congrats on all the participants for the entries and therefore upping the charity!
I won't make it in time, but I'd like to share some insights gathered using Construct 3 for a few days now.
Coming from C2, I must admit that I have been skeptical adopting to the follow-up, especially when per-se red flags like browser only, annual subscription etc. are raised, essentially the points that sparked some discourse with C3's release. This though was before the release of the C3 runtime.
Working with the C3 runtime + latest beta additions (the return of Z elevation) this weekend, it's safe for me to say now that my concerns have been ... mercilessly eradicated?!
A few hours in with a dev cycle, there is already an evident difference in performance. And by that I mean massive performance boosts, mostly gained through things like optimized for-each loops and built-in per tick object addressing, built-in colorizer (instead of tint, you can even boost the color range through events), spritesheet-esque additions to tiled backgrounds (offset, scale, angle, so basically sort of UV mapping), proper control of 9-patches (angle etc), automatic z-sorting through z-elevation, z elevation per layer, etc. the list goes on.
Workflow-wise I left with the impression that the tool as a whole is somewhat more responsive, fluid and clean in UI / UX design, so there is also definetely an improvement that pays in time saved getting things done.
The project I'm on now is an arena shooter that stress-tests the C3 runtime, with sometimes 1000+ enemies coming at the player + effect sprites, particle effects + all the logic. Having a similar build on C2, I'm coming from 30+ % CPU utilization to now 14%, leaving enough headroom for some either crazy effect additions or, well, unefficient code :D - always 60 fps, no frame drops.
So long story short, the Game Jam as a marketing effort left me impressed with the potential and I'm strongly considering adoption now.
One thing remains though: System expressions for raycast please! ;)