But Sim Coupe duplicates all known characteristics of the real hardware now, right?
I was actually involved in writing emualtors as early as 1998, and most of the early ones were horrible. Pretty much anything by Marat Fayzulin or derived from his source still is — they're usually "run the CPU for exactly this many cycles (usually one scanline of display), then give all the support chips a chance to update their state". Which is obviously just a cop out by unimaginative programmers and in most cases are trivially incorrect. Anyway, I think I've confused myself somewhere else, is Mode 2 no faster than Modes 3 and 4, or is it a little faster but not as much faster as having half the bytes to fetch should imply? On Wed, May 21, 2008 at 5:09 PM, Andrew Collier <[EMAIL PROTECTED]> wrote: > Hi, > > Well - at least, any of the things you can ever change, yes! screen on/off, > border colour, palette, screen mode, ram page, etc. > > Early versions of SimCoupe didn't support these changes better than once per > line. I remember Allan (or was it Simon?) telling me he had never realised > that the E-Tunes player from Fred 63 and on, even had a scrolly message - > because at the time, SimCoupe wasn't able to display it. > >> Also, one further question: am I right to think that the Sam has no >> means of producing interlaced video? > > Correct. Always non-interlaced (despite various examples on Fred of flickery > pictures which purported otherwise). > > Andrew > > > On Wed, May 21, 2008 at 04:36:34PM +0100, Thomas Harte wrote: >> Literally anything? On many of the machines I have written emulators >> for, most things are usually completely changeable but some things are >> loaded internally — especially on any machine that has a variable >> screen start address. >> >> Also, one further question: am I right to think that the Sam has no >> means of producing interlaced video? >> >> On Wed, May 21, 2008 at 3:58 PM, Andrew Collier <[EMAIL PROTECTED]> wrote: >> > On Wed, May 21, 2008 at 03:16:07PM +0100, Thomas Harte wrote: >> >> If you interrupt routine is small, could you not also just switch off >> >> the first one or two scanlines of your display, and even grab a few >> >> extra cycles out as a result? Or can you not enable and disable the >> >> display per scanline? >> > >> > Hi, >> > >> > You could, yes. You can alter pretty much anything about the screen on a >> > line-by-line basis - and if you're very careful, during those lines in the >> > screen area itself! For example, MMENOdemo 1 part 2 (on Fred 60) changes >> > the >> > screen page during the raster; repeatedly switching between foreground and >> > background screens to draw lit and unlit rectabgles - which togehter form a >> > scrolling message. >> > >> > Various other programs switch between mode 4 and mode 3, to get high >> > resolution in a certain area of the screen - Edwin Blink's menu on Blitz >> > (issue 6, was it?) and David Laundon's Fred menu come to mind. >> > >> > Anyway - in this particular case I wanted to display the whole screen; and >> > in >> > the case of one or two scanlines, it probably takes longer to set up the >> > line >> > interrupt handling, than you save by having the screen turned off. >> > >> > Cheers, >> > Andrew >> > >> > -- >> > --- Andrew Collier ---- >> > ---- http://www.intensity.org.uk/ --- >> > -- >> > r<2+ T<4* cSEL dMS hEn/CB<BL A4 S+*<++ C$++L/mP W- a-- Vh+seT+ (Cantab) >> > 1.1.4 >> > > > -- > --- Andrew Collier ---- > ---- http://www.intensity.org.uk/ --- > -- > r<2+ T<4* cSEL dMS hEn/CB<BL A4 S+*<++ C$++L/mP W- a-- Vh+seT+ (Cantab) 1.1.4 >
