Working with Office Automation from Java or .NET is often a frustrating endeavor. In addition to the problems that Java and C# have with the late binding and optional parameters so heavily used by the COM object model, there is the insistent need to release COM objects. Miss a single one and the associated Office process won't shut down until some indeterminate point in the future. And most people recommend not even trying to use it from a service or website.
With the new OpenXML format, there is the promise of an clean and efficient way to manipulate Office documents via XML. But with a 6000+ page spec, finding the exact nodes one needs to manipulate is a non-trivial task.
This is where SDKs and libraries come in. Java open source developers were the first to announce development plans with OpenXML4J.
This is the real started point now and we really hope to move on an alpha version early in July, and then a beta in August. We're still looking for skilled contributors and some help to give us - the community - a hand to spread this news around the web!
Their website includes a well defined roadmap and overall feels more like a well funded commercial project than an open source offering in its initial stages. If the resulting code is as good as the web site, Java developers are in for a real treat.
On the Microsoft side, we find an preview release of the Microsoft SDK for Open XML Formats. Microsoft has a head start over the Java project in that .NET framework already implements some of the very low level code needed to support the format.
The class library for the SDK is already online. Though much of it is still subject to change, the documentation is already approaching what one would expect from a Microsoft offering.
