Layers, layergroups, dockwindow, preview and dynamic layersize
cbr at boemann.dk
Sat Jul 3 16:06:54 CEST 2004
> No, I'm in favour of using layer-space coordinates for the paint devices,
> use paint device coordinates for paint devices. It's simple to give the
> these coordinates, as well as image space coordinates if they need them.
> I also think using the extent functions makes sense since then the layer's
> and y give the coordinates of its origin, which will only change when the
> layer's moved. It seems a cleaner way of doing things than having x and y
> (potentially) changing as you extend the layer.
> I think what you've done is fine, my main concern is that krita is already
> complex application and changing a core part of it can introduce many
> problems. But that doesn't mean it shouldn't be done, and it doesn't mean
> should be done in a different way.
Thanks Adrian. I must have misunderstood you previously regarding layerspace
Given that this is also what I have implemented and feels is the the right
thing I strongly favour committing my patch. If we were to require
paintdevice to take imagespace coords we would break encapsulation and
compartmentalisation rules. In this respect my patch is more clean. Wishing
for a facade is a good thing but this should be to hide the tile structure,
not to swallow things outside the paintdevice.
just my two cents worth.
Boudewijn, could you please make a decision, as I'm not getting anywhere
waiting. I would really like to finish this layer thing this weekend, and I
have the time and will, but unless you decide I'm stuck.
More information about the kimageshop