On Mon, 8 May 2000, Bart Oldeman wrote:
> 1. the README says:
>
> Borland's BGI requires the old IBM 8x8 ASCII font table to be present
> at memory address F000:FA6E. This hack loads a 8x8 font table at the
> expected addresses when VGA mode 12h is entered. (I sure hope nothing
> important lives there.) The font I used is ugly but free.
>
> I wonder whether this is necessary (it is not for me). It used to be in
> 0.98.*, but was fixed somewhere in the 0.99 series, and now copies this
> font from the real BIOS. Or are there some BIOSes which don't include this
> font any more?
Don't know. Wasn't an issue for me, But I'm not sure what hardware
David Hindman was using.
>
> 2. Isn't there some double work going on here w.r.t. the cpuemu code,
> which is present in 0.99.x but not in 1.0.0?
>
Quite possibly. Neither David or I have really been following
dosemu development. When 1.0.0 came out and and mode 12h wasn't
supported, we looked at the problem from the 1.0.0 code point
of view.
Basically we worked out vmware worked, then so did bochs. Bochs
was now lgpl and so the source was there. At that point David Hindman
really quickly went ahead and produced an implementation.
If there is a wider cpuemu effort, this code should definitely
be rolled in there. This really has been one of the "scratch your
own itch" implementations.
(apologies to people working in the cpuemu effort, we really
didn't mean to duplicate other work)
David Pinson
