A few days ago I saw the article Xamlon looks to beat Microsoft to the punch on C|Net which begins
On Monday, Colton's company Xamlon released its first product, a software development kit designed to speed development of user interface software for Web applications. Xamlon built the program from the published technical specifications of Microsoft's own user interface development software, which Microsoft itself doesn't plan to release until 2006.
I've been having difficulty processing this news over the past few days. Reading the Xamlon homepage gives more cause for pause. The site proclaims
XAML is a revolution in Windows application development. Xamlon is XAML today. Rapidly build Windows user interfaces with HTML-like markup Easily draw user interfaces and convert directly to XAML Deploy to the Windows desktop and to Internet Explorer with absolutely no changes to your application. Run XAML applications on versions of Windows from ’98 to Longhorn, and via the Web with Internet Explorer Write applications that port easily to Avalon
XAML is a revolution in Windows application development. Xamlon is XAML today.
What I find interesting about this are the claims that involve unreleased products that aren't expected to beta until next year and ship the year afterwards. I can understand that it is cool to claim to have scooped Microsoft but considering that the XAML and Longhorn are still being worked on it seems strange to claim to have built a product that is compatible with unreleased Microsoft products.
While writing this blog entry I decided to take a quick glance at the various Avalon folks' blogs and I stumbled on a post entitled Attribute grammar for xaml attributes from Rob Relyea, a program manager on the Avalon team. Rob writes
As part of this change, the flexibility that people have with compact syntax will be reduced. Today, they can use *Bind(), *Button(), *AnyClass() in any attribute. We'd like to restict this to a small set of classes that are explicitly in need of being set in an attribute. I'm not going into great detail in the description of our fix because I'd prefer to be the first company to ship our design.
As part of this change, the flexibility that people have with compact syntax will be reduced. Today, they can use *Bind(), *Button(), *AnyClass() in any attribute. We'd like to restict this to a small set of classes that are explicitly in need of being set in an attribute.
I'm not going into great detail in the description of our fix because I'd prefer to be the first company to ship our design.
Considering that XAML isn't even in beta yet, one can expect a lot more changes to the language before it ships in 2006. Yet Xamlon claims to be compatible with XAML. I have no idea how the Xamlon team plans to make good on their promise to be compatible with XAML and Longhorn before they've even shipped but I'd love to see what developers out there think about this topic.
I totally empathize with the Avalon team right now. I'm in the process of drafting a blog post about the changes to System.Xml of the .NET Framework that have occured between Whidbey beta 1 and Whidbey beta 2. Even though we don't have companies building products based on interim versions of System.Xml we do have book authors who'll have to rewrite [or maybe even eliminate] chapters about our stuff based on changes then off course there's the Mono folks implementing System.Xml who seem to be tracking our Whidbey betas.