Spike,

Something to consider when using the nested tree model:  it does not
scale well when implementing large or even medium sized trees.

We originally implemented the ntm for a client, and ran into performance
issues when manipulating tree data (adding new nodes, moving nodes to
new parents, deleting nodes) due to the fact that the left/right border
values for each affected node typically need to be reset to reflect new
node positions (Joe Celko refers to this as the nleft and nright values,
I believe).

If you have small trees, or you don't expect to manage/move nodes that
often, then you should be fine.  But in a large tree that is very
active, reordering the tree nodes adds a lot of overhead that really
isn't necessary.

As an FYI -- our client's tree had roughly 28,000 nodes in it. 

On another note, we used GUIDs as keys for nodes, parents, etc. in the
tree and while they do not take up considerable space, using them does
impact performance negatively when compared to ints.  Try doing some
benchmark tests against a million rows using GUIDs vs. ints and you'll
see what I mean.  It's only not a factor if your tree is very small.

Best Wishes and Regards,  

Abraham Lloyd
Project Leader
www.onproject.com
Working Together, Apart.

onProject.com Inc.
[EMAIL PROTECTED]
w 973.971.9970 ex. 111
f 973.971.9971
c 407.687.4152 


----------------------------------------------------------
You are subscribed to cfcdev. To unsubscribe, send an email
to [EMAIL PROTECTED] with the word 'unsubscribe cfcdev'
in the message of the email.

CFCDev is run by CFCZone (www.cfczone.org) and supported
by Mindtool, Corporation (www.mindtool.com).

Reply via email to