Re: tree model complaints

[D M <isaac belial pl eu org>]

On Sun, Aug 15, 2004 at 03:16:26PM +0100, Tim Müller wrote:
> On Sunday 15 August 2004 11:37, D M wrote:
>
> >     I'm implementing a custom tree model in my application. And one
> > thing itrigues me a little - why does it involve implementing
> > a functions which are complementary - getting n'th child is kind
> > of equivalent of getting first child and then iterating
> > over consecutive elements. 
>
> Just because the result is the same doesn't mean that those are equivalent 
> actions. Imagine a model that can access row data in a way where accessing 
> the 9999th row takes more or less the same time as accessing the 1st row. In 
> this case it would be madness to iterate through all previous 9998 rows just 
> to retrieve the 9999th row's data. In other words: it depends on the 
> underlying data structures whether those two methods of accessing the n-th 
> row are more or less equivalent or not.
Yes, thats true, but my point is :no matter how underlaying data 
are structred I have to provide both methods ( or even three 
- path to iterator conversion)

-- 
DM