Mapping

Will code for food!

A lot of code written today just moves data from one place to another and/or maps data from one structure to another. Right now I’m consulting on a project where the code I write is:

converting an socket stream to an xml document
converts the xml document to c# objects
converts the c# objects to linq objects (created by a database scheme)
uses the linq objects to update the database and something for going the other way as well.

The database scheme was written by me by hand based on the xml document.

This particular process could/should be totally automated since there is a 1-1 mapping from one way to the other. There are so many errors that can be made during the writing of this code (note: linq brings me some of the way - but not all, and I notice this pattern in modern programming WSDL's, POM's etc. etc.).

The real core of Dawn (the mindset behind it) is mapping. The mapping is coming from the library and the flowing of data. Conventions of mapping should be created and used when no special requirements are present. However we are forced to hand code the mapping over and over, hereafter debug it over and over.

I’m tired.