Autolayer merge done

Boudewijn Rempt boud at valdyas.org
Sun Jan 16 09:28:36 CET 2005


On Sunday 16 January 2005 09:21, Casper Boemann wrote:

> Yes KisRectIterator is broken. I never fixed it because I realised that the
> usage is quite limited so I implemented KisHLineIterator instead.

Maybe we should just get remove it then...

>
> The usage problem with KisRectIterator is that you dont know which pixel
> you have so unless its an one pixel function like change color it's no use.
> You could call x() and y() of the iterator, but that feels quite slow and
> clumpsy.
>
> Anyway if you think there is a need i'll fix it, but for now just use
> KisHLineIterator or the the familiar ones (now) build on top:
> KisIteratorPixel or KisIteratorPixelSelection (can't remember if thats
> their real names). Those are currenly broken according to boundwijn but
> I'll take a look, if I get a description of the problem... The selection
> one is probably better left to cyrille to fix

Well, the high-level iterators should use the low-level iterators to get at 
the array of Q_UINT8 to convert to KisPixel's using the colour strategy. I 
don't know how to do that, so I didn't.

The selection iterators do the same as the high-level iterators but over two 
paint devices: the layer one and the selection. I'm not decided what they 
should return, there are a number of possibilities:

* a KisPixel with the opacitity modified by the selectedness (i.e., not 
selected, completely transparent, completely selected, its own opacity).

* a KisPixel with an extra channel for selectedness.

* skip the pixels that are not at least X selected.

I'm leaning towards option 2, but I'm not sure...

-- 
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/20050116/f34ceb07/attachment.pgp


More information about the kimageshop mailing list