Looks like a release schedule to me

[Boudewijn Rempt]

On Monday 07 March 2005 02:08, Casper Boemann wrote:
> > > This could be done in to steps like in fill and gradient.
> >
> > I disagree, I think that something that should be hide inside the
>
> iterators.
> I (dis)agree with both of you. The iterators should provide this, and it
> would probably be the most common way to do it, but we should not disallow
> other ways like the two step approach.
>

I agree with Caspar: it's important to have the flexibility. The fact that a 
selection is a paint device makes it very easy to optimize working on areas. 
But if I could find a way to add the selection value to the pixel iterators, 
most filters would have an easier time of it.

My problem was that the pixel iterator does not encapsulate the underying 
paint device iterator but inherit them. Encapsualtion is necessary to make 
the pixel iterator read from two iterators (selection and paint device) and I 
didn't manage to create an iterator that did that: I also suspect that 
encapsulation would make using an iterator a lot more expensive, perhaps even 
more expensive than having two seperate iterators, one for the selection, one 
for the paint device.

-- 
Boudewijn Rempt 
http://www.valdyas.org/fading/index.cgi
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: not available
Url : http://mail.kde.org/pipermail/kimageshop/attachments/20050308/21ff7aad/attachment.pgp