Replies

First of all rit, ImaGraftedBranch can enlighten you about some of the issues you raise, so I'm pinging him in hopes that he jumps in to comment upon what you have posted. Let's get on to your comments now: "True only if by "between development platforms" you are referring to between Microsoft development platforms. . . ."

No, only one platform has to have .NET installed. The other platform can function without anything from Microsoft installed whatsoever.

". . . Otherwise, .NET has nothing to do with it. The mere fact that an application on a solaris OS may call a .NET app on a Microsoft box does not equate to .NET having removed any barriers. . ."

No, .NET has everything to do with it and it has removed the barriers of compiled programming language substituing a "language neutral" runtime compilation in its place. Let's say you have a server-side only Java application installed on a Solaris operating system on one machine and some RPG business logic installed on another machine running the OS/400 operating system and you want to call both from a third machine, which has .NET installed, to bring their information together for use in an application that integrates the two. Without access to a special Application Server on either the Solaris or OS/400 machines, such as an installation of WebSphere on the OS/400 example just given which would permit RPG code to be called by a Java application, you could only call these programs from another application written in the same language. The .NET Common Language Runtime changes all of that. It can call applications written in Java -- or any other object oriented programming language -- from any application written in any of the .NET programming languages (VB.NET, C#.NET, JScript.NET, COBOL.NET, etc.) and it can call RPG, provided that AVR.NET is installed (a plug in from ASNA), just as easily. The installation of additional application servers or interface applications on the Solaris and OS/400 machines are not necessary for a .NET app to communicate with either of the two applications described since .NET is "language neutral." And I haven't even begun to get into XML and XML Web Services where .NET's capabilities outshine all others.

". . . Interoperability implies heterogeneous (ie., windows and non-windows operating systems). . . ."

If by "heterogeneous" you mean the ability to install .NET on machines with distinct operating systems, that is not "interoperability" which implies the ability of software installed on distinct platforms to "operate" in communication and exchange with each other across those platforms. That is what .NET provides.

". . . The only barrier .NET may have removed (albeit not entirely), is the need to interact with certain MS apps that previously were difficult to communicate with. . . ."

Nope. .NET has removed the barriers of programming languages and has opened up XML and XML Web Service programming beyond anything running on any other development platform. If you want to contest this, give me a counter-example and I'll take it on.

". . . To argue Microsoft .NET does more to remove the barriers is akin to saying Microsoft has a lion's share of the market place and they are slowly being forced to license protocols and APIs that previously were proprietary. . . ."

You open source guys who are still concerned about operating system APIs don't realize that the world has changed and you're living in the past. Microsoft is revealing its APIs -- they don't have to be licensed -- because they don't matter any more now that .NET has made application code "language neutral" and the open and non-proprietary standards of XML and XML Web Services are becoming the industry norms for information exchange. Manipulating an operating system API was vital in the days when Microsoft's proprietary ActiveX technology was the centerpiece of its enabling of application code, but that day has passed.

I will admit that the licensing of protocols is a different matter and that there is reason to criticize Microsoft in some instances related to protocol licensing.

I disagree with the assertion that "Microsoft .NET does more to remove the barriers that have existed between development platforms, and which have impeded information exchange between them for years, than any other technology now available."

The reason is the statement is not qualified and my interpretation of your arguements is that you are implying .NET has done more for interoperability between Windows based, and non-Windows based systems, and that is just not true.

It would be fair to say that .NET has done more for interoperability between different applications of different languages all running on Windows based systems. But, to say it has done more for interoperability than any other technology that exist today, is, well, unqualified.

As noted before, I do not disagree with Microsoft charging for access to protocols or APIs, nor do I disagree with Microsoft providing a license that is inconsistent with the GPL. I do not favor the GPL. Specifically, there are issues in using GPL'd software for public accessible services that eventually a court will need to resolve. But that is a whole other area and I would be happy to discuss through freep-mail.

Counter example:
Interoperability between applications on a single system. The Unix shell. From a simple shell script, I can call any other program written in any other language, and I can pipeline output from one to the other.
Interoperability between applications on multiple platforms. XML, WSDL, SOAP, and UDDI.

Note that neither example mentions Microsoft. The later works because it is a service oriented architecture. In fact, hundreds (of thousands ) of non-Windows based systems can interoperate this way independent of .NET.

So what does .NET offer? Interoperability between applications callable from the .NET framework (the framework only runs on Windows). Again, that is not a bad thing, it is just a Microsoft only thing, and that is why I disagreed with the way you phrased your statements.