I've been following a heated discussions on XML parsing methods that exist in .NET right now: DOM and XmlReader, whereas the first one is the slowest and the second is fast but a more complicated to code against. A must read on this topic is the DonXML's post Waking Up From a DOM Induced Coma who underlines the existing problems with XML parsing and mentions a few other alternatives. All of the problems about XML parsing are relevant to .NetCF as well, therefore they are multiplied due to processing power and memory constraints of the WinCE devices. Don't forget a complete lack of XPath implementation in the current version of CF.
As lucky as we are, Dare Obasanjo and Co., a PM from Microsoft, have just posted another library: XPathReader. And as you can guess, I've immediately pulled the code from the article and tried to compile it against .NetCF... The result of the test was that it runs with a few minor changes to the original code!!! And it is more exiting since we get a some XPath functionlity as well. Here you can see a snapshot of the test program running in the PPC emulator:
Don't you love when something useful comes with the source?
You can download the test and a XPathReaderCF projects from here.
The next step will be to test a comparison performance of this library vs XmlDocument in CF.