Thanks for the info, we're looking into it.  It's a similar crash to what we've been getting reported, we just need to figure out why it's happening.

  Postign all your crash reports has been awesome.  The error we are fighting doesnt happen when the crash occurs.  The error happens well before and its only when the load/save/autosave process goes through to write the data that we get the crash.  Which makes it really tricky to track down what is corrupting the data in the first place.  But getting a bulk of crashes so we can try to find patterns to narrow our search is awesome and greatly appreciated.