Application development Toolkit

Story: .Net developers can write for Linux using Mono

Posted by: Gary Edwards (Friday 7 May 2004, 11:48 PM)

A Portable Application Environment:

The effort to establish a portable application environment took a big step forward with this release of the MONO beta. Once again the peasants with pitchforks are readying to storm the Bastille, to demand the Emperor release the prisoners. All prisoners.

The vast legacy of Win32 application users chained to the Windows platform. And the next generation of .NET framework application users being targeted for the great Longhorn roundup.
Microsoft pulled out all the stops to prevent Netscape and Java from establishing a cross platform environment developers could reliably write to. Chairman Bill risked the empire to make sure that these efforts to poach on the Windows platform, cannabolizing and pillaging the Win32 API franchise, would at best result in the uncertainties of developers having to hit a moving target. In this regard he succeeded brilliantly. The peasant rebellion was crushed. The franchise has to be protected at all costs. And it was.

What's interesting about the MONO project is that it doesn't try to carve out a beachhead in hostile territory, porting a new application environment ala Netscape and Java. Instead, they take an approach more similar to the WiNE Project.

Where WiNE simulates the Win32 API, MONO simulates the next generation Win32 franchise replacement, the .NET framework. A key difference being that while no one “writes” to WiNE, MONO was created as portable environment that could accommodate the porting of existing .NET applications, and, provide a reliable “target” developers could write to.
I think that this “dual” capability of being both a porting methodology (a simulated environment) and a full fledged application environment, truly separates MONO ambitions from Java expectations.

For any developer working within the muck of non inter operable heterogeneous systems, they need all the open source – open standards based tools and environment simulations they can find. They need access to open source frameworks, services, and core applications that can anchor the flow of inter op threads and channels.

Java is a wonderfully robust and well designed network application environment. So much so that it's platform specific clone, .NET, does next to nothing to push the envelope beyond standard Java functionality, expectations, and vision. What .NET does do, that Java cannot, is leverage the secret interfaces, class libraries, network services, communications and messaging protocols that maintain the monopolist's iron grip on the application marketplace. The hard truth is that Java was unable to crack Microsoft's hold on the Windows desktop.

The MONO simulation of those secret interfaces, protocols, and services is both open and portable. And because it's open source, given that enough developers write to the MONO environment, the simulation could very well become an open standards basis for network applications. And wherever open source technologies have a chance of becoming the basis of critical mass open standards, computational consumers and developers win in a big way. As Netscape proved, the mere possibility of cutting into the monopolists space, and releasing the grip, is enough to send the markets of speculation into a frenzy.
Many open source advocates say the MONO is a trap. That sooner or later Microsoft will kill every application written to MONO, or ported through MONO. Simply by making arbitrary changes to secret system calls, interfaces and services, Microsoft can choke of the competition and seize juicy opportunities others worked so hard to grow and ripen. After all, that's exactly how they manged to wipe out the entire Win32 API competitive landscape.

No one expects Chairman Bill to suddenly stop doing what has worked so well, legal of not. So, regarding MONO as a porting model for .NET applications, the skeptics have it right.
But that's not the whole story. While these tactics can and wi