пт, 15 мая 2020 г., 18:56 InkLab App <inklab...@gmail.com>: > > For 'layer effects' : do you mean, non destructive filters? If so, this > was a very interesting problem that I have been considering for a while, > that is, implementing arbitrary functions into a file format. IT seems very > difficult to do safely, but I did have a few ideas. I would love to hear > your thoughts on this. The same idea could be applied, in a more limited > scope, to define generalized layer filters. >
By the way, I think we missed an important stage here: (re)defining the goals of ORA. It used to be a long-term archival file format. Then suddenly it turned out to be a somewhat convenient file format for basic multi-layer project data exchange between applications (Krita, GIMP, and MyPaint were first to implement support for it, I think). At some point, Krita started pushing it forward with new features in their own namespace (extra blending modes). Other projects haven't followed suit yet, as far as I can tell. So what do we actually want ORA to be and why? If we want it to become a lingua franca multilayer project file format w/ basic vector graphics, a kinda libre PSD, then it makes sense to list certain expectations based on real-life use cases _before_ jumping to features implementation. Also, if/when we add features like layer effects, how much (many?) roundtripping errors can we realistically live with? Because you _will_ get those unless you standardize a fixed set of effects between applications (settings and semantics, rendering model etc.). P.S. As for speed etc., here you go: https://blender.stackexchange.com/a/148466/213 (pointed out by Troy). Alex
_______________________________________________ CREATE mailing list CREATE@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/create