Update on: For the next release

[Boudewijn Rempt]

On Monday 22 August 2005 10:41, Casper Boemann wrote:
> On Monday 22 August 2005 09:24, Boudewijn Rempt wrote:
> > On Thursday 11 August 2005 14:04, Boudewijn Rempt wrote:
> > > * Colorspacee independence 20%
> > > 	- make a base class for 8-bit/int, 16-bit/int colorspaces with common
> > > code
> >
> > The base classes are created, but there's not much common code refactored
> > into the base classes.
>
> just thinking that maybe it isn't as useful after all.

It's not useful for every colorspac, but it can help a lot with the more 
common colorspaces, like gray, rgb, cmyk, lab, xyz, yuv: all those 
colorspaces can exist in 8/16/32 bits/channel.

> Some colorspaces could have both 8 bit  and 16 bit and what else. We would
> have to do multiple inheritance then. No problem, just remember to call the
> functions different things, but that would a problem if we want to template
> actual colorspaces.

If your colorspace isn't one of these (like wet, wet&sticky or test) then you 
inherit the abstract colorspace and implement what's necessary.

> I assume that the intention was to move kis_integer_math.h into a class

I haven't given kis_integer_math.h much thought yet.

> Anyway. Just wanted to say that there is more to this, and that we should
> be careful.

It if it doesn't work out, we can always go back.


-- 
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/20050822/dd6f335c/attachment.pgp