VT100's

[Paul Koning]

> On Oct 1, 2018, at 2:46 PM, Ethan Dicks via cctalk <cctalk at classiccmp.org> wrote:
> 
> On Thu, Sep 6, 2018 at 2:16 PM, Tapley, Mark via cctalk
> <cctalk at classiccmp.org> wrote:
>>        I have more desire to own systems to play on than I have space or time.
> 
> True for most of us, I suspect.
> 
>>        Addressing the former, I have to say my favorite VT-100-alike is a Rainbow. One box (plus monitor plus the dreaded LK-201), three functions in the collection: VT-100 emulation (not perfect but not bad), CPM-80/86 (is that one or two functions?), MS-DOS 3.11b.
> 
> I have only recently learned of the built-in VT100 emulation.  I'm
> curious how it's "not perfect".

I don't know that particular one.  But a possible answer would be: because the VT100 had a bunch of strange corner cases that were not documented and not necessarily well understood.

DEC created an internal standard for terminal behavior; that specification was extremely detailed and very well written.  It became the functional specification for the VT200 series.  I used it to write the terminal emulator for RSTS on the Pro.  It was understood at the time that this spec was close to VT100 behavior (apart from 8 bit characters instead of 7) but not exactly that, and deliberately so.

Similar things have happened in other places.  There is DDCMP, and "DMC compatibility mode" which is best described as "DDCMP with certain bugs".  It hard to find a reasonable description of the latter.  If you want to do DDCMP, you're best off implementing the spec (which is easy) but if you do, it won't work 100% with the "high speed" variant of the DMC-11.

	paul