Mono roadmap?
I was a bit pessimistic about the Novell aquisition of Ximian and it's impact on mono. Novell has a habit of burying good technology under inscrutable marketing, ( i.e. did you know they sell a Java App Server? It's got a really flashy name: 'Extend'), but with the Suse aquisition things are falling into place. Mono's importance and longevity at Novell now looks assured as a reasonable differentiator in the commodity Linux-distribution market. A play identical to Red Hat's involvement in the ObjectWeb consortium - a piece of differentiating middleware and keep up with Sun's middleware-moniker-de-jeur (Orion?) and MS's middleware-moniker-de-jeur (.NET/COM+?/Indigo?).
But it was a bit disturbing to read Miguel's roadmap, written after attending the Microsoft Professional Developers Conference (PDC).
http://www.go-mono.com/mono-roadmap.html
Here he describes how he's going to translate all (or most of) the .NET 2.0+Longhorn innovations MS has in the works into marching orders for the open source mono project. Doesn't that seem deeply wrong? Open source is supposed to imply some degree of vendor independence, but Mono is committing to just reimplementing large chunks of stuff Microsoft will probably just open source anyway. And the likelihood of implementing enough of the stack to offer true compatibility and portability seems to fade as .NET pushes further down into the operation system. Copying that type of integration on other platforms doesn't seem either wise or likely.
Miguel gives the rationale (eloquently, as usual) here:
http://primates.ximian.com/~miguel//texts/linux-developers.html
But I don't buy it. Vendors shipping C applications that are hassled by ABI problems will retrain all their developers and port to C# to gain stability? Hurrumph. Say for a moment that they did. They'll want to use all the whizzy new Longhorn features of .NET, and how or when will they see them on Linux? For example, if something pretty cool-sounding like XAML is released in 2005, then mono will probably have it implemented and stable by 2007 based on their progress with ASP.NET etc. Is that a platform release schedule you, the ISV, want to depend on? I think by that time virtualization on the OS level (Xen,VMWare,UML) and Moore's law will combine to be just as competitive with any other technology for easily distributing software and eliminating versioning issues, and mono will lose it's raison-d-etre.
I like to see big, ambitious projects like mono succeed - it's just makes me happy. But I think the scope creep of going from "Good platform for Linux developers" to "Mirror image of MS deployment environment" has mono flirting with death-march status.
9:43:55 AM 
|
