I like look at the amount of posts in a forum as an indicator of the interest in the real world. In your opinion why webservices seem not so popular in JavaRanch, is it because it is a tough certification or because it's relatively new?
I like look at the amount of posts in a forum as an indicator of the interest in the real world. ...which in my opinion is not correct since forums by themselves are not necessarly authoritative entities, that reflect real world settings...
In your opinion why webservices seem not so popular in JavaRanch, is it because it is a tough certification or because it's relatively new? More the latter than the former. As I explain in another post, web services are still in an embryonic stage compared to the potential they can unleash. Just because this forum, the SCDJWS certification or even web services themselves are *not so* popular (which I refute for obvious reasons) doesn't mean that they have no future or no potential. Just think how much time it took the object-oriented paradigm to become as widely used as it is today. OO is only one of the numerous steps in the software engineering history. Aspect-oriented technologies have emerged to palliate to OO's weaknesses. And I'm pretty sure that many other paradigms will emerge in the years to come. The bottom line is that web services are definitely here to stay and that it is a (work)horse you can safely bet on [ January 18, 2005: Message edited by: Valentin Crettaz ]
Do you think that GRID is more popular and acceptable in real world project ? can it do the same things that web services do or they are completly differents ? it�s like comparing apples with oranges.
SCJP 1.4 / 5.0 - SCBCD 1.3 - SCWCD 1.4 - IBM 484
Joined: Aug 26, 2001
In my opinion, they are two different things that, if utilized efficiently, elegantly complement each other and can return real value.
Since one of the main goals of grid computing is to discover and utilize idle processes to carry out massive operations (mainly over HTTP) and since web services provide support for discovering remote services and invoke them over any protocol (for the moment mainly SOAP over HTTP), we can easily imagine how web services and grid computing could collaborate.
Let's say that a grid should be set up to compute the first 10^100 (or something big enough...) prime numbers. That would take an enormous time on a single machine. Let's say we have developed an algorithm as a web services that we have made available on 100 machines and that are exposed as WSDL. It would be possible to carry out the computation by utilizing the 100 remote algorithm processes and benefit from the added values of the web services infrastructure, which are retrieval of processes and simple communication between nodes. Actually, web services hide the complexity of remote invocations and let you concentrate on your business (compute prime numbers). I'm pretty sure there are grid computing infrastructure out there that also provide everything you need to carry out your tasks but web services would definitely be one solution among others.
Note that in some cases SOAP over HTTP would not be the protocol of choice since some computations may take a loooooong time. Exactly for this reason, web services would be a good choice since SOAP can be used over pretty much any synchronous (HTTP, FTP, ...) OR asynchronous protocols (SMTP, JMS, etc). [ January 18, 2005: Message edited by: Valentin Crettaz ]