, is far from the most pleasant user experience)
4) Blackberry OS (while business's love it, it scores very low in user friendliness)
I haven't personally used an Android phones, but I've seen demos and it ranks right up there in usability and is a solid #2 to the iPhone, which I'll admit has by a safe margin the best user interface of all.
I suggest you actually try an Android phone, and try out some of the Windows Mobile/Symbian/Blackberry phones before declaring it a "laughing stock".
Oh yeah, and take a look at Android's partners:
http://www.openhandsetalliance.com/oha_members.html
Humm lets see... the iphone, the 2007 invention of the year, has sold 22 million copies worldwide and over 1 billion apps have been downloaded from the app store.
The android has sold a copy to Sergey, one to larry, and the CEO also has one because it would be inappropriate for him to show up to board meetings without one. Android was created by Sergey because he was getting too bored on the "bang-plane".
Need I go on? LOL!!!
This whole thread is absolutely ridiculous.
The only thing that can stand up to the iphone is blackberry.
Humm lets see... the iphone, the 2007 invention of the year, has sold 22 million copies worldwide and over 1 billion apps have been downloaded from the app store.
I never disagreed with the fact that the iPhone is a runaway success. The subject is your statement that Android is a "laughing stock". This part of your post is a non sequitur.
The android has sold a copy to Sergey, one to larry, and the CEO also has one because it would be inappropriate for him to show up to board meetings without one. Android was created by Sergey because he was getting too bored on the "bang-plane".
I suppose you have proof to back up your statement that Android was "just for fun". Of course you.
Here again is the list of partners in Android's Open Handset Alliance:
http://www.openhandsetalliance.com/oha_members.html
Need I go on? LOL!!!
Apparently you do, because you're making your case based on hyperbole and ad hominem statements. In other words, you've failed to leave out logical reasoning and facts so far.
This whole thread is absolutely ridiculous.
Since you started it by stating that Android is a "laughing stock", I wholeheartedly agree.
The only thing that can stand up to the iphone is blackberry.
Since four Android phones are being sold thus far, with many more on the way before the end of the year, I wholeheartedly disagree.
http://en.wikipedia.org/wiki/Google_android#Forthcoming
Not sure I see your point. The only other language that's even close is C.
The slope is downward not upward.
Well, actually, it's Linux that hurt them more than NT. That's their real competitor, and prompted them to port Solaris to x86 and open source it as OpenSolaris. OpenSolaris is really coming along nicely, and has a lot of customers. The problem is, they should have started that initiative at least 5 years before they did.
No. In 1997 NT server market share was 36%. Microsoft took the low end server market and workstation markets that unix makers used to own.
By the time linux became a real server OS (as in after IBM dumped an assload of IP into Linux) Microsoft had already taken a large chunk out of the mid-range server markets too with 4-way x86 boxes competing against Sparcs, PA-RISC, etc...and winning.
Sure, you can look at companies like Google and see linux as dominant...but Exchange, windows file shares and Sharepoint are what's deployed by IT for their enterprise services on those low and mid-grade servers. Shops that used to be Solaris or HPUX or AIX or IRIX or Digital Unix.
If linux killed Solaris, the pray tell why folks aren't using sendmail but Exchange? Why don't folks nfs rather than smb? Why are folks authenticated via active directory vs nds/fds?
Windows won. Unix lost. It wasn't linux that killed Solaris although it sure hastened its death. The fatal shot was fired by MS.
Linux just took away what meager share Sun still had in the lower end of the unix market. IBM did what it did because they cared more about selling services than selling servers with AIX. Solaris was beating AIX anyway so by going Linux they hurt HP and Sun far more than themselves.
Not sure I follow you. I thought we were talking about Sun's business prowess, not its technological prowess. How is supporting OpenStep a business decision? From what you're telling me, it was more a question of picking the best technology.
It was a business decision on what to focus on. The desktop or a new language. One is a user centric focus. The other a developer one.
One gives you desktop penetration and the ability to fight on MS's incursion into the server market and defend your critical workstation market hardware sales. The other, not so much.
I thought you said Sun was in the business of selling hardware, not software.
That's the whole point. OpenStep is a competitive advantage for selling Sun equipment running Solaris. Even if that hardware is commodity x86 hardware with a "Sun Tax" (vs "Apple Tax").
Java is not given it runs just as well on any JVM living anywhere.
Besides, OpenStep would have failed even with Sun, and NeXT would have still been taken over by Apple.
NEXTSTEP and OpenStep were MILES ahead of CDE and any other unix desktop on the market. It would have taken share from HPUX, AIX, IRIX. It was also better than either MacOS or Windows. It would not have failed.
Sun should have bought NeXT and pulled Jobs in. After black Tuesday Jobs would have sold if Sun offered Jobs a free hand for the workstation market.
Jobs and the NeXT guys did for unix what Jobs and the Pixar guys did for animation.
There are millions of unix users today who've never seen a command line. There are millions of unix users today thrilled with their desktop experience.
It sure as hell wasn't because of either Java or Linux.
NeXT's products lost credibility once Apple took over the company. One important example is WebObjects, which was miles ahead technologically and supported by many large corporations. Once Apple took over, it lost to Microsoft's inferior IIS/ASP framework.
More like once Jobs took over Apple, he focused on the consumer market which has always been his focus. He was precluded from competing in that market and had to do the workstation market. That WebObjects fell by the wayside was to be expected.
OpenStep was never abandoned by Jobs. It's what OSX is based on.
It would still have been a wash. IBM would have still rallied behind Linux and if the Unix desktop market existed they would have found a way to compete.
The Unix desktop market exists today. It's called OSX. It could have existed in 1998...called OpenStep on Solaris x86 at NT pricing.
And IBM wouldn't have anything to counter that.
Ultimately, Java succeeded on the server. OpenStep was purely a GUI platform, and Objective-C never took off as a server-side language.
Ultimately, Java failed to save Sun. OpenStep became the dominant unix desktop UI, and Objective-C took off on the iPhone and there are many thousands of new Objective-C developers today as a consequence. More importantly, Apple has billions in the bank while Sun is dead. OpenStep made Apple far more money than Sun has ever made from Java.
Because Java killed OpenStep interest at Sun, effectively Java killed Sun.
I was a NEXTSTEP and SunOS/Solaris developer in the early 90s. It was a crucial moment for Sun and they went down the wrong path. That path was Java.
What part of "no common runtime for developers to develop and test against" do you not understand?
Actually, the other thing you seem to forget is that Google's use of Java isn't standard, either in GWT or in Android or in ChromeOS.
So no common runtime...Android isn't J2ME. ChromeOS won't be J2SE. And Android's NDK allows C/C++. Heh...and maybe in the future they'll add ObjC.
Dalvik isn't you dad's JVM.
Actually, the other thing you seem to forget is that Google's use of Java isn't standard, either in GWT or in Android or in ChromeOS.
So no common runtime...Android isn't J2ME. ChromeOS won't be J2SE. And Android's NDK allows C/C++. Heh...and maybe in the future they'll add ObjC.
Dalvik isn't you dad's JVM.
It's standard in the sense that there aren't multiple implementations of Android to code against, as there are multiple browsers to code against. This means if I write an Android app, I only have to test it once, not multiple times as I would for a HTML/CSS/JavaScript app for each browser.
Sun's JVM supports multiple languages, and this is not a weakness. It allows developers a choice of languages to use to code against the JVM. Java contains JNI (Java Native Interface), which Google could re-implement in Android's JVM.
BTW, you're on crack if you think anybody except Apple and maybe a couple of obscure open source projects (ex Étoile) is ever going to use Objective-C. It's a language from the 80's with a bizarre syntax. Apple has obviously done amazing things with it, but the only reason it has any traction outside the company is because it's the only way to code iPhone apps currently. Apple would be wise to enhance multiple language support for Cocoa.
It's standard in the sense that there aren't multiple implementations of Android to code against, as there are multiple browsers to code against. This means if I write an Android app, I only have to test it once, not multiple times as I would for a HTML/CSS/JavaScript app for each browser.
Sure, but this is the same as saying .NET or Silverlight is standard. Well...yeah, for MS based platforms (ignore Mono and Moonlight for the moment).
Sun's JVM supports multiple languages, and this is not a weakness. It allows developers a choice of languages to use to code against the JVM. Java contains JNI (Java Native Interface), which Google could re-implement in Android's JVM.
BTW, you're on crack if you think anybody except Apple and maybe a couple of obscure open source projects (ex Étoile) is ever going to use Objective-C. It's a language from the 80's with a bizarre syntax. Apple has obviously done amazing things with it, but the only reason it has any traction outside the company is because it's the only way to code iPhone apps currently. Apple would be wise to enhance multiple language support for Cocoa.
Heh, I could say the same about being on crack if you think the multi-language support in JVMs will ever be any more significant than ObjC.
Way to go ignoring all the other points though.
Sure, but this is the same as saying .NET or Silverlight is standard. Well...yeah, for MS based platforms (ignore Mono and Moonlight for the moment).
Yes. .NET is a standard according to the point that I was making.. Developers can write to the .NET runtime and expect consistent performance on all installations. Aside from Mono, there's only one .NET runtime.
Heh, I could say the same about being on crack if you think the multi-language support in JVMs will ever be any more significant than ObjC.
They'll be just as significant as the JVM as a whole is, which is a hell of a lot more significant than Objective-C is for enterprise development.
Way to go ignoring all the other points though.
I'm responding to your points in this particular post, not your other post about the history of OpenSTEP.
The slope is downward not upward.
True, but the JVM has more than one language. The Java language has been in legacy mode since Java 5. Groovy, Scala, JPython, JRuby, and JavaFX are picking up much of the ground that Java is leaving behind. Scala in particular is interesting, since it has all the advantages of Java with more compact syntax and modern features such as closures and the ability to use domain specific languages such as XML and SQL.
No. In 1997 NT server market share was 36%. Microsoft took the low end server market and workstation markets that unix makers used to own.
By the time linux became a real server OS (as in after IBM dumped an assload of IP into Linux) Microsoft had already taken a large chunk out of the mid-range server markets too with 4-way x86 boxes competing against Sparcs, PA-RISC, etc...and winning.
Sure, you can look at companies like Google and see linux as dominant...but Exchange, windows file shares and Sharepoint are what's deployed by IT for their enterprise services on those low and mid-grade servers. Shops that used to be Solaris or HPUX or AIX or IRIX or Digital Unix.
If linux killed Solaris, the pray tell why folks aren't using sendmail but Exchange? Why don't folks nfs rather than smb? Why are folks authenticated via active directory vs nds/fds?
Windows won. Unix lost. It wasn't linux that killed Solaris although it sure hastened its death. The fatal shot was fired by MS.
Linux just took away what meager share Sun still had in the lower end of the unix market. IBM did what it did because they cared more about selling services than selling servers with AIX. Solaris was beating AIX anyway so by going Linux they hurt HP and Sun far more than themselves.
Ah...screw it! I have to admit defeat on this one.
It was a business decision on what to focus on. The desktop or a new language. One is a user centric focus. The other a developer one.
One gives you desktop penetration and the ability to fight on MS's incursion into the server market and defend your critical workstation market hardware sales. The other, not so much.
So, are you saying that Sun should have totally ignored Java when it came out of their R&D lab? Java still has a lot of tech that Objective-C doesn't have. There's a reason Apple switched WebObjects from Objective-C to Java, and I doubt that it was political.
That's the whole point. OpenStep is a competitive advantage for selling Sun equipment running Solaris. Even if that hardware is commodity x86 hardware with a "Sun Tax" (vs "Apple Tax").
Java is not given it runs just as well on any JVM living anywhere.
NEXTSTEP and OpenStep were MILES ahead of CDE and any other unix desktop on the market. It would have taken share from HPUX, AIX, IRIX. It was also better than either MacOS or Windows. It would not have failed.
Unless Sun's marketing really screwed up, which even I would admit is a possibility
Sun should have bought NeXT and pulled Jobs in. After black Tuesday Jobs would have sold if Sun offered Jobs a free hand for the workstation market.
Jobs and the NeXT guys did for unix what Jobs and the Pixar guys did for animation.
There are millions of unix users today who've never seen a command line. There are millions of unix users today thrilled with their desktop experience.
It sure as hell wasn't because of either Java or Linux.
From what I hear, Apple was dead set on buying Be before they changed their mind at the last minute, and that was just pure chance. Nobody apparently wanted NeXT as nobody really saw the point of the technology.
More like once Jobs took over Apple, he focused on the consumer market which has always been his focus. He was precluded from competing in that market and had to do the workstation market. That WebObjects fell by the wayside was to be expected.
OpenStep was never abandoned by Jobs. It's what OSX is based on.
It was abandoned as a standalone product sold/distributed to other companies. That was my point.
The Unix desktop market exists today. It's called OSX. It could have existed in 1998...called OpenStep on Solaris x86 at NT pricing.
And IBM wouldn't have anything to counter that.
Ultimately, Java failed to save Sun. OpenStep became the dominant unix desktop UI, and Objective-C took off on the iPhone and there are many thousands of new Objective-C developers today as a consequence. More importantly, Apple has billions in the bank while Sun is dead. OpenStep made Apple far more money than Sun has ever made from Java.
Because Java killed OpenStep interest at Sun, effectively Java killed Sun.
I was a NEXTSTEP and SunOS/Solaris developer in the early 90s. It was a crucial moment for Sun and they went down the wrong path. That path was Java.
There's one inconsistency in this story though. NeXT allowed an open source implementation: GNUStep. It's what WindowMaker and some other window managers use. Sure, it was GPL'd, but in theory any Unix vendor could have taken that tech and built a GUI out of it. Nobody did.
True, but the JVM has more than one language. The Java language has been in legacy mode since Java 5. Groovy, Scala, JPython, JRuby, and JavaFX are picking up much of the ground that Java is leaving behind. Scala in particular is interesting, since it has all the advantages of Java with more compact syntax and modern features such as closures and the ability to use domain specific languages such as XML and SQL.
Actually, Java 6 with finally a new LAF seems finally be moving forward a bit.
JavaFX is just bizzaro to me given how well WPF was designed by MS. When your avowed architecture is MVC and you add a new View technology, you kinda want Java to be able to cover the Model and Controller parts since it does it fairly well and there's a huge assload of legacy Java Model/Controller code tied to crappy Swing Views..
But you can't...because Java can't call into JavaFX directly except through hacks. Something that has been broken since 1.0. How on earth do consider something 1.0 when you can't do that? Beats me.
This has been the Sun Java story all along. Incredibly brilliant engineering crippled by an almost total lack of attention to what doesn't make life suck for the average developer.
Something that both Apple and MS does a much better job at IMHO.
Ah...screw it! I have to admit defeat on this one.
Yah, I am (or was since Sun is no more) a big Sun fan. If I bitch about Sun it's because they had the potential for greatness. Or at least lasting greatness. Or at least an ending not quite this pathetic.
Sun is kinda like how Apple would have ended up without Jobs coming back.
So, are you saying that Sun should have totally ignored Java when it came out of their R&D lab? Java still has a lot of tech that Objective-C doesn't have. There's a reason Apple switched WebObjects from Objective-C to Java, and I doubt that it was political.
Totally ignored? Heck no. Build the corporate strategy around it? Heck no. VB, .NET, C#, ObjC are all simply a means to an end for MS and Apple. Even C/C++ via gcc is simply a means to and end for the FSF.
Java could have been another great means to an end. The end being selling Sun hardware and not going out of business. But that requires also attention to the user, not to a language technologies for developers who are most likely on Windows.
Unless Sun's marketing really screwed up, which even I would admit is a possibility
Yah, sure.
It was abandoned as a standalone product sold/distributed to other companies. That was my point.
Sure, you don't give your competitors your competitive advantage if you have the resources to leverage it yourself. Licensing NEXTSTEP happened only because NeXT couldn't sell its own hardware in qty.
There's one inconsistency in this story though. NeXT allowed an open source implementation: GNUStep. It's what WindowMaker and some other window managers use. Sure, it was GPL'd, but in theory any Unix vendor could have taken that tech and built a GUI out of it. Nobody did.
The value of an API is far lower than the amount of software required to implement it. Sun purchased a lot of code from NeXT. I'm not certain what happened to it but it sure didn't get open sourced. I think I remember Apple buying the licenses back or Sun letting lapse or something.
The GNUStep stuff is half done at best, isn't a desktop (but a "window manager") and frankly the GPL aspects kills any incentive for commercial companies to invest in it.
It's different when you GPL your own IP. You still get to use it for your own proprietary products. Adopting someone else's GPL code kills the monetary value of any IP you attach to it.
You do so because it doesn't matter to you anymore like with IBM and it's unix technology.