I don't write too many articles when it comes to announcements like this, but the trend around F# is particularly interesting and I am really curious about the outcome. A while ago (actually, at the end of October), it was announced by Microsoft that it dropped internal support for IronPython and IronRuby, putting these projects in the hands of the community under the Apache 2.0 open source license. Will it have a bad effect on the Python and Ruby layers on top of the .NET Framework? It's too early to say anything, but from general trends I think that these languages won't pick up much momentum and will eventually fade away, suppressed by C# as the main programming language on the .NET platform.
Today, I realized that F# is going down an almost similar road. From what I see, Microsoft wants developers to not only use their tools, but also become familiar with what's under the hood, and there are several advantages to that which I will talk about later on here. Mind you, there is a distinct difference between the F# news and the situation with Iron-languages - Microsoft isn't dropping support for F#, but rather they are releasing the source code for educational purposes and tool development, as per Don Syme.
The compiler source code is now bundled with the F# PowerPack, available on CodePlex. The source code itself was out there in public for quite a while as a part of the CTP release - what changed is the delivery model. The license is the same for IronPython and IronRuby - Apache 2.0. I can see the reasoning behind the decision - Microsoft definitely wants to push F# into public and get its name out there - it provides a reasonably new approach on the .NET platform compared to C# and VB.NET, which have been dominating for quite a while. Will F# be the third wheel? Not at all - it was gaining popularity and positive reviews since the very beginning - not a lot of people expected to see functional programming and .NET linked together.
Eventually, this will enable Microsoft to take the language even further, to the point when the compiler and the set of core libraries will be ported - yep, I am talking about cross-platform usage. Will it be ported by independent developers? Not so much - the current situation as well as past experiences tell us that independent developers won't be the leading force behind various forks and improvements. I think the developers from Novell (the guys who are leading the Mono project) will be the ones who will mainly consume the new releases of the source drop. They already have a set of cross-platform tools for F# (also available on CodePlex), speaking of which - today Miguel de Icaza announced the F# plugin for MonoDevelop that will support IntelliSense and type checking - the language is definitely a solid candidate for living along with C# and VB.NET and the guys at Mono see it. Besides, we could eventually see a F#-Contrib project.
Looking deeper into the fact that the F# compiler is open gives us the opportunity to see that it is possible to fix bugs much faster (if any are found) now that people know there is source code out there. And once again, this brings me back to Mono as I see them being the primary audience - since they do so much cross-platform work on existing .NET tools, they will also be the ones detecting some of the flaws and problems in the existing toolset.
As F# gets a stronger position in the Microsoft language stack, I would expect that it will become a full-fledged part of the framework - it's almost there, since you can develop Windows Phone 7 applications, XNA games and ASP.NET web applications. All that is needed now is full Microsoft support when it comes to tool integration and maintenance - including the toolset in Visual Studio by default for a variety of projects, because right now it is all scattered across the web.