If I remembered correctly, image API allows you to copy parts of an image. Why not import your font as actual bitmap, copy each 'letter' into an array, and create your own rendering routine?
A couple of reasons. First, I think the font rendering APIs should work. Bitmap fonts should also work and not render the entire environment unstable.
Also, I'm using the dialog extension, and the problem is evident there, too. I'm not going to reimplement the dialog extension, or something equivalent to it.
And, the final reason why I'm not going to roll my own version of a basic piece of functionality that's supposed to just work - I don't have time to do that. I have a deadline to meet, I need to keep some promises, and I'd like fonts to display correctly on supported devices. The project I'm working on needs to be done, one way or another, in a few weeks. It won't get done if I have to repeatedly work around or reimplement large pieces of engine functionality that simply don't work as they should.
