> On Jul 16, 2018, at 3:35 PM, David McMackins <cont...@mcmackins.org> wrote: > > I was replying to Jerome who is the original author of the Pascal one. > > I get that VGA is widely available. I'm writing a library, so I want to > support as much as possible. My intended targets (in the order of planned > implementation) are VGA, VESA, CGA, and EGA. I want to have contained tests > for each of them (and separate for different VESA modes, but I already have a > document on doing that) so that an application could list to the user all the > supported modes for their hardware.
You are testing on real hardware... Not a VM or DOSBox? There are several possibilities, none really more likely than the others. I goofed nearly 30 years ago and a bug was undetected until now. Not impossible. I ain’t perfect. On top of that, I don’t think code ever ran on anything less than VGA. So, there could be issues with MCGA, CGA, Hercules and the like. The original reference material was wrong. Oh, well. Would not be the first time. For a lot of the earlier stuff, there was sometimes several ways to do tests. Some better than others. It is possible the method used in that function to check for EGA was not the became invalid with later hardware. Your VGA card does not respond correctly for this test. Also, not impossible. I once had to write a TSR video BIOS for my 286. Whenever you polled it for current color RGB values it returned pure white for each. So, anytime I exited a game I got a white screen of death. Although looking at Mateusz excerpt from PC magazine, it seems very similar. So, it’s probably a bug. You should probably just assume support for a minimum of EGA and test for VGA and better. This wasn’t even an issue in 1990. But, if you insist on supporting MCGA, CGA and others. Try some of the other tests and hopefully they’ll work out. ------------------------------------------------------------------------------ Check out the vibrant tech community on one of the world's most engaging tech sites, Slashdot.org! http://sdm.link/slashdot _______________________________________________ Freedos-devel mailing list Freedos-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/freedos-devel
