Going year back, I knew I would take a somewhat a risk by using Leadwerks engine (that was – and still is – in development). My main reason for usin LE boiled down to being able to get more done in shorter timeframe – the learning curve for me was minimal.
I knew a compatibility was an issue, and about 6 months ago (or so) I was re-checking if I should abandon the engine for good, as there was a major change that greatly impacted in my development.
I recently got a community version out, which can be called to be a “release candidate” (although it was missing some content). While testing the version, it was spotted that there was a driver bug. The game graphics wouldn’t work on new gfx drivers.
I wrote about this in the Leadwerks forums and it was revealed to me that in 2.25 this was fixed (I was using 2.24), and that I’d need to do another upgrade. The last upgrade I did took quite a bit of time (since for example physics system had changed, and supported graphics file format had changed). Now, I started doing the upgrade and the next step is to get the graphics to show (meshes are not shown anymore). This upgrade is again an essential one, since if nobody can play the game with new drivers… nobody probably will be interested in buying it (I wouldn’t buy those flickering shadows).
I’m working on this to get the upgrade done, and then I’m much closer to doing the actual release and getting people to buy the thing.
The lesson from all this goes:
Whatever engine you plan to use, make sure you don’t rely on getting “engine patches” later. Make sure the engine is so stable that you can rely on using the one version (the version of today, not version of the future) to release your game.
The costs of relying that “engine will be patched” are risky. It may be suitable for you, but I would be very careful on this matter. It’s better to pick a stable and perhaps older engine over a new, fancy – but in development – engine.
At least in my bloody humble opinion.