I’ve been toying around with the idea of turning away builder application into a full game engine, am I wasting my time or would anyone benefit from it?

I believe there was some discussion about “game engining” away3d a couple of years ago being the AB the initial step.

But as you can realize, there isn’t much move around here for a while. So aside from rumors about an update to 4.2 and an eventual release of Away 5.0, its not very clear (at least to me) what is the plan of the away team.

I am not part of the team, so I can only say by my point of view as an ordinary project follower.

What I can say is that currently most of the code available are out dated. I myself struggled through the AB code a couple of months ago, because I couldn’t even have it compiled properly. The same can be said about the Bullet codebase embedded in AwayPhysics.

Also, the AB code has a robotlegs document model that is deeply attached to AWD format (obviously) and I believe it will be pretty hard to adapt it without messing with the AWD format itself.

Anyway, if your engine works well it will benefit someone eventually.
Just remember there are others around (like minko, flare, etc) that does a pretty decent job already. Yours must be better in someway.

Sorry if I didn’t sound very positive, but I still think anything that could help shake up the away community is welcome.

I am far from being an away specialist, but you are welcome to contact me if you wish to share some experiences.

Keep it up.

After much deeper digging into the code your right in some ways, like AB being deeply attached to AWD format. I feel I will need to create a custom program editor, with a custom away3d viewer, as well as creating a new file format to go with it, a few ideas I do have for the engine which I have run test are multi-threading physics, file streaming and AGAL compiling, and currently looking at integrating as3 compiler into an air application to allow attaching scripts to objects like there done in unity3d.