Is Pjax ( Push technology for Ajax-like paradigm ) valuable?

I received a feedback from a programmer a week ago, and this fellow seemed to get the point of Pjax wrong. He wrote below:

"As long as it is easy to call a classical AJAX request object and other javascript functions from within a Flash object, I don't see why to pay for that. I guess I can do that in about half an hour, or so... for a simple news marquee, let's say."

Well, let me describe the problems we will face when developing and deploying RIAs equipped with push technology, and we will find out the answer of the problem: "Is Pjax valuable?"

According to the description in Wikipedia, COMET has problems in scalability and guaranteed delivery:

"There are some potential concerns regarding how well a web server implementing Comet can scale. Managing this large number of connections which are kept alive until events have occurred poses a considerable load on the system. The problem worsens if different connections wait for different events. Unfortunally, not all http servers provide intrinsic support for COMET, and that’s why servers and other supporting applications are being developed which have better support for such conditions.

The problem arises with guaranteed delivery of the message to the client. The client may be busy doing something else, and if it cannot temporarily suspend that job to receive the results, it will be lost. To get around this, some Message Oriented Middleware is required, that will allow the client to retrieve the received message at its own pace. Also, browsers, which host the client application, need to be designed to allow the client to connect to as many servers as it requires, and stay connected as long as required, to receive messages from them."

Besides, it seems to need higher programming skill to develop a COMET application. It may be a good news to experienced developers, but definitely not to project managers, CIOs and proprietors.

So, here comes the value of Pjax. Pjax combines the benefit of high installation base of Flash Player plug-in on browsers, and adopting a powerful push server which can solve the problems of scalability and guaranteed delivery. And most of all, the coding skeleton of Pjax is quite clear and simple; it doesn’t need a skillful developer to develop an application with Pjax. I think it’s an angel’s voice to proprietors.

Wanna to get further proofs? Please visit http://www.icetechnology.com/develop...wto_pjax.shtml . This is a simple programming tutorial proposed by ICE Technology Corp.

Of course I don’t think the skeleton is perfect. So, what are your opinions? Please share with us.