Silvio Esser

I'm looking for an economic Java web hosting site. I do not want a shared web server. Any recommendations?

What if I have two requests that access the singleton bean at the same time? Does one request have to wait
until the other to finish. How does Spring container find the singleton bean instance for my requests? Using static method or
variable?

I know the session is cached by default. What happens when same query is issued from multiple users at the same time?

Using the Singleton/Double Checking Locked pattern?

Originally posted by Ulf Dittmer:
That's a good point - using a class file hierarchy instead of a jar file would circumvent this particular problem.

How to use class file hierarchy in Applet?

Is there a way to instruct the Classloader of the JVM running in a Browser to load the classes from the internet, not local jar file?

Encryption can make things very complex and slow.

I'm hoping that the Classloader of the JVM that runs the Applet can load the
classes from the internet (not from the jar). If it could be done, the Java classes would be just in the RAM. When the browser is closed, the RAM that the Applet classes used will be available for other things. So we can say no residual.

Originally posted by Ulf Dittmer:
...

So, from a security standpoint, the cache doesn't make anything worse than it already is.

It does.

Since it is a secure application, any user who can retrieve the jar has to login first because the URL of the jar is protected. A person who is not authenticated will not be able to establish a secure session, and therefore the server will deny his/her access to any application URLs except the login screen.

Now the jar is cached in HD. Technically, anyone who can get hold of the HD can get the jar.

Originally posted by Ulf Dittmer:
Applet jar files are cached, and remain somewhere on the hard disk. The location and size of the jar cache can be adjusted with the Java Plugin control panel.

So someone can get the jar file and de-compile it. It is an security issue.
Is there a way to delete it after browser is closed or just keep all the classes in RAM?

http://news.bbc.co.uk/1/hi/business/7437908.stm

I heard this company before. There are many this type of companies and people in popular US job websites. Be careful.

Not sure what happens to the resources RBG placed in the world.
[ June 07, 2008: Message edited by: Silvio Esser ]

Once you closed the browser that ran an Applet, is the jar file containing the Applet left on the hard disk or the RAM of computer?

Which is better in terms of usability (look & feel), learning and development for Java developers, maintaining, integration with popular Java frameworks, future?

It seems to me that Flex is more or less like a Applet which give much richer behaviors than HTML.

What's the problem with applet plus swing?

Originally posted by Jim Yingst:


Of the Java techniques listed, cloning and deserializing are the only two which do not ultimately invoke any constructor. In that sense, we may say that there are just three ways to create an object in Java:

The default implementation of Object.clone() in Java uses the "new" operator
to create an new instance and shallow-copy the contents from the old one.
So the constructor is called.

(1) The messages are not distributed by a load balancer, but by an internal queue-to-queue mechanism through the queue managers themselves. When CLIENT_A talks to its queue manager an exit will choose which of the actual "queue instances" to put the message on.

(2) CLIENT_B just connects up to the clustered queue the same way it would connect to a local queue. The redbook I referred you to has some detailed examples of how this works.

Kyle, Thanks for your reply.

1. I guess the "queue-to-queue manager" is a software module that is installed on both nodes. One of the "queue-to-queue manager" is actively serving CLIENT_A and distributing
the messages to MQ service providers on the two nodes while another is passively standby. When the active "Q2Q manager" is down, the passive "Q2Q manager" on the other node will find it and become active. There must be some IP porting happening on the background. Am I right?

2. How does the MQ cluster ensures CLIENT_B that the messages out of the cluster are in correct order (FIFO)?

I want to create a cluster of MQs to support high availability. In the cluster, there are two MQ service providers installed on two separated nodes. The logical Q name for this cluster is �my_order_q�. My questions are:

1.When CLIENT_A puts messages on �my_order_q�, how are the messages distributed between the two MQ service providers? Using a hardware load balancer in front of the two nodes?

2.How does CLIENT_B retrieve the messages from �my_order_q�? How does the cluster handle the messages from two separated MQ service providers?