[bdbxml] .NET design questions

All,
	I have a design question from the DBXML user perspective.  I'm using managed c++, so I've written a managed wrapper around the core DBXML functionality.  The wrapper queries the DB and then loads the results into the MS DOM.  I commit the query transaction after loading the latter action.  The higher software layers only see the DOM and perform all modifications on that object.  Later, I transactionally commit the changes by using the object's ToString() method.  This approach creates data hazards, though, as I will eventually have multiple processes accessing the database.  Here are my questions:

1) How do I prevent data hazards (read after write, etc.).  Currently I plan to implement my own locks.  Is there a better solution?
2) Should I wait to commit the query transaction until AFTER I'm through with the data?  I suppose this approach would prevent hazards to an extent.
3) Are there any other issues involved with multithreading the database?


I would appreciate any comments.  Thanks!


Spence Green
[email protected]



------------------------------------------
To remove yourself from this list, send an
email to [email protected]