Sajid Raza

Ranch Hand
+ Follow
since Jul 13, 2001
Merit badge: grant badges
For More
Cows and Likes
Total received
In last 30 days
Total given
Total received
Received in last 30 days
Total given
Given in last 30 days
Forums and Threads
Scavenger Hunt
expand Ranch Hand Scavenger Hunt
expand Greenhorn Scavenger Hunt

Recent posts by Sajid Raza

Well, all I can add is that unless you have experience certification do not mean anything. I've got both the programmer and developer certification with respectable scores 90% (1999) 96% (2001) and the employers showing interest in me only do so because of my VB/COM experience.
Plus, most Java jobs really are J2EE jobs, therefore, SCJP and SCJD aren't really helpful with those. Maybe the SCWD would help, but I wouldn't count on it.
22 years ago
I'm sorry to ask this again, but previous answers to this question have been ambiguous and Sun's site has no clear answer either.
What specifications (Servlet and JSP) are applicable to the test?
The SCJD reflects the process of a real project: following specification, deriving detailed requirements, researching implementation technologies, implementing the application, justifying decision, and creating documentation.
Frankly, the argument contending that the SCJD has less value because it is not an exam taken in a "secure" class-room is incredulous. First of all, a person who really wanted to could somehow outsmart the system on any test. Secondly, any person who did not do his project is quite likely to fail the follow up exam. Last of all, the exam requires research and consultation. If you ask someone or read a book, it is no different. I don't really see the big deal behind research and consulation involved in the exam.
The harder certifications such as the CCIE require some sort of practical work. Why would they include a project component in the SCEA certification?
An equivalent of dynamic class loading in the C++ world would be linking/loading DLLs at runtime. But, that's more an OS feature rather than a C++ specific feature.
22 years ago
This might help peoplee trying to verify their lock/unlock implementations: Test the functionality by setting break-points in the code at appropriate location. What I did was to set a break point after a call to lock, starting the clien in the debugger, and then starting other clients that try to lock the same record to see if the call blocked properly.
I think simulating many threads/clients is still tricky as compared to testing one or two clients with the right break-points set to see how they operate step by step. There's not much too it, but it makes life easy when you want to see how your locking code operates.
I think it depends on how you'r starting the rmiregistry. In order to avoid specifying a server codebase I used to start the rmiregistry in the same current directory as the root directory of my classpath.
Were you starting rmiregsitry in c:\scjd ?
Compile/execute/test your client and server using a non-beta release of the JDK.
I'm guessing that your server class is actually located in the package suncertify.db. You might try
java -Djava.rmi.server.codebase=file:///c:/scjd/
instead of
java -Djava.rmi.server.codebase=file:///c:\scjd/suncertify\db/
Basically, I think the codbase you're giving for your rmi server doesn't include the suncertify.db package because the corresponding codebase would have been C:\scjd otherwise.
In some high-performance oriented libraries/systems inline-assembly is used to optimize function calls. Is there ever any equivalent in a Java method? I understand its possible to call a native method. However, would it be possible to insert inline byte-code in the middle of a methods statements? Would such a practice be of any use?
22 years ago
I created a few resource bundles for externalized strings. The resource bundles were placed in the nested (object-code) jar archives (server.jar, client.jar). I also included the resource bundles in the source code directory.
I did not use any policy files. I didn't even justify that I didn't use any policy files, I was too sick of the whole thing at the end. However, I suggest that you do include policy files for the sake of completeness. He will have to extract the policy file.
Regardless, for the examiner to test your submission he is going to need access to the README.txt file which he will have after he extracts the submission jar. In the readme just document the location of the policy file and how to use it. I think the submission jar file serves more to zip all of your submission into a package rather than serve as an executable.
I used VAJ generated code for the submission. No problems. I also tested it using JRE 1.3.3 prior to submission.
They're going to have to extract the db file. I don't believe you can access a resource within a jar file for writes.
You need to have the README.txt inside the submission jar file.
Are there any resources on how the JVM is implmented on various platforms (specifically Win32). I've been reading Inside the Java2 Virtual Machine, and I want to delve further into the subject matter. Any pointer on JVM implmentation and general VM theory would be appreciated.
22 years ago
Most people who went the client id route changed the method signatures. All you need to do is justify your decision to change the signature in your documentation.