"Convenience Over Correctness" - The Case Against RPC

[michi]

And it is still think it is the only alternative to other "cool" frameworks
Don't forget that, in the end, you wrote a wonderful book on that topic. With Steve Vinoski

Right, I haven't forgotten that. CORBA was the best thing going at the time. But CORBA got worse over time instead of better, and it took years of experience and learning to realize how all of the CORBA problems relentlessly added up to something that was very hard to use. (The run-away standards process didn't help either. I wrote about this at length in The Rise And Fall Of CORBA.)

As usual, the problem is not the gun but the man behind it.
I bet that it is possible to find a lot of people out there that can make disasters even with ICE.

That is no doubt correct. Ice is a powerful tool and, as with any powerful tool, there are ways to use it incorrectly. Ice makes things as easy and efficient as possible, but that doesn't mean that it guarantees a successful distributed system. That still takes experience, insight, good design, and good implementation.

As the old saying goes, "a good programmer can write FORTRAN in any language..."

Cheers,

Michi.

[ganzuoni]

Right, I haven't forgotten that. CORBA was the best thing going at the time. But CORBA got worse over time instead of better, and it took years of experience and learning to realize how all of the CORBA problems relentlessly added up to something that was very hard to use. (The run-away standards process didn't help either. I wrote about this at length in The Rise And Fall Of CORBA.)

Got it !!
I still remember when I was a happy Orbacus user and I was asking for a mean to let the application "register" and "deregister" objects in the ORB (it was in the 2.0 old days) and Matthew suggested the reading of a "new" proposal from Douglas Schmidt.....
Apart from the well-described complications, POA concept is a corner stone for what a DOC framework should offer on the server-side.
Client-side portable layer is another bright example of the good things CORBA has put on the table.
These concepts are totally absent in SOAP-based specs and the marvelous J2EE.
And no one seems to care about....

Guido.

[kwaclaw]

Just because the web is good for distributed computing with a human at one end of the connection does not mean that it also is good for distributed computing with computers at both ends of the connection. As far as I can see, no-one has ever really examined why the web is as successful as it is, and how that might affect distributed computing between computers, instead of between humans and computers...

I think that is a crucial observation.
The web/REST architecture might work if the computers on both end had some "artificial intelligence" that takes all kinds of context into account (incl. situation, cultural, ...) in order to interpret the incoming message unambiguously.

The error in the REST thinking is that the bar for that AI is so low that it is practical to achieve today. In addition, the processing overhead would mean that such messages should be very high level and of coarse granularity (just like human communication), which they not always are.

I do actually think that at some point we may have such style of computer to computer communications, but they will in part be used to dynamically establish and delegate to low level, strongly typed and efficient communications that for instance ICE provides.

Karl