Dirk Reiners wrote:
Hi Allen,
On Tue, 2005-04-19 at 14:04 -0500, Allen Bierbaum wrote:
It appears from the API documentation for CoredNodePtr that this code
would issue a begin edit on the "node" part of the cored node pointer
and not on the core part (because of the operator FieldContainerPtr).
You missed a little detail at the bottom of OSGCoredNodePtr.inl. ;) The
documentation is non-existent, really, as I just added that quickly to
see if it works.
Ok. That looks good. I will start using it right away. :)
This seems to be exactly what I would not expect it to do. :(
I think there are 3 cases for what it could do:
- flag the core for edits
- flag the node for edits
- flag both of them for edits
Since the dereference operator makes the root_transform "look" like a
core pointer I would expect that most people would think of the first
option being the one to happen.
To keep things as simple as possible it actually does 3. Which is a
little inefficient, as it might flag changes that didn't really happen,
but it was the simplest thing to do.
I can see a point for 1 though, if 3 is deemed too expensive.
I think 3 is fine. It is closer to expectations and it shouldn't be too
much more expensive.
Thanks,
Allen
Dirk
-------------------------------------------------------
This SF.Net email is sponsored by: New Crystal Reports XI.
Version 11 adds new functionality designed to reduce time involved in
creating, integrating, and deploying reporting solutions. Free runtime info,
new features, or free trial, at: http://www.businessobjects.com/devxi/728
_______________________________________________
Opensg-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/opensg-users
