can an os be wrritten in 6 months time

[Brynet-Inc]

AVINASH posted that in 2001... For what purpose is there to revive such old topics and argue with someone who probably isn't even visiting these forums?

[mathematician]

I reckon any of the long term OS Developers here could do that in only a few months.

You wouldn't want me to hold you to that would you? I would guess that it is probably around 50,000 lines of assembly code you're talking about.

[Combuster]

Far from all OSes are 100% assembly, Mine is rare enough to have the entire kernel in assembly.

For that matter, C kernels are easier to build once you know how to start. Plus, that you can borrow code from linux and earlier work and such

[mathematician]

But MS-DOS was 100% assembly. I doubt whether there were any compilers around when version 1 was written.

[Tyler]

I reckon any of the long term OS Developers here could do that in only a few months.

You wouldn't want me to hold you to that would you? I would guess that it is probably around 50,000 lines of assembly code you're talking about.

Well i wrote a compatable to MS-DOS four in four months, so assuming MS-DOS 1.00 contains less features, i'll happily take up your challenge next time i win the lottery and can quit my job in order to work on it full time for a few months.

[Tyler]

But MS-DOS was 100% assembly. I doubt whether there were any compilers around when version 1 was written.

...

...



...

http://en.wikipedia.org/wiki/Compiler

...

Only 30 years out.

[Brynet-Inc]

Are you kidding mathematician? No C compilers when MS-DOS 1 was written?

The language has been used for quite a long time, It was designed by Dennis Ritchie in 1972.. He along with Ken Thompson and others developed Unix as well.

And even MS-DOS in it's later incarnations used the C language.. Most of the unities included were though.

http://en.wikipedia.org/wiki/C_%28progr ... anguage%29 --I advice you to read this...Carefully.

[urxae]

Perhaps he meant "compilers targeting the 8086" instead of compilers in general?

[mathematician]

Are you kidding mathematician? No C compilers when MS-DOS 1 was written?

Don't teach your grandmother to suck eggs. I know where the C language come from. Prior to the days of ANSI C it was difficult to read anything about C without repeatedly tripping over the intials K&R; I just assumed that you would have been able to insert the words "for the x86" yourself.

MS-DOS programs only had 640kb of memory available to them, and even that was partly used up by the operating system. If MS-DOS hadn't been written in assembly, there would have been very little room left for anything else.

[mathematician]

I reckon any of the long term OS Developers here could do that in only a few months.

You wouldn't want me to hold you to that would you? I would guess that it is probably around 50,000 lines of assembly code you're talking about.

Well i wrote a compatable to MS-DOS four in four months, so assuming MS-DOS 1.00 contains less features, i'll happily take up your challenge next time i win the lottery and can quit my job in order to work on it full time for a few months.

How may lines of assembly language did that run to? The shell alone cost me 20,000; admittedly it improved slightly upon the default option. Hopefully you will have included code to locate the InDOS byte, otherwise you would have had my TSR's crashing all over the place.

[Tyler]

I have no idea how many lines of assembly it would have taken... as i wrote it in C, it was specification compatable but i doubt it would have actual run on a DOS period computer. As for undocumented stuff, i purposely included very little because i only wrote it to run old Apogee DOS games fast with the ability to easily implement a background watch on the code (something i never actually did)

Unfortunately for the Compiler argument; you stated neither that the compiler was C or x86... Making you 30 years out, but if you had mentioned that you meant 8086 compiler you would have been around 12 years out, and had you sepcifically said C (maybe you did actually but nevermind) you would have only been 3-5 years out as ones existed in the late 70's as Microsoft BASIC (the first real x86 program) became less useful.

[mathematician]

Unfortunately for the Compiler argument; you stated neither that the compiler was C or x86... Making you 30 years out, but if you had mentioned that you meant 8086 compiler you would have been around 12 years out, and had you sepcifically said C (maybe you did actually but nevermind) you would have only been 3-5 years out as ones existed in the late 70's as Microsoft BASIC (the first real x86 program) became less useful.

I'm not sure I understand any of that. As for x86 compilers being around in the late 70's, it must have been the very late seventies, because the first in the x86 family of chips wasn't released until 1979. Microsoft began life, it is true, as a small company turning out stand alone Basic interpreters for the early micro-computers, which at that time were still very much toys.

[Tyler]

Unfortunately for the Compiler argument; you stated neither that the compiler was C or x86... Making you 30 years out, but if you had mentioned that you meant 8086 compiler you would have been around 12 years out, and had you sepcifically said C (maybe you did actually but nevermind) you would have only been 3-5 years out as ones existed in the late 70's as Microsoft BASIC (the first real x86 program) became less useful.

I'm not sure I understand any of that. As for x86 compilers being around in the late 70's, it must have been the very late seventies, because the first in the x86 family of chips wasn't released until 1979. Microsoft began life, it is true, as a small company turning out stand alone Basic interpreters for the early micro-computers, which at that time were still very much toys.

Apologies when i say x86 there imeant 8080... though i do know those compilrs where moved almost immediately to the 8086/8088

[osdp.vnx]

I think writing OS from scratch for a novice is a bit uphill task but neverthless not impossible. Just a strong will power high motivation and dedication...
Check list:
1. Good command over C and Assembly
2. X386 Arc understanding
3. Bios and boot process understanding


Lots of dedication....
well gud luck to u

[bontanu]

Yes it is possible at least for me

250.000+ lines of assembly in less that 6 months (project time in IDE and I have left it idle for some days or night when I have fell asleep or went to shopping all day)

However the project did spread over many years since I could only write a few ASM lines every other weekend or so... some years I never had the time. The longest full work period was approximatively 2 months.

So yes it is possible even in full ASM but very unlikely if you are not very very experienced ...

For the average experienced coder it is much wiser to consider a few years at least.