There is no such thing as a serious ASM OS

[AlonzoTG]

I noticed the postings about pathetic little assembly OSes and had to make a response. Anyone who thinks an assembler OS is worth-while knows alot but understands little.

The key here is an understanding of computers and what they are. They are best thought of as language processing devices. So any sucessful OS will be designed to support the broadest variety of high-level languages.

Furthermore, to permit the advancment of hardware design, it is necessary that as much of the OS as possible be written in a high level language so that it can be easily ported to new platforms.

I have spent many years designing my own OS, Sphere. The design is VERY good though not quite perfect. Sadly I have neither the resources nor the talent to code even a prototype on my own. =\

[suselinux]

Anyone who thinks an assembler OS is worth-while-----understands little.

I have neither the resources nor the talent to code

So Calum dosen't understand anything, but nieither do.

Your argument is a little thin

[mobrien_12]

Someone told me that os/2 Warp 4 was an ASM OS.....

I'm doubtful, but I don't know enough about it.  Can anyone confirm this?

[Laukev7]

I used OS/2, and it's certainly not assembly, since there is also a very rare version for PPC. Like Windows NT, which contains OS/2 code and has a similar API, it's probably made of C/C++, and a bit of REXX.

[suselinux]

I used OS/2, and it's certainly not assembly, since there is also a very rare version for PPC. Like Windows NT, which contains OS/2 code and has a similar API, it's probably made of C/C++, and a bit of REXX.

OS/2 for PPC.....really? when?


did you imply that NT was on PPC as well?

[AlonzoTG]

Actually, to contradict my previous post to a point, OS/2 is a very strange OS. It was co-developed by IBM and Microsoft back in the '80s as a replacement for DOS. It was designed with a high-performance object-oriented C API. However, as was the common practice in the '80s, it did rely heavily on code written directly in assembler.

However, for reasons of cost and portability this practice should not be repeated in this day and age.

[Laukev7]

OS/2 for PPC.....really? when?


did you imply that NT was on PPC as well?

IBM was at the very least developping a version of OS/2 for the PowerPC (which was, after all, their own processor). That was during the days of the RISC hype and the rise of the PPC. Check the package os2ppc.zip here:

http://www.os2site.com/sw/hardware/ppc/

And yes, Windows NT 3.51 was also available for the PPC, and the SPARC, MIPS and Alpha processors as well.

[choasforages]

FOOL. the best operating system is the one that does exactly what it is needed to do. will java work to well for an ECU? i doutb it, so "pathetic little assembly OSes" is what keeps your ass moving on the roads with computer controlled vehicles. if you talk about pre computer applications, they are the ultimate example of programing. they put the whole engine control system into hardware. no python, perl, java, .net. just pure mechanical engineering, which is what most issue's break down into anyway.

[Calum]

this is interesting to say the least.

i agree with grimes' comments, strangely enough because i understand what he is saying. i know exactly what he means when he says that someone making a system purely in assembly has a lot of knowledge but not much of a clue (or whatever he said) however you know i think the responses have a lot of good points too, and i think the assembly OS debate has developed enough that i want to just list what i think are the main points if i may:

* if they get big assembly language systems will be really difficult to overhaul

* assembly language systems (and programs) are really stable and fast if they are written and maintained properly

* assembly language is NOT portable for fairly obvious reasons. if you write in the language of one chip, you must rewrite from scratch when talking to another chip

and that's mainly it. so what i am saying is that so long as whoever is writing a aystem keeps those things in mind when they write their system and evaluate their needs,fine. here are my thoughts about assembly in real world systems:

* general purpose GUI based desktop PC system - NO!!! read the reasons above!!!

* GPS reception system - YES! it's not portable but by the time a new GPS client hardware comes out, you'll be wanting to rewrite your software anyway

* embedded system with minimal memory - great! we can make the most of what we have

see what i mean? there's a time and a place, and i don't like seeing people who like the idea of an asm OS getting verbal abuse.