[Simh] Questions regarding future simulator development

dott.Piergiorgio d' Errico dott.piergiorgio at fastwebnet.it
Wed Apr 10 18:38:41 EDT 2013


Il 10/04/2013 18:27, Christian Gauger-Cosgrove ha scritto:

> I have a few quick questions about what could be added to SIMH. Not
> that I could add the features, my skills at C are not quite excellent,
> so in general, I wouldn't really be able to add anything to the
> project... or it would be better to say, I wouldn't be able to add
> anything remotely competently.
>
> Any way on to the questions! How difficult, and better to ask, how
> possible would it be to add the following to SIMH at some point in the
> future?

My wishlist is already known (and debated & criticized) but basically 
what I suggest is:

- a skeleton emulation, having an "empty" (that is, w/o opcodes) CPU.h 
and the basic terminal interface; currently one needs to apply Occam's 
razor to the smallest emulators's sources (currently 8080 Altair and 
LGP) with the risk of inadvertently removing functions actually needed 
(happened to me more than once...)

-a better, up-to date documentation, both user's and developer's

-looking into means to interfacing to console/grh library (implementing 
a curses.h interface can allow emulation of non-TTY consoles, esp. those 
non-80x25, as some earlier S-100 TV/CRT cards have)

-a blinkenlighten interface !!! and also, adding binary input/output to 
EX and DE commands should be useful in dealing with emulation of big irons

-on graphic interface, a vector2raster routine (read: PDP1 Spacewar ;) )

-perhaps isn't a bad idea implementing a discrete logic emulation, or at 
least provisions/hooks for implementing this (similiar to what MAME are 
doing for emulating some earlier arcade machine's sound)

- a minor thing, renaming AltairZ80 in S100 (Herr Schorn's work is now 
much more than an Altair emulation, capable potentially of emulating a 
bewildering array of S100 cards & machines)

Personally, I'm looking conceptually on means of emulating unit record 
equipment and on how to implement their "programming"; provisionally, 
I'm estimating that their cabled "control panel" can be "emulated" with 
a sort of description language, like a circuit simulator. (This can 
enable the emulation of some early computers based on a combination of 
stored program and cabled control panels)

SIMH has as strong point an unique flexibility, and the emulation can be 
carefully configured set prior of booting the emulation proper (I can 
have around even an half-dozen or more of configuration scripts for SIMH 
emulators...) and some of my suggestion are more or less linked to this 
forte

hope to have done a clear wishlist worth of being debated, and

Best regards from Italy,
dott. Piergiorgio.




More information about the Simh mailing list