This week's book giveaway is in the OCAJP 8 forum. We're giving away four copies of OCA Java SE 8 Programmer I Study Guide and have Edward Finegan & Robert Liguori on-line! See this thread for details.
I think the test in very good in most of it's contents. However, about objectives 10.4 and 10.5, there is a very scary thing: Expect to see .Net questions!
The questions asked directly about how would .Net 3.0 webservice behave or not behave, including code written in C# and citing a webservice creation with Visual Studio. And worse, there was one question purely with .Net with nothing to do with java!!! So you should actually know .Net to answer correctly some questions in a java certification!
This type of thing, IMHO should not be in the test in any way!
I understand the need for interoperability with non-java webservices, however non-java != .net. There are webservices in lots of languages, including perl, ruby and python. However, the test failed to focus interoperability with non-java webservices focusing instead interoperability specifically with .Net.
So, I came here to protest against these objectives and alert who will take the exam (and i think most of the candidates would agree with me). I would appreciate if instead focusing operability with .Net the exam focused interoperability with generic non-java webservices. Because non-java != .Net. The exam could even state that some webservice is .Net, but it should not ask for any .Net knowledge. REWORK THESE 10.4 AND 10.5 OBJECTIVES!
By the way, there are a few questions about migration from JAX-RPC to JAX-WS, however the objectives does not state this! Obviously this is important and probably someone just forget to put this in the objectives, but candidates should be aware of this.
EDIT: BTW, there was a answer saying specifically about webservice creation with netbeans and ant. This clearly are not in the objectives and should not be.
[ November 09, 2008: Message edited by: Victor Williams Stafusa da Silva ] [ November 09, 2008: Message edited by: Bear Bibeault ]
Personally, I don't see the major issue with 10.4 and 10.5. So yes, they're asking us questions about .NET and how Java and .NET interoperate, but that's not uncommon in the business world. WCF is sufficiently wide-spread now that it's not unfair to expect some experience with it.
Admittedly, I may be biased - I have had experience with MS's technologies and I have not yet taken the beta (that's coming up this weekend) - but.. there was fair warning, from the way it was termed, that they might be asking direct questions on .NET web services.
Theodore Jonathan Casser
SCJP/SCSNI/SCBCD/SCWCD/SCDJWS/SCMAD/SCEA/MCTS/MCPD... and so many more letters than you can shake a stick at!
Victor, Thanks for sharing your experience. It will certainly help the rest of the community in taking the test. I do want to respond to some of the things you said. I also have a question for you at the end of my post. Disclaimer: I am not related to Sun Microsystems in any way.
Section 10 of the exam is about WSIT. Sun has explicitly made it clear that the goal of WSIT is to interoperate with .NET so I wouldn't expect any non .NET questions in this section. Sun participates in Microsoft plug-in fest (I think that's what it is called) to test the interoperability. Microsoft and Sun support the exact same versions of the specifications implemented in WSIT set of technologies and test each other's web services. For example, if WSIT Vx.y says it supports .NET v3.0 then both sides are supporting the exact same versions of all WS-* specifications. I do agree on your point that non-java != .NET but the intent of section 10 was never to test interoperability with non-java services, it was to test interoperability with .NET. Is it fair? May be not. Objective 10.4 does explicitly talk about creating WCF client but I wasn't expecting .NET code. But then who knows, may be that question won't be counted, especially if a lot of people get it wrong.
I would appreciate if instead focusing operability with .Net the exam focused interoperability with generic non-java webservices... REWORK THESE 10.4 AND 10.5 OBJECTIVES!
May be Sun should consider an architect level exam in web services and SOA and move these things in that exam. My personal experience tells me that if you are an architect, you will end up reading all kinds of code to design good services that interoperate between them. BTW, I do agree with your points about migration and Netbeans questions.
Finally, my question(s). Could you please help the rest of the community understand the exam little bit better? Of course you will need to follow this website's disclosure rules, but I was hoping if you could tell us whether the exam is more for architects or for developers? Did you see a lot of API questions or questions that require 'remembering' stuff or did you see a lot of analysis, design, evaluate a scenario type of questions? I am guessing not many beta takers have experience using web services under Java EE 5, what would you say to them about taking the test without much practical experience in JAX-WS? [ November 10, 2008: Message edited by: Chintan Rajyaguru ]
Originally posted by Victor Williams Stafusa da Silva: I understand the need for interoperability with non-java webservices, however non-java != .net. There are webservices in lots of languages, including perl, ruby and python. However, the test failed to focus interoperability with non-java webservices focusing instead interoperability specifically with .Net.
Which is justifiable. Have a look at the topic that I just posted in the web services forum. It seems that back in 2006 there was a +66% likelihood that a public web service that you were accessing was implemented with ASP.NET. I doubt that the percentage is that much smaller these days. What is worse is that these web services are not likely to be carefully crafted by developers with a decent competence in web service technologies; they are most likely created by someone who just knows where the right button is buried in Visual Studio (there are always exceptions of course). Which means that you will have to deal with the quirky WSDLs that those code generators (lets face it, JAX-WS has them too) crank out.
Theoretically there shouldn't be any interoperability problems - reality is always a bit different.
I just feel they have made SCDJWS overly complicated and boring. I have been swimming in the world of web services since the last 3 years and I feel the objectives are in no way reflect the practical difficulties faced.
I am in no way shocked the world is losing faith and interest in SCDJWS