Reacting to Suns recent claim that Java is the only language suitable for for the growth of web services, an article on vnunet cites a number of analysts who think that this attitude will hurt their web services strategy.
While Java definitly has strong advantages as a platform to implement web services, I feel that most of these 'analysts' objections are irrelevant given that Sun is truly using Java to support open and interperable standards such as UDDI, XML and SOAP.
I think that this article does a poor job of discussing Suns strategy. If you want to know what Sun really recommends, read the Sun-sponored whitepaper in TheServerSide.com's resources section entitled "XML-based Web Services with the Java 2 Platform, Enterprise Edition (J2EE)".
Read Sun blinded by devotion to Java tactics
Jelousy is never attractive.
Isn't the message from these analysts that the whole point of "web services" technologies (such as SOAP and UDDI) is that takes the implementation architecture off of the critical path. So you can truely use whichever implementation makes the most sense (cost/performance) or leverage of existing systems with simple adaptor layers.
So I _could_ use:
1. SOAP Tookit onto COM+ components
2. Java classes to manage SOAP/web service layer.
3. Microsoft.Net (_still_ beta)
5. (any other implementation on any platform)
None of this centres on Java, and whether this is the most appropriate implementation technology is open to debate (despite the _oh so_ unbiased Sun-sponsored document written by TSS ;-).
I wouldn't take this too easy. I still remember the old chaotic days when a miriad of solutions arose to make the web interactive (PHP, Pearl, CGIs,...). Java was a great step, giving us a standard for making enterprise and web applications, with the promise of abstracting us from the underlying platform. Java became a well known language for developers, and a rather sensible choice for enterprises to adopt.
Now we have moved a step further. Web services propose, as Lee and those analyst point, independence from the underlying platform, as once Java promised. So I think the war now will be focused on products giving the best time to market characteristics in building web services.
I'm a Java fan from it's very beginning, but I don't think we can impose Java as the language for building web services based in its APIs, standardization or anything like that. If M$ then comes with its .NET strategy, and gives developers a "point & click & rest while we do the job" GUI like it uses to do, I think Java will have a hard time, even having the most portable architecture, the most complete APIs, and the most whatever you can imagine.
Let's admit it, enterprises want things done, and they want it fast. I think the one offering the tools for building web services faster will be the one in the leading position. I really hope that being Java, cause if M$ wins on this, I think I leave this and become a gardener ;)
Personally, I fancy becoming a lumberjack. ;-)
Don't chain saws come with embedded Java VMs now a days?
Personally, I believe that webservices are a pre-mature technology. I don't see it as as taking off or being very big for a very long time. the foundation just isn't there.
In the short term I see nothing but hand-waving by technology experts as most companies aren't even spending money to hire needed IT people as they fear this market recession. After that were do we go?
When the idea of web services was conceived there wasn't a dot-com shakeout going on. The whole purpose of the technology is to bring together business into a common directory for the publishing of services online. They came up with this idea when selling toads online was a business plan.
And at what market is this technology advancement pointed at? Are we looking to get fortune 500 companies to throw away their current communication channel for this. Think about it. These companies have been working off of established contracts and communication lines for years. What is an online directory service going to provide them with. It's not for the dot-com companies that for sure. Is it poised to service the needs of the needy consumer? Maybe its pointed at making sys-con and their new web services magazine some money?
I heard many a wild tale of how web services will benefit I just don't see it as being a solid venture anytime soon. I think that once wireless is more established, and the internet generation makes it out of high school and college that such services might be useful.
I completely agree with Matthew.
To add a few points:
1. How many often do you look for new partners rather than working with established partners?
2. Would you changes you existing propritary mecahnism to do business with your partners IMMEDIATELY just because a standard has come out?
3. How mature are todays web services? Probably most web services implemented on J2EE today are based on Apache SOAP. Here are few of fixes in latest release (May 30):
a. Removed many printlns from the server-side code, replacing them with thrown Exceptions.
b. Removed non-threadsafe private variables from Servlets.
I guess its going to be long time before web services can be called matured enough.
I agree till broadband communications becomes commonplace for both wireline and wireless communications web services are DOA. I don't think these analysts have ever even worked in a service environment to even understand the kind of infrastructural and man power needs of supporting it.
On the wider question of relevence of Web Services to the current commercial environment, no-one seems to be discussing the commercial realitites of electronic trading. It's pretty clear that interoperability is the key to allowing seamless integration of services. The growth of e-procurement, supply chain integration (manufacturing, banking & investment, etc)... All these companies have some <optimism>strategy</optimism> around Java/J2EE (I work across these verticals, and if they don't have it, they're trying hard to get there now).
How do you suppose all the little banks can seamlessly offer their investment products to all the other banks? How do you manage complex supply chain integration across different businesses?
Well, some technologies that standardise the communication, foster interoperability are essential. Just take a look at the UK e-GIF2 initiative to get the importance of this emerging standard... (government interoperability framework)
In conclusion, all technologies are going to be immature after just a year or so, but we need *some* global interoperability solution that all manufacturers of tools buy in to.... and this would appear to be it..
Once again, analysts are showing that they have no idea what they're talking about, and are unfortunately going to cause a lot of harm with these go-nowhere web service strategies that are bound to pop up everywhere.
Please, show me where there is substance in these quotes...
"It isn't about vendors creating the best platform, but how users will interact with others using different systems"
It's not about creating the best platform. Think about that for a second... Okay, so it's about creating a way of interacting with ... what.. mediocre platforms?
This one's even better...
"Sun doesn't seem to realise that it's not about the platform, but the ability to reduce the time needed for business processes,"
Okay, how does one build business processes? With a point and drool GUI tool? Or with a platform consisting of business components, frameworks, and a sound programming language with a large literature of patterns, idioms, that is battle-tested in the real world?
There seems to be this aversion to the "P" word among analysts -- i.e. the fact that in the end, "P"rogrammers have to make it all work. That requires a quality platform. I'm not implying J2EE is the be-all-end-all, but it certainly is one of the best available today.
"Sun can't see the wood for the trees. It needs to realise there's a world beyond Java, or Sun One will be too limited to be a major success."
Limited by what? Java? Where's the study showing that companies are moving away from it? Why should Sun branch out and compete as a generic web services framework when there already are Oracle and IBM doing that? Shouldn't sun focus on it's strengths (the momentum and growing quality of J2EE), and not face Microsoft head on again? I rather like the idea that .NET and J2EE are somewhat oblique competitors at this point. Another "platform war" might sell more newspapers and analyst reports, but I think it will generate more heat than light, especially since .NET needs at least another 18 months before a port to non-Windows platform could even conceivably be mature enough to use.
Well-said. I think it's just industry jelousy over how excellent Java really is. There's nothing a company can't do with Java.