Comment by phil-martin | Hacker News Reader

Comment by phil-martin | Hacker News Reader

I’m no the target user, I haven’t built games in years, but I loved watching the videos, and your approach of making a sprite editor for doing real work.

Many many years ago I built Quake 3D model editors. One of the features added in was being able to paint on the 3D model and have it update the original texture. The ability to make little tweaks to texture in-situ was really valuable to fine tune the art asset.

When I watched the video of Stipple Effect, I loved the live output window showing the result of the combined sprites.

The first thing that came to mind was “If I was the artist I would love to draw I the result window and have it update the input sprites for me”

I fully get the challenges with implementing that, but it might be a valuable addition to the workflow, being able to work on the art with one less abstraction level

Thank you!

That's a really interesting idea, but unfortunately I don't think it would be feasible to implement.

3D model texture mappings are one thing. It is quite straightforward to create a bijection to edit either the texture or the model with the texture applied and have the changes propagate to the other. For my program, 1) this 2D UV mapping texture to animation workflow is only one specific application of preview scripting, and 2) you would have to take additional factors into account besides the texture and the animation in order to implement a two-way function between the texture and animation in 2D. This is because a pixel coordinate from the texture may map to multiple pixels in the animation. These correspondent pixels in the animation may have different color data, and so propagating a change back to the texture would have to account for which pixel was overridden most recently, or a different type of heuristic.

I hope that makes sense.

Oh it makes total sense, for sure. Sorry if I came across pretending it is simple, it absolutely isn’t.

I guess that’s the role of tool builders - is to find ways to do the impossible to support artist workflows.

The first step is to find out if the end user would even want that workflow in the first place. Maybe it makes no sense, to work that way. But if it is valuable then starts the hard part of figuring out if it is worth the programming acrobatics to implement it.

Reply on news.ycombinator.com