NEC V20: BRKEM, CALLN and RETEM

[krebizfan]

No, no no. Don't start using markup languages in the 70s. My entire career in the late 80s and early 90s was built around getting bad software to work with markup languages.

 

[Chuck(G)]

Are you referring to ANSI X3.159-1989 or ANSI X3.64?

The former. Heck it doesn't matter today, anyway, what with script kiddies running the show. But early K&R really was full of pitfalls. I recall that on USEnet, I think it was Brian who would occasionally post a statement and ask "what does it do?" I still have a bunch of code I wrote from 1982--it looks like foreign dialect to me now.

Since much of m own work involves MCUs, I stick with C--the toolsets are written for it. And Linux always has a C compiler in the standard installation. No sense in straying from what everyone understands.

 

[kerravon]

The former. Heck it doesn't matter today, anyway, what with script kiddies running the show. But early K&R really was full of pitfalls. I recall that on USEnet, I think it was Brian who would occasionally post a statement and ask "what does it do?" I still have a bunch of code I wrote from 1982--it looks like foreign dialect to me now.

That's fine - my question is more about ANSI X3.64 than K&R 1 vs 2.

So the code from 1982 - in the alien language - hell, it may as well be Pascal - is there any way to make a Pascal etc program use ANSI escapes on a 4.77 MHz 8088 and be able to be fast enough (so that the user doesn't care - even if it is slower) to not require direct screen buffer manipulation?

Assuming you're willing to change or bypass (by hooking an interrupt) the OS or replace ansi.sys and potentially change the C library too. In fact, that's another issue - I basically want a 16-bit msvcrt.dll rather than statically-linking the C library to every application.

I heard that there were computers competing with the IBM PC at the time that actually allowed even more than 640k, but the direct screen writes were what won the day for IBM. I can find the names if required. Rainbow? Olivetti? Just guessing.

 

[Chuck(G)]

(Anent display speed: you forget that there were x86 systems out there using serial terminals.) At any rate, I know of at least one ANSI escape sequence driver that bypasses int 10 BIOS services completely and goes for direct hardware access. It may (but i"m not certain about that) also manipulate the 6845 display start register to handle scrolling. So with with a Pascal, Modula-2 program driving the display with VT100 escape sequences might actually be quire tolerable, particularly in MDA.

 

[Svenska]

So the code from 1982 - in the alien language - hell, it may as well be Pascal - is there any way to make a Pascal etc program use ANSI escapes on a 4.77 MHz 8088 and be able to be fast enough (so that the user doesn't care - even if it is slower) to not require direct screen buffer manipulation?

Are you talking about an MS-DOS machine or the IBM PC in particular? Huge difference, especially in its early years. I've got a machine running MS-DOS (8 MHz 80186) which does handle ANSI escape sequences in its BIOS, no additional software required (it can also do VT-52). It's not fast; the drawing speed is comparable to a serial terminal at 9600 bps.

I know of a few ANSI.SYS replacements which access the hardware directly for speed. Obviously, they won't work on that system - and many don't do MDA, either.

Assuming you're willing to change or bypass (by hooking an interrupt) the OS or replace ansi.sys and potentially change the C library too. In fact, that's another issue - I basically want a 16-bit msvcrt.dll rather than statically-linking the C library to every application.

Do you want it in 1985 or 2023? That's a relevant question.

I heard that there were computers competing with the IBM PC at the time that actually allowed even more than 640k, but the direct screen writes were what won the day for IBM. I can find the names if required. Rainbow? Olivetti? Just guessing.

I'm very sure that most systems could reach a similar or better drawing speed than the IBM PC using similar approaches. It's not an IBM exclusive feature - but the ubiquity of IBM PC (and its clones) was.

Also, I'm relatively sure that you don't want to use a universal graphics standard designed in 1984 today. Oh, and on a less technical note, you can actually answer to multiple posts in a single post; no need to post multiple replies separately. Just click on "Quote" for the relevant post, then "Insert quotes" in your answer window. Quite handy, and better for everyone.

 

[kerravon]

Are you talking about an MS-DOS machine or the IBM PC in particular? Huge difference, especially in its early years. I've got a machine running MS-DOS (8 MHz 80186) which does handle ANSI escape sequences in its BIOS, no additional software required (it can also do VT-52). It's not fast; the drawing speed is comparable to a serial terminal at 9600 bps.

I know of a few ANSI.SYS replacements which access the hardware directly for speed. Obviously, they won't work on that system - and many don't do MDA, either.

I have only seen IBM PCs being used. I only got my XT in 1987 and that is when I saw people writing applications that didn't work on the Amiga. I was expecting a simple recompile of a (draft-standard-compliant) C program.

Do you want it in 1985 or 2023? That's a relevant question.

I want to prepare for a rerun of the 1980s on the assumption that I will have access to a time machine if I ask the gods nicely. But in 2023 I need to gather the knowledge before I step into the machine. I still don't understand what I'm going to encounter when I arrive in the 1980s. Is it a cultural or technical problem? As early as possible I want to get people to switch to ANSI X3.64 in preparation for recompiling on the Amiga so that the Amiga is a viable alternative. And that means (at a minimum) ANSI X3.64 on an XT needs to be acceptable performance.

Oh, and on a less technical note, you can actually answer to multiple posts in a single post; no need to post multiple replies separately. Just click on "Quote" for the relevant post, then "Insert quotes" in your answer window. Quite handy, and better for everyone.

Do you want me to produce a single reply to messages from multiple people or just a single person? I don't think the former sounds very good.

Sorry for the delay in replying - other priorities.

 

[Svenska]

I have only seen IBM PCs being used. I only got my XT in 1987 and that is when I saw people writing applications that didn't work on the Amiga. I was expecting a simple recompile of a (draft-standard-compliant) C program.

Either you do direct hardware accesses (which is guaranteed to not be standard-compliant), or you use the libc functions (which is standard-compliant). The latter will handle ANSI if your underlying output stream supports ANSI. Which is not the case on IBM PCs without ANSI.SYS (or a replacement for it).

ANSI.SYS has the advantage of using only BIOS functions, making it compatible with all MS-DOS machines out there. It also has the disadvantage of only using BIOS functions, making it slow. Other applications bypass the BIOS, making them fast and incompatible. Choose your poison.

As early as possible I want to get people to switch to ANSI X3.64 in preparation for recompiling on the Amiga so that the Amiga is a viable alternative. And that means (at a minimum) ANSI X3.64 on an XT needs to be acceptable performance.

I don't see ANSI X3.64 a viable solution for an Amiga without giving up on the main advantages of the Amiga itself. Basically, you are trying to reduce all machines to the capabilities of a text-only terminal. Doesn't seem like a good idea.

Do you want me to produce a single reply to messages from multiple people or just a single person? I don't think the former sounds very good.

One single forum post can contain multiple quotes from different people, and each quote box lists who it is from. It's quite natural way to respond to multiple posts with a single post instead of having many short posts by the same person following each other. At least that is my opinion. Double posting used to be frowned upon.

Sorry for the delay in replying - other priorities.

Yeah, the 80s can wait. They've got plenty of time.