We are reading all the feedback in this thread and we're still interested in hearing more. However much like the feature suggestion platform, it's challenging to deal with the volume and scope of comments. People sometimes suggest things encompassing year's worth of work. It can be difficult both for users, and even for us, to establish how long anything will take or how feasible it would be. Some things would take several weeks just to review, understand, design a plan for, and investigate the necessary technical work involved. Often what sounds like a throwaway remark or a "simple" feature can open up a whole load of unexpected challenges, raise the prospect of fundamental redesigns, or have complicated tradeoffs like simplifying one specific type of use while making another type of use more difficult. All this plus the fact we get huge amounts of feedback and have limited time makes it difficult to respond to every single comment. However we are reading and we are interested in more feedback.
Just on drawing directly on the layout, I like the idea and I appreciate how it would be useful to animators who want quick results. However I think it would involve a fundamental reworking of the editor, which makes it difficult to act on. The Animations Editor is complex and has already had year's worth of development on it. It has loads of user interface sections that would be challenging to fit in to the existing interface without just making a complicated mess. It's also not clear to what extent its features apply to free-form drawing (what about frame-by-frame animations? what does cropping or resizing mean in a free form drawing? are collision polygons necessary to? etc. etc.) It would be good to discuss all this if we were building from the ground up. But the practical reality is we're working from the starting point of Construct 3. It would be much more feasible to think about how to make it easier to use what we've got. I know that imposes limitations, but meanwhile I've seen other animation tools that have nothing remotely sophisticated as Construct's Animation Editor for drawing and seem to rely largely on importing resources, let alone Construct's event sheets for interactivity or procedural logic. We're not trying to out-do Adobe Animate, just like Construct 3 isn't trying to out-do Unity. Our goal is to design something that is useful to at least some portion of the market, and I am confident we can do that, even if it doesn't do everything imaginable.