Hello Troels,

On Sun, 2 Dec 2007, troels knak-nielsen wrote:

> We're about to implement an application, which will use a tree 
> structure as the main taxonomy. Instead of implementing the algorithm 
> ourself, we are considering using the tree from ezc. We are keen on 
> using a modified preorder (nested set), but looking through the table 
> at http://ezcomponents.org/docs/tutorials/Tree I had a few questions. 
> When the phrase "Simple operation" is used, I suppose that it means 
> that it's "cheap" to execute -- Not that the implementation is simple?

Yes, it's often a simple query, which uses an index of some sorts. As 
the queries are simple, so is the implementation of course.

> Also, does "simple" mean that it's a constant cost or what happens, 
> when the size of the tree scales? I'm sure, an in-depth answer gets 
> rather complex, but some indication of how each operation scales, 
> would be most informative.

That's basically what the table tries to say without reverting to O 
notations. The value "Simple Operation" tells you that the backend only 
has to do one (or two or three) queries to get the result - and that the 
number of queries is always constant.

> For example, I'm not entirely sure, what the implications of using 
> LIKE on large datasets is? I suppose, that in theory, LIKE should be 
> able to perform reasonably well, since the match is anchored, but I'm 
> not sure how well MySql (We use MySql) works.

I'm not a real MySQL guru either, but I could run some benchmarks on 
this next week I suppose. If you have a more specific question about 
the cost of an operation, I'd be happy to explain you the algorithm.

regards,
-- 
Derick Rethans
eZ components Product Manager
eZ systems | http://ez.no
-- 
Components mailing list
[email protected]
http://lists.ez.no/mailman/listinfo/components

Reply via email to