I am bringing NHibernate into an existing project where there is a pre-
existing roll-your-own persistence.  These two persistence mechanisms
have to co-exist. The application is a single user, desktop
application.  It is not a typical line of business application but
more of a designer tool for specialized hardware.

To date and in the preceding generation of application, the user goes
about their work and then does a "Save".  We can't get away from the
Save model as this is what the users have come to expect.

1) Should I be building the persistence components to avoid actually
calling Save methods until the Application calls Save, then
explicitely call Save?

If I go this route then I lose some of the goodness about the
generation of IDs, move to assigned IDs, and then  have to track if it
is a new\update.

I had a thought that maybe I could be using a session (keeping it
open) and only flushing the session on save.

2) My colleague feels that on update (FileSave), both persistence
mechanisms should share the same transaction.  My thought is that from
an application perspective the object models represented by differing
persistence models don't relate except at a very high level, so they
don't need to be in the same transaction.

I am also having some difficulty in thinking that a transaction needs
to wrap up all of the activity between saves.  A single error might
roll-back a whole lot of un-related activity.  And, in the grand
scheme of things we aren't doing banking here...

I am still new to the NHibernate model so I appreciate any comments
that anyone may have.

Thanks,
jeff

--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups 
"nhusers" group.
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to 
[email protected]
For more options, visit this group at 
http://groups.google.com/group/nhusers?hl=en
-~----------~----~----~----~------~----~------~--~---

Reply via email to