Show us some data supporting your claims that Java is replacing C and C++ in areas such as shrink-wrapped desktop software development, numerical computing, systems software, or even custom desktop software development.
Firstly, I have never said that Java is significantly replacing C or C++ in the areas of shrink-wrapped desktop software, numerical computing or system software. All I have said is that I am happy to use Java for numerical work - it was one of my criteria for switching to Java. This is not to say that some people aren't using Java for numerical work: one of the examples I often point out is that one of most significant numerical research centres - based at the University of Edinburgh - now has Java as one of the primary languages for high-performance computing on its supercomputer systems. Java is totally inappropriate for system software - that is where C and C++ have their role, and they are likely to remain strong there indefinitely.
[However, there are some very interesting non-server areas where Java is replacing C and C++ significantly - embedded and realtime systems. For example, Boeing have recently stated that Java is their primary language for such systems and will replace C/C++ and Ada.]
What I have been going on and on about is Java replacing C and C++ for most development - the truly boring stuff that most developers actually do; financial coding - dumping SQL into databases and getting results back, producing reports, working out tax or stock changes, predicting future profit, tracking customer information, handling product orders, etc, etc. It used to be that all this stuff was written in C++, perhaps interfacing to legacy COBOL (or even, on some legacy IBM machines, to assembler!). Now almost all new development of this is in Java, and fortunately there are more exiting interfaces to these things - such as the web, which is where J2EE with its scalability and ease of development in areas such as high-volume request processing is the de-facto standard.
However Java is now also having at least a minor impact in client-side development (there was some transition to Java a few years ago when many unhappy VB6 developers did not like the poor upgrade path to VB.NET and C#). For evidence of this go to a job site like monster.com. Look for Jobs that specifically mention the Swing GUI. Now look for the.NET equivalent (Winforms). Surprising, isn't it? (It surprised me). Of course, these are both a very small subset of what is really going on client-side, but what matters is the trend.
But anyway I have given you evidence of replacement - I have pointed you at a range of resources. Monitor the job adverts at sides such as monster, dice, jobserve; keep a check on the TIOBE internet resource monitoring site. There is no point looking at these things once - you need to track them for months or years. I was hoping to give figures, but the TIOBE site is down right now! From memory, the decline in C++ has been significant (with C declining only a little), and this has been tracked by a rise in Java. Other languages have remained relatively static for some time (especially C#).
We do agree that C and C++ are misused and misapplied and ought to be replaced. And, in a sense, it's Java that's replacing them, just not Sun Java; it's being replace by Microsoft Java, aka C#.
This is certainly happening in some cases - but mainly client-side. Combined with VB.NET it is taking over from existing VB6 and VC++ development as those products have come to their end-of-life. This is certainly not the entire use of C#, but it is its main use as far as I can tell. This is a very good use for it, but it is hardly having much impact on Java in the area where Java is by the strongest - general server-side development.
C# has all the advantages of Java, but it adds to that platform agnosticism (instead of cross-platform zealotry) and the few low-level facilities that Java lac
Contrary to what you seem to think, programming languages aren't just a matter of fashion. Java won't be replacing C or C++ because it can't--it lacks the low-level, platfom dependent, and systems programming capabilities of C and C++, and the Java 5 design decisions have made it clear that Sun is not going to fix this. Many of the early Java adopters (including our organization) have gone back to C++ for the time being.
This is all irrelevant. Java IS replacing C and C++ right now, for the simple reason that the vast majority of C and C++ development has no need whatsoever for low-level or system programming capabilities, platform dependent or not. Although they are vital for system development, C and C++ have never been ideal languages for general purpose work, but there were no alternative languages that had the performance and reasonable cross-platform source portability of C/C++. Smalltalk seemed a possible alternative for a while in the late 80s, but fragmented into incompatible versions from different vendors. Java is now the alternative that is actually working. And, you can do an awful lot without having to resort to low levels. You can write high-performance databases (like HSQL), you can write high-performance app and web servers (Tomcat 5.5.x), and you can actually write financial and business code in a language that is safe and secure in ways that C and C++ can never be - you don't need low-level system access to write code that calculates a mortgage, or submits a query to a database, or that emits a database.
I was not an early Java adopter - I dislike fashions in the IT industry and I avoid over-hyped approaches; I waited until the language had matured in terms of robustness and performance - I was not going to give up the performance of C++ - one of my main development languages since it first arrived on the scene more than 20 years ago. (I have high performance requirements for a development language - part of my work involves serious math ). However, I found that a few years ago, Java was finally a suitable replacement. This is why Java is still growing - because I am very far from alone in this.
In any case, this discussion is pointless: you just keep quoting back irrelevant job site statistics. The question will resolve itself over the next few years; you'll see.
I get the impression that you have made your mind up, which is why you are dismissing any evidence to the contrary as 'irrelevant' without explaining why. However, I agree with your last phrase - you will indeed see.
Which part does it fail to be the definition of according to you? "Most" or "development"?
Most development means what most developers produce on a day to day basis.
Have you actually ever tried to use mobile Java applications? They're a pain to install and function poorly on any of the major mobile platforms: PPC, Palm, or the various J2ME/MIDP phones.
This has no relevance to the argument. I agree that the performance can be awful on some devices, but this does not counteract the fact that Java mobile applications are highly popular and downloaded on a vast number of devices.
"I don't like Java on mobiles" is not evidence for "Java is not present much on mobiles", no matter how correct you may be.
You usually don't write your HR or finance system at all, you buy them from a small number of established vendors. Since the generation of software that they're now selling was developed over the last half dozen years when Java was hot, yes, a lot of it is in Java.
No, you don't, unless you are a very simple organisation indeed. A key aspect of almost all HR, CRM and finance systems these days is customisability. You have some combination of pre-written code and in-house applications. All major systems from companies like Salesforce, Oracle and SAP work like this. Simple Sage Accounts software packages only work in limited cases!
You can say 'Java was hot' as many times as you like, but that won't change the fact that it is still a developing technology with expanding use.
You're making a serious mistake if you project the future of IT, corporate or otherwise, based on the current job openings on dice.com. If the future of IT were determined by job postings, we'd all still be programming in COBOL on IBM mainframes, or maybe Visual Basic on NT, the past frontrunners. By the time a platform makes it to the top of the job postings, it's already past its prime.
The thing is, Java hasn't made it to the top of the job postings yet; there are still phenomenal numbers of C and C++ jobs out there. On the client side, these are being rapidly replaced by.NET, and on the server side by Java.
Yeah, it's still you. You see that Java is well-established in some significant areas, and you falsely extrapolate future growth from it.
No. There are a wide range of statistics that allow the use of different development languages to be monitored, from job adverts to traffic on blogs.....
Java has gotten about as big as it's gonna get--competition from.NET and PHP, as well as other issues, mean it's on a plateau, followed by a slow but steady decline. Like all platforms before it.
This is clearly false - all statistics show Java use growing; it is forming an increasing proportion of projects on sites like sourceforge, and indices like TIOBE (which measures the volume of web resources) show Java growing. There are good reasons for this - a large number of projects based on C and C++ are being replaced or ported: Java's growth in recent years has been as the expense primarily of C++, and there is no sign of that stopping.
I suggest you keep track of C, C++ and Java in job sites, software sites and language resources sites (like TIOBE). Over time you will see C++ falling significantly, C falling (but less so) and Java growing, replacing them. Check back in a year, and we will see who is right!
After some more careful thought, I concede, Decaf, that you are correct on nearly all points - at least, at this moment in time.
That is very generous of you. I hope you were not too offended by my gentle mockery of your post. If you were, I apologise.
For my own part, however, as long as Sun's relationship with Java remains as it is now, I will continue to use Java only when all other more open options are exhausted. And I will also continue to work towards the ultimate success of open source, no matter what others may say, because I believe in its cause and potential. To me, it is more important to do what I think is the Right Thing, than to do what is most convenient.
I respect that view. You already have pretty good quality open source non-certified Java implementations right now. I have heard that later this year some of them will be supporting Java 1.5 features and Swing.
However, I do have a question for you. What is the alternative to Java? If you want to use something that has support from many sources, is portable, really high performance, has a cross-platform GUI, has a safe memory model, and is free or inexpensive?
I use Java because Smalltalk was too slow and too fragmented between different incompatible versions, and because C++ a maintenance nightmare (I find Java's strack traces on exceptions to be a huge benefit).
I think you're living in a corporate dreamworld. When I look at hosting providers, I find that there are 10x as many PHP providers than Java hosting providers, and having installed both of them myself, it's not hard to see why. Surveys (eg Netcraft) also suggest that ASP.NET has already overtaken JSP and servlet hosting in 2004. PHP and ASP hosting are so easy to set up and get started with that it's the obvious first choice for anybody needed to do a little bit of server-side programming. In terms of applications, on Freshmeat, there are more than 5x as many CMS's listed that are written in PHP than in Java. And, perhaps most importantly, PHP seems to be more widely used for people trying new ideas on the web, including many Web 2.0 applications.
I think you are living in a OSS dreamworld. Numbers of applications does not, and never has, equalled volume of development. PHP, although it can be used for more substantial applications, is ideal for simple page scripting and hosting - the requirements of perhaps the majority of simple websites; but that is not what most of the hundreds of thousands of developers employed writing server side generally do. They write scalable middleware that links to database. They deploy on J2EE.
"a little bit of server-side programming"
Exactly! A little bit! But most server side code is not 'a little bit' - you usually don't write your HR or finance system in PHP.
Java clearly has a good chunk of the high end of server-side development: banks and corporations use it a lot, and it will stay entrenched there for another decade or two. But that's a small part of overall server-side development. Most server-side development is probably developing small scripts, putting together a few existing components, or small customizations of CMSs.
That is not the definition of 'most development'. I do server-side scripting, but that scripting does not form anything but the smallest part of the coding that I do, and it does not form anything but the smallest part of what goes on in general server side development. The vast majority of lines of code written are in Java, and this is not limited to banks and corporations - Java is widely used by small development teams and consultants in all areas of IT. Java is certainly not limited to the high end. There are very widely used Java tools like Hibernate that allow agile development of even the smallest applications. In fact the biggest use of Java of all is at the lowest end - mobile applications.
You really do need to have a deeper look at what is going on in the IT world. Lets have a look at the jobs on dice.com: out of about 90,000 jobs....
You're dancing around the fact that at present, there are no cerified open source versions. Java has been around for many years now. Why the continued delay?
I am not dancing around anything; you have not responded to the points I made. I really don't know exactly why there are no certified open source versions, although I have a good idea - where is the need? For the vast majority of developers they can get a fully compatible JRE or JDK for free any time they want. Unfair as this is, there probably aren't enough users of unsupported systems for OSS developers to get the resources to develop a compatible JRE. On the other hand, Kaffe/GNU Classpath is making fast progress.
Trust me on this one -- the only reason you think of Sun when you think of Open Systems is that Sun's proprietary products never gained traction because they never had enough marketshare. Remember NeWS? How about NIS+? OpenLook, anybody?
Sorry, but I don't trust you on this one - I was there at the time:)
You can pick up one or two things that were proprietary and did not take off, but that is completely missing the point. Sun was pushing UNIX as a whole - the idea of using a standards-compliant systems like UNIX was still far from mainstream at the time. Also, you seem to be forgetting APIs like NFS that Sun donated freely to the community.
Sun did a lot, but what about all of the other Unix vendors: DEC, IBM, HP, and SGI?
At the time most of them were trying to push closed systems. IBM did not join the UNIX bandwagon until much later. DEC was still hugely promoting VMS. SGI was about the only other one actually fully into UNIX at the time.
Why the constant attempts to revise history to erase the huge beneficial influence of Sun?
What about my NetBSD/mac68k machine? How about my NetBSD/mips machine? I won't even raise the possibility of my SGI IRIX machine.
Nope. Go ahead and praise the virtues of java if you like. Don't pooh-pooh the issue of it's closed nature and think you will get away with it.
What closed nature? The only thing that is closed is the source code of Sun's implementation. To call Java closed because of this is as silly as to call C closed because there are some proprietary compilers.
Java is documented, as is the VM and the standard libraries. There are many open-source (but not certified) implementations. There are projects to build certified open source versions.
Yes, I usually try to figure out what sites run. For most of the sites I use, I succeed and it's not Java.
Well you are not typical then.
I agree that Java is widely used for internal desktop application development and for some server side development. I believe you are overestimating the economic importance of those areas, or the hold Java has on them.
'Some' server side development? Sorry, but it sounds like you are wildly out of touch with what most software developers are up to.
The majority of server side development is done in Java.
As for overestimating the economic importance, that is very hard to do. For most corporations Java is now a key part of their financial infrastructure and their data processing. Java is now the standard way to write the 'middleware' that interfaces to corporate databases. J2EE has become the standard platform for high-performance, high-reliability server-side infrastructure, and forms the foundation of a large number of financial and trading systems, with trading systems often handing millions or even hundreds of millions of transactions a day. Java has now become a standard way to develop Customer Relationship Management systems - running all those call centres... Salesforce, one such system, and one of the most important software products, is built using Java.
So no, I don't think I am overestimating the economic important of Java.
McNealy was trying to ride the wave that Microsoft was, at the time, willfully blind to. He showed more vision than Microsoft. But Sun shouldn't get the credit for creating that wave. The Internet had been around for a while, and was going to burst on the commercial and public scene in a big way, thanks to many factors, of which Sun was just a small part.
No, Sun was not just a small part. Sun was a dominant part of the promotion of 'Open Systems' in the 1980s - encouraging the use of UNIX with documented and standard protocols. It had quite a battle, with vendors such as IBM attempting to encourage use of closed and proprietary systems that tied IT installations in to one vendor (sound familiar?).
Sun had a great vision - that wide use of compatible systems would allow customers to mix products from different vendors and combine hardware and software. This would create a large market for standard systems that Sun could compete in. They helped grow this market by freely donating standards such as NFS to the community.
It was this use of standard and largely compatible systems based on common software (C/C++, UNIX) that helped provide the basis for the growth of open source, and later, Linux.
(Microsoft, meanwhile, had their collective heads in the sand, or rather, in their hindquarters, trying to deny that this potential Windows-dominance threat was anything worth thinking about. Remember when they thought MSN was an *alternative* to the Internet? Anything they don't utterly control, they hate.)
Nothing has changed here.
It is true that for a long time, Java was one of the all-important buzzwords, but it didn't pan out quite as well as it might have.
It is the dominant language serverside, growing for corporate client-side development, has rapidly growing use for open source development (check out Sourceforge projects), on just about every mobile phone, turning up in cars, and aeronautics, and even helping to control Mars missions.
It has done extremely well, and it is far, far more than 'one of the all-important buzzwords'.
Sun was important, but not *that* important. CERN was far more important....
The web was based on the open systems and protocols that Sun was a major factor in promoting.
Who in their right mind would proclaim Linux to be the ultimate game platform. Not even Windows is the ultimate game platform. For games you need (want) to squeeze out every ounce of performance out of the plaform. You don't really want an OS. You want more of a graphics/sound library on top of a lean microkernel. Just like the PS2. Right?
For a start, I was proclaiming Linux to be a major games platform as an extreme point of view. I wasn't seriously suggesting it was. But I am not sure about the rest of what you say. Why not have an OS - especially one that provides things that are useful for many games, like storage APIs and GUI APIs? Most of the performance you want is numeric and graphic. All good OSes today do a good job of getting out of the way when you need to do that, so whether or not you are running on a rich operating system would seem to me to be largely irrelevant.
It's more than a little funny that, thanks to Mono, the newest version of C# will run on my laptop (Linux/PPC) sooner than the new version of Java will.
I've done a fair amount of J2EE and J2SE development (never a full blown ejb aplication, just servlets/jsps/jstl etc), but aren't those both just application servers? I guess they might be comparable to say J2EE is to JBoss/GlassFish as J2SE is to Sun/Ibm/?kaffe? JVM's. And there are open source JVM's and Open source Application servers. But *Java* isn't really open source and I don't beleive J2EE really is in that sense either, though I may be wrong.
You are mostly right, because, of course, a J2EE implementation is unusable without a J2SE implementation to run it!
What else has Microsoft copied? The WinAPI, nope, they created it from scratch
And what a professional job they did, as revealed in the name of API calls that they thought they would never have to make public..... such as 'PrestoChangoSelector'
Java has grown into a complex mess, and cators or organizations that are complex messes. It tries to use buzzwords and fancy-sounding hoopla to protect itself from reality, but the reality is that it is a big fat mess.
No, it hasn't. Apart from the addition of generics and annotations in Java 1.5, very little has been added to the language in years. There is an increasing range of libraries and tools that you can use, but they are all optional - you can still code small apps with VI if you like. Java has deliberately avoided the traps that other languages have fallen into of frequently adding new syntax and features.
It takes 3 times longer to write and maintain Java code, and this is why it is a jobs machine.
3 times longer that what? Have you actually used NetBeans with its Matisse GUI designer? Have you used Eclipse, with its phenomenal code maintenance and refactoring tools?
Good languages/tools actually kill jobs. Thus bragging that Sun stuff increases jobs may be true, but does not necessarily bode well for their ideas or technology.
Ah - because development teams are so dumb that they actively take on tools and languages that make things slower to develop and harder to maintain? Nonsense, of course. Java is largely replacing C++ for commercial development because it is far faster to develop with and maintain. There are few porting issues, memory management is phenomenally easier, programs are safer, and there are standard deployment platforms like J2EE.
Let's see: there is one Java application I use regularly (and it looks like shit). I haven't seen a Java applet in years, I have not bothered to install Java on any machine that it didn't ship on, and I rarely see Java-powered web servers. Seems pretty obscure to me. Tell us: where exactly do you think Java is going?
Perhaps you could explain how you know you are rarely seeing Java-powered web servers - some interesting techniques to probe the structure of the HTML? They are everywhere. Go to EBay. See that logo at the top right of the screen? 'Java Technology'. Java used by other sites is less obvious. Sometimes you can indeed see where Java is being used, by the 'struts' or 'jsp' type URLs. From this I can see that my bank, one of the biggest in the UK, used Java for its website.
But anyway, why are you judging Java by whether or not it is being used for desktop applications, or applets, that you have personally heard of? Have you heard of Delphi? It is has been a major development platform for desktop applications for over a decade, but I can't name a single desktop application I have used that is written in it. Does this mean it is not successful? Of course not. The shrink-wrapped desktop application market is totally insignificant compared to the number of applications developed for internal desktop use within companies, and even that is negligible compared with server side development. Java use is growing in the first area (with IDEs such as Eclipse and NetBeans often being used a Rich Client platforms ), and hugely dominant (but still growing) in the second. Don't believe me? Check out dice.com or monster.com for jobs, and see where the development effort is going on.
I humbly suggest you are wrong in you statement. If you are doing anything commercial on the web, you are most likely using Java applications all the time.
Even though they did little but stand in the corner with their vision?
What a distorted view of IT history!
Sun certainly did not 'do little but stand in the corner'. They were vigorous promoters of Open Systems in the 80s, working hard to encourage the use of UNIX and standard protocols against the attempts of competitors (like IBM) to keep things closed and entirely proprietary. The resulting wide use of UNIX and standards helped the original development of open source, and then Linux.
Because it's supposed to be 'write once, run anywhere.' At the moment, it's write once, and run on Solaris, Windows, Linux (x86 or IA64 only), Mac if you don't mind waiting a bit, and maybe IRIX if you are lucky.
You have missed out a large number of important platforms. For example IBM provides certified VMs to run on AMD 64-bit, Power 64-bit, z-Zeries 31-bit, zZeries 64-bit. HP provides other platforms, and I don't know what you are supposed to be waiting for on the Mac.
Only very recently have FreeBSD been allowed to distribute the port to their OS. If you look somewhere slightly more obscure, like OpenBSD, then you start to have problems.
The license means that OpenBSD are not allowed to distribute a ported version of Java. They can distribute diffs to the source code, but not the source or compiled binaries. Since Java requires Java to build, installing Java on OpenBSD requires following these steps:
1. Install the Linux JDK, which runs through the binary compatibility layer.
2. Download the source from Sun (manually).
3. Compile the JDK. This takes a long time and requires a lot of RAM.
Since the Linux JDK is only supported on i386, you can't use Java on any other architecture (such as PowerPC or SPARC).
I would much rather Sun used the trademark to protect Java. Make the JDK open source, but do not allow any patched versions to be called Java (so OpenBSD, for example, could include a binary of 'Columbian' that would run Java apps, but would not be called Java). Provide a mechanism for pushing patches upstream, so that if someone does port it to a new platform there is a good chance that the next release will actually support that platform.
To be honest, it's the second of these that is the clincher. The number of hoops the FreeBSD team had to jump to in order to be allowed to ship Java with their OS was insane.
This only applies if you want to use Sun's source code. There is nothing to stop an independent open source version of Java being developed and certified; indeed, the Apache Harmony project is doing exactly that.
Sorry, but this sort of post does seem to be typical of a highly demanding attitude from open source supporters:
First Sun releases Java.... at no cost. Then there are demands for it to be on more platforms. Sun provides it on more platforms. Now there is a demand for Sun to release their source code! What is the big deal? Projects like Mono seemed to have little difficulty re-implementing the basic.NET architecture with reasonable performance - why the hold-up with Java? Perhaps it is because there are already free (as in beer) versions of Java for the majority of developers, there is not such much demand.
Java has several Free Software implementations, and the only real limitation they have is that they can't use the Java trademark. Kaffe, after a long absense, is back and is already mostly 1.4 compliant, GCJ takes an entirely different approach to implementing Java which is both innovative and supportive of what already exists.
And there is Harmony, an Apache project to create an open source Java platform which is certified as Java compliant, so can use the trademark.
Why? This seems to be a popular opinion on Slashdot, and I'm curious why people need it to be any more open than it is? I mean afaik, the only thing that isn't "open" about it is the spec. If you want to create your own implementation of a JVM you're allowed but it must conform to the spec. This is a very *good* thing IMO. It would really suck if MS had been able to complete its "embrace and extend" manuever on Java (which is what MS has done with the open web standards and browsers) and it would suck even more if there were 5 different JVM's out there and you had to tailor your code to run on each one. You would completely lose the WORA (or you'd have to do all sorts of gimmicky crap to figure out what jvm you were running on -thats a lot of fun with browsers and html, I think it would be even more annoying with code). So I ask again, not rhetorically, but honestly: why open source it? Am I missing something?
I absolutely agree that having conformant java specifications is a great thing, and is perhaps the single most important reason for Java's success. But I don't see why being open source should conflict with this - there should be no reason why an open source product should not have to pass the tests in order to be called 'Java'. In fact there is a current project, called 'Harmony', that intends to do exactly this. Open source need not permit 'embrace and extend'.
I can understand that there are potential problems with open sourcing Sun's implementation of Java - there are most likely huge amounts of code that involve patented techniques or are licensed from other sources.
I like open source (or at least having the source) because I have had to deal with problems in closed source products that won't be fixed by the vendor. I am not after re-selling the product, or re-distributing the source, but the possibility of patching something myself is pretty appealing.
I would lay claim that Perl and PHP power more online transactions than Java does. Look at the world's most popular forum system, PHPBB. Yes, that runs on PHP. And people don't run LAMJ servers, they run LAMP servers: the P stands for PHP or Perl or Python, all of them open languages.
And I would lay claim that they really don't. You can combine all the online forums you like but they don't come near the phenomenal combined volume of stock trading systems, banking systems, airline booking systems etc. We are talking of system which individually handle hundreds of millions or even billions of transactions each day. Consider the combined volume of transactions of all these systems....
And before you mention google - that uses a considerable amount of Java as well.
I didn't say that Java is now unpopular in all domains, that is false. But I think that it will degrade because it will not have the ability to adapt like other languages can. If Sun goes down, methinks it would be all over for Java.
Java is constantly adapting, with regular releases with new features (well, new to Java anyway!): Generics, improved concurrency and higher performance for the GUI in Java 5; scripting language support and web services support and far better client side integration in Java 6. How is this not adapting?
Apart from the wild idea that Sun is going down (their annual losses are trival compared to their net worth, and that worth is not largely dependent on share value), there are companies with far, far bigger investments in Java than Sun, like IBM. They are constantly producing new VMs for internal and external uses.
If Sun did 'go down', Java would certainly continue (in fact, IBM could well buy up the rights and open source it!). That is one of the reasons why I find it such an appealing language - it is not a one-vendor language.
And if you want to chide somebody for wanting to overcome the competition, fine. But don't forget that the origin of all open projects is the desire to build a better product, and it's only because we want to be better that we can achieve that. Wanting something is the root cause for it happening. That's not a guarantee, but it's as close as we can get.
I was not chiding anyone for wanting anything. What I was gently ridiculing is a Slashdot speciality - stating what someone might want (for whatever reasons) as if it has already happened.
I want better products - I would rather that more people adopted MacOS that Windows. I wish I could play more games on Linux. I would prefer Java to be open source. However, we have to face reality.
There is not the slightest sign (at least yet) that Java has stopped growing in terms of its adoption - it is still in the growth phase. This may change in a few years, but to say now that 'Java will remain on the path to obscurity' is ridiculous in many ways - it implies not only that Java is going to be obscure, but it is already on that path, which is obviously false.
Like it or not, Java works, and works very well for a very large number of developers. It would be nice if it were open source, I agree, but it seems to me that its current status has had little impact on its adoption, no matter now much open source supporters may wish otherwise.
No amount of marketing can change this: if Java is not sufficiently opened, it will remain on the path to obscurity.
I love comments like this! They indicate what a strange reality some slashdotters live in - it almosts make me believe in parallel universes.
I eagerly await other posts from this other dimension:
"Intel - will it ever take off?"
"Windows - how it lost out to Apple"
"Linux - the ultimate game platform".
Actually I guess the message here is that no matter how much you really, really want something to be be true (Java on the decline) this does not make it true.
As an old c/c++ unix programmer who is forced to work in java I see little real inovation in java. What started out as a highlevel embedded language has morphed into a text processing language for web development. Java is just differnet not better or visionary.
Apart from the fact that Java is used for far more than web development (it is becoming dominant in areas such as embedded and realtime code - an area that interests me), I agree that Java in itself is not that innovative. But what it does is combine the individual strengths of so many other languages in one package: Garbage collection, the use of a really high-performance VM for runtime portability, a robust security manager, built-in threading support at the core of the language, a cross-platform GUI shipped as standard, unicode as standard etc. And, it has strong multi-vendor support and is free (as in beer).
To dismiss Java as 'not innovative' is to miss the point.
Java, rebuilding the wheel one spoke at a time.
Yes, but it is a better wheel - one that goes faster and runs on more roads that most of the others.....
Slashdot Mirror
Show us some data supporting your claims that Java is replacing C and C++ in areas such as shrink-wrapped desktop software development, numerical computing, systems software, or even custom desktop software development.
.NET equivalent (Winforms). Surprising, isn't it? (It surprised me). Of course, these are both a very small subset of what is really going on client-side, but what matters is the trend.
Firstly, I have never said that Java is significantly replacing C or C++ in the areas of shrink-wrapped desktop software, numerical computing or system software. All I have said is that I am happy to use Java for numerical work - it was one of my criteria for switching to Java. This is not to say that some people aren't using Java for numerical work: one of the examples I often point out is that one of most significant numerical research centres - based at the University of Edinburgh - now has Java as one of the primary languages for high-performance computing on its supercomputer systems. Java is totally inappropriate for system software - that is where C and C++ have their role, and they are likely to remain strong there indefinitely.
[However, there are some very interesting non-server areas where Java is replacing C and C++ significantly - embedded and realtime systems. For example, Boeing have recently stated that Java is their primary language for such systems and will replace C/C++ and Ada.]
What I have been going on and on about is Java replacing C and C++ for most development - the truly boring stuff that most developers actually do; financial coding - dumping SQL into databases and getting results back, producing reports, working out tax or stock changes, predicting future profit, tracking customer information, handling product orders, etc, etc. It used to be that all this stuff was written in C++, perhaps interfacing to legacy COBOL (or even, on some legacy IBM machines, to assembler!). Now almost all new development of this is in Java, and fortunately there are more exiting interfaces to these things - such as the web, which is where J2EE with its scalability and ease of development in areas such as high-volume request processing is the de-facto standard.
However Java is now also having at least a minor impact in client-side development (there was some transition to Java a few years ago when many unhappy VB6 developers did not like the poor upgrade path to VB.NET and C#). For evidence of this go to a job site like monster.com. Look for Jobs that specifically mention the Swing GUI. Now look for the
But anyway I have given you evidence of replacement - I have pointed you at a range of resources. Monitor the job adverts at sides such as monster, dice, jobserve; keep a check on the TIOBE internet resource monitoring site. There is no point looking at these things once - you need to track them for months or years. I was hoping to give figures, but the TIOBE site is down right now! From memory, the decline in C++ has been significant (with C declining only a little), and this has been tracked by a rise in Java. Other languages have remained relatively static for some time (especially C#).
We do agree that C and C++ are misused and misapplied and ought to be replaced. And, in a sense, it's Java that's replacing them, just not Sun Java; it's being replace by Microsoft Java, aka C#.
This is certainly happening in some cases - but mainly client-side. Combined with VB.NET it is taking over from existing VB6 and VC++ development as those products have come to their end-of-life. This is certainly not the entire use of C#, but it is its main use as far as I can tell. This is a very good use for it, but it is hardly having much impact on Java in the area where Java is by the strongest - general server-side development.
C# has all the advantages of Java, but it adds to that platform agnosticism (instead of cross-platform zealotry) and the few low-level facilities that Java lac
Contrary to what you seem to think, programming languages aren't just a matter of fashion. Java won't be replacing C or C++ because it can't--it lacks the low-level, platfom dependent, and systems programming capabilities of C and C++, and the Java 5 design decisions have made it clear that Sun is not going to fix this. Many of the early Java adopters (including our organization) have gone back to C++ for the time being.
This is all irrelevant. Java IS replacing C and C++ right now, for the simple reason that the vast majority of C and C++ development has no need whatsoever for low-level or system programming capabilities, platform dependent or not. Although they are vital for system development, C and C++ have never been ideal languages for general purpose work, but there were no alternative languages that had the performance and reasonable cross-platform source portability of C/C++. Smalltalk seemed a possible alternative for a while in the late 80s, but fragmented into incompatible versions from different vendors. Java is now the alternative that is actually working. And, you can do an awful lot without having to resort to low levels. You can write high-performance databases (like HSQL), you can write high-performance app and web servers (Tomcat 5.5.x), and you can actually write financial and business code in a language that is safe and secure in ways that C and C++ can never be - you don't need low-level system access to write code that calculates a mortgage, or submits a query to a database, or that emits a database.
I was not an early Java adopter - I dislike fashions in the IT industry and I avoid over-hyped approaches; I waited until the language had matured in terms of robustness and performance - I was not going to give up the performance of C++ - one of my main development languages since it first arrived on the scene more than 20 years ago. (I have high performance requirements for a development language - part of my work involves serious math ). However, I found that a few years ago, Java was finally a suitable replacement. This is why Java is still growing - because I am very far from alone in this.
In any case, this discussion is pointless: you just keep quoting back irrelevant job site statistics. The question will resolve itself over the next few years; you'll see.
I get the impression that you have made your mind up, which is why you are dismissing any evidence to the contrary as 'irrelevant' without explaining why. However, I agree with your last phrase - you will indeed see.
Which part does it fail to be the definition of according to you? "Most" or "development"?
.NET, and on the server side by Java.
.NET and PHP, as well as other issues, mean it's on a plateau, followed by a slow but steady decline. Like all platforms before it.
Most development means what most developers produce on a day to day basis.
Have you actually ever tried to use mobile Java applications? They're a pain to install and function poorly on any of the major mobile platforms: PPC, Palm, or the various J2ME/MIDP phones.
This has no relevance to the argument. I agree that the performance can be awful on some devices, but this does not counteract the fact that Java mobile applications are highly popular and downloaded on a vast number of devices.
"I don't like Java on mobiles" is not evidence for "Java is not present much on mobiles", no matter how correct you may be.
You usually don't write your HR or finance system at all, you buy them from a small number of established vendors. Since the generation of software that they're now selling was developed over the last half dozen years when Java was hot, yes, a lot of it is in Java.
No, you don't, unless you are a very simple organisation indeed. A key aspect of almost all HR, CRM and finance systems these days is customisability. You have some combination of pre-written code and in-house applications. All major systems from companies like Salesforce, Oracle and SAP work like this. Simple Sage Accounts software packages only work in limited cases!
You can say 'Java was hot' as many times as you like, but that won't change the fact that it is still a developing technology with expanding use.
You're making a serious mistake if you project the future of IT, corporate or otherwise, based on the current job openings on dice.com. If the future of IT were determined by job postings, we'd all still be programming in COBOL on IBM mainframes, or maybe Visual Basic on NT, the past frontrunners. By the time a platform makes it to the top of the job postings, it's already past its prime.
The thing is, Java hasn't made it to the top of the job postings yet; there are still phenomenal numbers of C and C++ jobs out there. On the client side, these are being rapidly replaced by
Yeah, it's still you. You see that Java is well-established in some significant areas, and you falsely extrapolate future growth from it.
No. There are a wide range of statistics that allow the use of different development languages to be monitored, from job adverts to traffic on blogs.....
Java has gotten about as big as it's gonna get--competition from
This is clearly false - all statistics show Java use growing; it is forming an increasing proportion of projects on sites like sourceforge, and indices like TIOBE (which measures the volume of web resources) show Java growing. There are good reasons for this - a large number of projects based on C and C++ are being replaced or ported: Java's growth in recent years has been as the expense primarily of C++, and there is no sign of that stopping.
I suggest you keep track of C, C++ and Java in job sites, software sites and language resources sites (like TIOBE). Over time you will see C++ falling significantly, C falling (but less so) and Java growing, replacing them. Check back in a year, and we will see who is right!
After some more careful thought, I concede, Decaf, that you are correct on nearly all points - at least, at this moment in time.
That is very generous of you. I hope you were not too offended by my gentle mockery of your post. If you were, I apologise.
For my own part, however, as long as Sun's relationship with Java remains as it is now, I will continue to use Java only when all other more open options are exhausted. And I will also continue to work towards the ultimate success of open source, no matter what others may say, because I believe in its cause and potential. To me, it is more important to do what I think is the Right Thing, than to do what is most convenient.
I respect that view. You already have pretty good quality open source non-certified Java implementations right now. I have heard that later this year some of them will be supporting Java 1.5 features and Swing.
However, I do have a question for you. What is the alternative to Java? If you want to use something that has support from many sources, is portable, really high performance, has a cross-platform GUI, has a safe memory model, and is free or inexpensive?
I use Java because Smalltalk was too slow and too fragmented between different incompatible versions, and because C++ a maintenance nightmare (I find Java's strack traces on exceptions to be a huge benefit).
What would you use?
I think you're living in a corporate dreamworld. When I look at hosting providers, I find that there are 10x as many PHP providers than Java hosting providers, and having installed both of them myself, it's not hard to see why. Surveys (eg Netcraft) also suggest that ASP.NET has already overtaken JSP and servlet hosting in 2004. PHP and ASP hosting are so easy to set up and get started with that it's the obvious first choice for anybody needed to do a little bit of server-side programming. In terms of applications, on Freshmeat, there are more than 5x as many CMS's listed that are written in PHP than in Java. And, perhaps most importantly, PHP seems to be more widely used for people trying new ideas on the web, including many Web 2.0 applications.
I think you are living in a OSS dreamworld. Numbers of applications does not, and never has, equalled volume of development. PHP, although it can be used for more substantial applications, is ideal for simple page scripting and hosting - the requirements of perhaps the majority of simple websites; but that is not what most of the hundreds of thousands of developers employed writing server side generally do. They write scalable middleware that links to database. They deploy on J2EE.
"a little bit of server-side programming"
Exactly! A little bit! But most server side code is not 'a little bit' - you usually don't write your HR or finance system in PHP.
Java clearly has a good chunk of the high end of server-side development: banks and corporations use it a lot, and it will stay entrenched there for another decade or two. But that's a small part of overall server-side development. Most server-side development is probably developing small scripts, putting together a few existing components, or small customizations of CMSs.
That is not the definition of 'most development'. I do server-side scripting, but that scripting does not form anything but the smallest part of the coding that I do, and it does not form anything but the smallest part of what goes on in general server side development. The vast majority of lines of code written are in Java, and this is not limited to banks and corporations - Java is widely used by small development teams and consultants in all areas of IT. Java is certainly not limited to the high end. There are very widely used Java tools like Hibernate that allow agile development of even the smallest applications. In fact the biggest use of Java of all is at the lowest end - mobile applications.
You really do need to have a deeper look at what is going on in the IT world. Lets have a look at the jobs on dice.com: out of about 90,000 jobs....
Java 15,354
J2EE 6,905
PHP 1093
(Oh, and the much-hyped Ruby on Rails: 38!)
I think that clarifies who is in the dreamworld!
You're dancing around the fact that at present, there are no cerified open source versions. Java has been around for many years now. Why the continued delay?
I am not dancing around anything; you have not responded to the points I made. I really don't know exactly why there are no certified open source versions, although I have a good idea - where is the need? For the vast majority of developers they can get a fully compatible JRE or JDK for free any time they want. Unfair as this is, there probably aren't enough users of unsupported systems for OSS developers to get the resources to develop a compatible JRE. On the other hand, Kaffe/GNU Classpath is making fast progress.
Trust me on this one -- the only reason you think of Sun when you think of Open Systems is that Sun's proprietary products never gained traction because they never had enough marketshare. Remember NeWS? How about NIS+? OpenLook, anybody?
:)
Sorry, but I don't trust you on this one - I was there at the time
You can pick up one or two things that were proprietary and did not take off, but that is completely missing the point. Sun was pushing UNIX as a whole - the idea of using a standards-compliant systems like UNIX was still far from mainstream at the time. Also, you seem to be forgetting APIs like NFS that Sun donated freely to the community.
Sun did a lot, but what about all of the other Unix vendors: DEC, IBM, HP, and SGI?
At the time most of them were trying to push closed systems. IBM did not join the UNIX bandwagon until much later. DEC was still hugely promoting VMS. SGI was about the only other one actually fully into UNIX at the time.
Why the constant attempts to revise history to erase the huge beneficial influence of Sun?
There are also lots of COBOL jobs; that doesn't mean the platform is doing well...
There aren't lots of COBOL jobs.
What about my NetBSD/mac68k machine? How about my NetBSD/mips machine? I won't even raise the possibility of my SGI IRIX machine.
Nope. Go ahead and praise the virtues of java if you like. Don't pooh-pooh the issue of it's closed nature and think you will get away with it.
What closed nature? The only thing that is closed is the source code of Sun's implementation. To call Java closed because of this is as silly as to call C closed because there are some proprietary compilers.
Java is documented, as is the VM and the standard libraries. There are many open-source (but not certified) implementations. There are projects to build certified open source versions.
Yes, I usually try to figure out what sites run. For most of the sites I use, I succeed and it's not Java.
Well you are not typical then.
I agree that Java is widely used for internal desktop application development and for some server side development. I believe you are overestimating the economic importance of those areas, or the hold Java has on them.
'Some' server side development? Sorry, but it sounds like you are wildly out of touch with what most software developers are up to.
The majority of server side development is done in Java.
As for overestimating the economic importance, that is very hard to do. For most corporations Java is now a key part of their financial infrastructure and their data processing. Java is now the standard way to write the 'middleware' that interfaces to corporate databases. J2EE has become the standard platform for high-performance, high-reliability server-side infrastructure, and forms the foundation of a large number of financial and trading systems, with trading systems often handing millions or even hundreds of millions of transactions a day. Java has now become a standard way to develop Customer Relationship Management systems - running all those call centres... Salesforce, one such system, and one of the most important software products, is built using Java.
So no, I don't think I am overestimating the economic important of Java.
McNealy was trying to ride the wave that Microsoft was, at the time, willfully blind to. He showed more vision than Microsoft. But Sun shouldn't get the credit for creating that wave. The Internet had been around for a while, and was going to burst on the commercial and public scene in a big way, thanks to many factors, of which Sun was just a small part.
No, Sun was not just a small part. Sun was a dominant part of the promotion of 'Open Systems' in the 1980s - encouraging the use of UNIX with documented and standard protocols. It had quite a battle, with vendors such as IBM attempting to encourage use of closed and proprietary systems that tied IT installations in to one vendor (sound familiar?).
Sun had a great vision - that wide use of compatible systems would allow customers to mix products from different vendors and combine hardware and software. This would create a large market for standard systems that Sun could compete in. They helped grow this market by freely donating standards such as NFS to the community.
It was this use of standard and largely compatible systems based on common software (C/C++, UNIX) that helped provide the basis for the growth of open source, and later, Linux.
(Microsoft, meanwhile, had their collective heads in the sand, or rather, in their hindquarters, trying to deny that this potential Windows-dominance threat was anything worth thinking about. Remember when they thought MSN was an *alternative* to the Internet? Anything they don't utterly control, they hate.)
Nothing has changed here.
It is true that for a long time, Java was one of the all-important buzzwords, but it didn't pan out quite as well as it might have.
It is the dominant language serverside, growing for corporate client-side development, has rapidly growing use for open source development (check out Sourceforge projects), on just about every mobile phone, turning up in cars, and aeronautics, and even helping to control Mars missions.
It has done extremely well, and it is far, far more than 'one of the all-important buzzwords'.
Sun was important, but not *that* important. CERN was far more important....
The web was based on the open systems and protocols that Sun was a major factor in promoting.
Who in their right mind would proclaim Linux to be the ultimate game platform. Not even Windows is the ultimate game platform. For games you need (want) to squeeze out every ounce of performance out of the plaform. You don't really want an OS. You want more of a graphics/sound library on top of a lean microkernel. Just like the PS2. Right?
For a start, I was proclaiming Linux to be a major games platform as an extreme point of view. I wasn't seriously suggesting it was. But I am not sure about the rest of what you say. Why not have an OS - especially one that provides things that are useful for many games, like storage APIs and GUI APIs? Most of the performance you want is numeric and graphic. All good OSes today do a good job of getting out of the way when you need to do that, so whether or not you are running on a rich operating system would seem to me to be largely irrelevant.
It's more than a little funny that, thanks to Mono, the newest version of C# will run on my laptop (Linux/PPC) sooner than the new version of Java will.
IBM already have a Java 1.5 VM for Linux/PPC.
I've done a fair amount of J2EE and J2SE development (never a full blown ejb aplication, just servlets/jsps/jstl etc), but aren't those both just application servers? I guess they might be comparable to say J2EE is to JBoss/GlassFish as J2SE is to Sun/Ibm/?kaffe? JVM's. And there are open source JVM's and Open source Application servers. But *Java* isn't really open source and I don't beleive J2EE really is in that sense either, though I may be wrong.
You are mostly right, because, of course, a J2EE implementation is unusable without a J2SE implementation to run it!
What else has Microsoft copied? The WinAPI, nope, they created it from scratch
And what a professional job they did, as revealed in the name of API calls that they thought they would never have to make public..... such as 'PrestoChangoSelector'
Java has grown into a complex mess, and cators or organizations that are complex messes. It tries to use buzzwords and fancy-sounding hoopla to protect itself from reality, but the reality is that it is a big fat mess.
No, it hasn't. Apart from the addition of generics and annotations in Java 1.5, very little has been added to the language in years. There is an increasing range of libraries and tools that you can use, but they are all optional - you can still code small apps with VI if you like. Java has deliberately avoided the traps that other languages have fallen into of frequently adding new syntax and features.
It takes 3 times longer to write and maintain Java code, and this is why it is a jobs machine.
3 times longer that what? Have you actually used NetBeans with its Matisse GUI designer? Have you used Eclipse, with its phenomenal code maintenance and refactoring tools?
Good languages/tools actually kill jobs. Thus bragging that Sun stuff increases jobs may be true, but does not necessarily bode well for their ideas or technology.
Ah - because development teams are so dumb that they actively take on tools and languages that make things slower to develop and harder to maintain? Nonsense, of course. Java is largely replacing C++ for commercial development because it is far faster to develop with and maintain. There are few porting issues, memory management is phenomenally easier, programs are safer, and there are standard deployment platforms like J2EE.
Let's see: there is one Java application I use regularly (and it looks like shit). I haven't seen a Java applet in years, I have not bothered to install Java on any machine that it didn't ship on, and I rarely see Java-powered web servers. Seems pretty obscure to me. Tell us: where exactly do you think Java is going?
Perhaps you could explain how you know you are rarely seeing Java-powered web servers - some interesting techniques to probe the structure of the HTML? They are everywhere. Go to EBay. See that logo at the top right of the screen? 'Java Technology'. Java used by other sites is less obvious. Sometimes you can indeed see where Java is being used, by the 'struts' or 'jsp' type URLs. From this I can see that my bank, one of the biggest in the UK, used Java for its website.
But anyway, why are you judging Java by whether or not it is being used for desktop applications, or applets, that you have personally heard of? Have you heard of Delphi? It is has been a major development platform for desktop applications for over a decade, but I can't name a single desktop application I have used that is written in it. Does this mean it is not successful? Of course not. The shrink-wrapped desktop application market is totally insignificant compared to the number of applications developed for internal desktop use within companies, and even that is negligible compared with server side development. Java use is growing in the first area (with IDEs such as Eclipse and NetBeans often being used a Rich Client platforms ), and hugely dominant (but still growing) in the second. Don't believe me? Check out dice.com or monster.com for jobs, and see where the development effort is going on.
I humbly suggest you are wrong in you statement. If you are doing anything commercial on the web, you are most likely using Java applications all the time.
Even though they did little but stand in the corner with their vision?
What a distorted view of IT history!
Sun certainly did not 'do little but stand in the corner'. They were vigorous promoters of Open Systems in the 80s, working hard to encourage the use of UNIX and standard protocols against the attempts of competitors (like IBM) to keep things closed and entirely proprietary. The resulting wide use of UNIX and standards helped the original development of open source, and then Linux.
Because it's supposed to be 'write once, run anywhere.' At the moment, it's write once, and run on Solaris, Windows, Linux (x86 or IA64 only), Mac if you don't mind waiting a bit, and maybe IRIX if you are lucky.
.NET architecture with reasonable performance - why the hold-up with Java? Perhaps it is because there are already free (as in beer) versions of Java for the majority of developers, there is not such much demand.
You have missed out a large number of important platforms. For example IBM provides certified VMs to run on AMD 64-bit, Power 64-bit, z-Zeries 31-bit, zZeries 64-bit. HP provides other platforms, and I don't know what you are supposed to be waiting for on the Mac.
Only very recently have FreeBSD been allowed to distribute the port to their OS. If you look somewhere slightly more obscure, like OpenBSD, then you start to have problems.
The license means that OpenBSD are not allowed to distribute a ported version of Java. They can distribute diffs to the source code, but not the source or compiled binaries. Since Java requires Java to build, installing Java on OpenBSD requires following these steps:
1. Install the Linux JDK, which runs through the binary compatibility layer.
2. Download the source from Sun (manually).
3. Compile the JDK. This takes a long time and requires a lot of RAM.
Since the Linux JDK is only supported on i386, you can't use Java on any other architecture (such as PowerPC or SPARC).
I would much rather Sun used the trademark to protect Java. Make the JDK open source, but do not allow any patched versions to be called Java (so OpenBSD, for example, could include a binary of 'Columbian' that would run Java apps, but would not be called Java). Provide a mechanism for pushing patches upstream, so that if someone does port it to a new platform there is a good chance that the next release will actually support that platform.
To be honest, it's the second of these that is the clincher. The number of hoops the FreeBSD team had to jump to in order to be allowed to ship Java with their OS was insane.
This only applies if you want to use Sun's source code. There is nothing to stop an independent open source version of Java being developed and certified; indeed, the Apache Harmony project is doing exactly that.
Sorry, but this sort of post does seem to be typical of a highly demanding attitude from open source supporters:
First Sun releases Java.... at no cost. Then there are demands for it to be on more platforms. Sun provides it on more platforms. Now there is a demand for Sun to release their source code! What is the big deal? Projects like Mono seemed to have little difficulty re-implementing the basic
They have already open sourced J2EE (without the J2SE portion) in the form of Glassfish.
There has been an open source version of J2EE around for a long time: JBoss.
Java has several Free Software implementations, and the only real limitation they have is that they can't use the Java trademark. Kaffe, after a long absense, is back and is already mostly 1.4 compliant, GCJ takes an entirely different approach to implementing Java which is both innovative and supportive of what already exists.
And there is Harmony, an Apache project to create an open source Java platform which is certified as Java compliant, so can use the trademark.
Why? This seems to be a popular opinion on Slashdot, and I'm curious why people need it to be any more open than it is? I mean afaik, the only thing that isn't "open" about it is the spec. If you want to create your own implementation of a JVM you're allowed but it must conform to the spec. This is a very *good* thing IMO. It would really suck if MS had been able to complete its "embrace and extend" manuever on Java (which is what MS has done with the open web standards and browsers) and it would suck even more if there were 5 different JVM's out there and you had to tailor your code to run on each one. You would completely lose the WORA (or you'd have to do all sorts of gimmicky crap to figure out what jvm you were running on -thats a lot of fun with browsers and html, I think it would be even more annoying with code). So I ask again, not rhetorically, but honestly: why open source it? Am I missing something?
I absolutely agree that having conformant java specifications is a great thing, and is perhaps the single most important reason for Java's success. But I don't see why being open source should conflict with this - there should be no reason why an open source product should not have to pass the tests in order to be called 'Java'. In fact there is a current project, called 'Harmony', that intends to do exactly this. Open source need not permit 'embrace and extend'.
I can understand that there are potential problems with open sourcing Sun's implementation of Java - there are most likely huge amounts of code that involve patented techniques or are licensed from other sources.
I like open source (or at least having the source) because I have had to deal with problems in closed source products that won't be fixed by the vendor. I am not after re-selling the product, or re-distributing the source, but the possibility of patching something myself is pretty appealing.
I would lay claim that Perl and PHP power more online transactions than Java does. Look at the world's most popular forum system, PHPBB. Yes, that runs on PHP. And people don't run LAMJ servers, they run LAMP servers: the P stands for PHP or Perl or Python, all of them open languages.
And I would lay claim that they really don't. You can combine all the online forums you like but they don't come near the phenomenal combined volume of stock trading systems, banking systems, airline booking systems etc. We are talking of system which individually handle hundreds of millions or even billions of transactions each day. Consider the combined volume of transactions of all these systems....
And before you mention google - that uses a considerable amount of Java as well.
I didn't say that Java is now unpopular in all domains, that is false. But I think that it will degrade because it will not have the ability to adapt like other languages can. If Sun goes down, methinks it would be all over for Java.
Java is constantly adapting, with regular releases with new features (well, new to Java anyway!): Generics, improved concurrency and higher performance for the GUI in Java 5; scripting language support and web services support and far better client side integration in Java 6. How is this not adapting?
Apart from the wild idea that Sun is going down (their annual losses are trival compared to their net worth, and that worth is not largely dependent on share value), there are companies with far, far bigger investments in Java than Sun, like IBM. They are constantly producing new VMs for internal and external uses.
If Sun did 'go down', Java would certainly continue (in fact, IBM could well buy up the rights and open source it!). That is one of the reasons why I find it such an appealing language - it is not a one-vendor language.
And if you want to chide somebody for wanting to overcome the competition, fine. But don't forget that the origin of all open projects is the desire to build a better product, and it's only because we want to be better that we can achieve that. Wanting something is the root cause for it happening. That's not a guarantee, but it's as close as we can get.
I was not chiding anyone for wanting anything. What I was gently ridiculing is a Slashdot speciality - stating what someone might want (for whatever reasons) as if it has already happened.
I want better products - I would rather that more people adopted MacOS that Windows. I wish I could play more games on Linux. I would prefer Java to be open source. However, we have to face reality.
There is not the slightest sign (at least yet) that Java has stopped growing in terms of its adoption - it is still in the growth phase. This may change in a few years, but to say now that 'Java will remain on the path to obscurity' is ridiculous in many ways - it implies not only that Java is going to be obscure, but it is already on that path, which is obviously false.
Like it or not, Java works, and works very well for a very large number of developers. It would be nice if it were open source, I agree, but it seems to me that its current status has had little impact on its adoption, no matter now much open source supporters may wish otherwise.
No amount of marketing can change this: if Java is not sufficiently opened, it will remain on the path to obscurity.
I love comments like this! They indicate what a strange reality some slashdotters live in - it almosts make me believe in parallel universes.
I eagerly await other posts from this other dimension:
"Intel - will it ever take off?"
"Windows - how it lost out to Apple"
"Linux - the ultimate game platform".
Actually I guess the message here is that no matter how much you really, really want something to be be true (Java on the decline) this does not make it true.
As an old c/c++ unix programmer who is forced to work in java I see little real inovation in java. What started out as a highlevel embedded language has morphed into a text processing language for web development. Java is just differnet not better or visionary.
Apart from the fact that Java is used for far more than web development (it is becoming dominant in areas such as embedded and realtime code - an area that interests me), I agree that Java in itself is not that innovative. But what it does is combine the individual strengths of so many other languages in one package: Garbage collection, the use of a really high-performance VM for runtime portability, a robust security manager, built-in threading support at the core of the language, a cross-platform GUI shipped as standard, unicode as standard etc. And, it has strong multi-vendor support and is free (as in beer).
To dismiss Java as 'not innovative' is to miss the point.
Java, rebuilding the wheel one spoke at a time.
Yes, but it is a better wheel - one that goes faster and runs on more roads that most of the others.....