Yes. I agree it is complex. We have given it fairly a bit of though and if you want to support legacy applications then it is pretty much about the only way to do this.
Sure we could go LinuxBIOS version of grub, but unless you add "legacy application support" it won't help you with booting anything else other than linux. So while it may fit with your particular problem, usefullness of such hack won't be as big as one could expect.
-- Adam Sulmicki
Well, I certainly did not mean to imply ADLO or BOCHS was complex unnecessarily. These are brilliant developments that greatly advance the capabilities of the open source community.
My point was that I am only focused on booting linux, and so the legacy BIOS is just flash baggage after linux gets going. Grub offers some booting flexibility that I need, and since it is C code, also is more easily customized for some other things I need to do during boot (embedded application). So the interest in a non-BIOS adaptation (I prefer that word to "hack"). It doesn't really matter to me if it has wide interest or not, as long as it solves my purposes and doesn't take too long to do and isn't really stupid for some reason or other. I just wanted to see if it was already done or there was a giant flaw in my thinking. Based on a previous post, perhaps the tiara project may have already done it, so I am going to check that out.
Thanks for your excellent work on ADLO and mods to BOCHS-BIOS.
-Steve