[25% OT] C library integral part of OS/kernel???

[Keith Thompson]

Keith Thompson wrote: [...]

Now do you understand?

I think I do. BTW, I never meant to be rude with you or anybody else. I
earnestly apologize if I did end up.

Apology accepted.

 

[Jonathan Mcdougall]

Would the lord allow me to nitpick on the last sentence?

Speaking on His/Her behalf, I give you permission
to do so.


Dan Pop: Please make sure you include something
like "sometimes" or "may or may not" when you talk
about things like inlining function calls.


Jonathan

 

[Minti]

Speaking on His/Her behalf, I give you permission
to do so.

The lord seemeth to be busy or maybe just a bit arrogant .

<snip>

 

[Minti]

Oh, I think we've all understood quite clearly:

Indeed. I appreciate that.

Says it all, really.

Yup it does. Though just replace "top-post for top-post for top-post"
with "bullet-for-bullet."


Anwyay I have apologized twice for this so I don't see your point.
Moreso the flame DID NOT start with Mark commenting me on top-posting.
He just mentioned that one should not not assume that the reply is
directly aimed at one's own post. It WAS I who mentioned the toppost.
And if you look ( did you? ) at that post I KNOWINGLY added a signature
with "Damn Top Posters". Why do you think I wrote that?

So who is posting without thought now?

Because I respect a lot of guys here, I have already said that I won't
be top-posting even if someone top-posts me. Why is that such a simple
reply is made a issue here? Why can't you just resist the tempation to
post?

HTH

 

[CBFalconer]

.... snip a lot of childish nonsense ...

Because I respect a lot of guys here, I have already said that I
won't be top-posting even if someone top-posts me. Why is that
such a simple reply is made a issue here? Why can't you just
resist the tempation to post?

Why are you insisting on having the last word? Go and sin no more.

 

[Keith Thompson]

Because I respect a lot of guys here, I have already said that I won't
be top-posting even if someone top-posts me. Why is that such a simple
reply is made a issue here? Why can't you just resist the tempation to
post?

Usenet is an asychronous medium. If multiple people post a followup,
it may be because they haven't seen each other's posts.

You've apologized. I advise you to drop it and move on, unless
there's actually something new to say (which seems unlikely).

 

[Minti]

... snip a lot of childish nonsense ...

Yeah well sort of but just when I start to avoid these posts someone
just pops up ;-).

Why are you insisting on having the last word?

How can I NOT? I have got to reply to these polite questions? ;-).
Don't tell me you have also joined the forces of evil. (-:

 

[Keith Thompson]

CBFalconer wrote: [...]

Why are you insisting on having the last word?

How can I NOT?

By not posting. Really.

 

[gtippery]

The very important point that favors C being used in Operating Systems
is that no other compiler-for-language-other-than-c has been able to
been able to beat code-generated-by-c-compilers.

"Beat" it how? Convenience, size, speed, ubiquity, or what? There are
certainly cases where another compiler will generate a smaller and/or
faster executable than any of the C compilers available on that
platform, or for that architecture. Highly parallel machines are a
likely case, but there are others. Some architectures are rather
C-hostile, having for instance no user stack or no indirect addressing.


I believe I've heard that the system programming language on CDC
mainframes was (at least at one time) FORTRAN, to the extent that no
assembler was provided. (Or perhaps it was ALGOL. Not C, in any
case.)

It is very much
possible to write an Operating System within a programming language
like Java, with a pinch of salt actually, and then generating the
native code for a platform like i386, again with a pinch of salt. There
is at least one Operating System Project that attempts to do so.

What's this "salt" you refer to? And when you say "a programming
language like Java", what Java-like characteristic are you referring
to? A bytecoded, interpreted language? An OO language? A declarative
language? A strongly-checked language? And what's special about
i386's?

OS's have been written in many languages, but you usually require at
least some assembly language unless there's already at least some sort
of bootstrap on the platform, and usually some low-level device
drivers.


Not specifically to Minti:

Of course, nobody's even referenced an explicit definition of
"Operating System" in this entire thread, as far as I can see. (There
were references to the kernal plus other parts.) Some of the
disagreement appears to be due to different, unstated definitions.

 

[Minti]

"Beat" it how? Convenience, size, speed, ubiquity, or what? There are
certainly cases where another compiler will generate a smaller and/or
faster executable than any of the C compilers available on that
platform, or for that architecture. Highly parallel machines are a
likely case, but there are others.

Yup, but how many Operating System developers seem to be convinced with
this argument?

Some architectures are rather
C-hostile, having for instance no user stack or no indirect

addressing.

How many of these are in _common_ and _popular_ usage? (*)

I believe I've heard that the system programming language on CDC
mainframes was (at least at one time) FORTRAN, to the extent that no
assembler was provided. (Or perhaps it was ALGOL. Not C, in any
case.)

Read (*).

What's this "salt" you refer to?

Have you read about projects like Jalapeno? Though Jalapeno[%] is a
Virtual Machine implemented (totally) in Java, the developers had to
use several hacks like storing references of several different objects
in a single array.

[%] http://jikesrvm.sourceforge.net/

And when you say "a programming
language like Java", what Java-like characteristic are you referring
to? A bytecoded, interpreted language? An OO language? A declarative
language? A strongly-checked language? And what's special about
i386's?

Is Java a declarative language? And what's so special about i386's? ask
yourself.

OS's have been written in many languages, but you usually require at
least some assembly language unless there's already at least some sort
of bootstrap on the platform, and usually some low-level device
drivers.

In as sense YES! But only if it were just _writing_ an Operating
System. _Usability_ _Efficiency_ _Speed_ are just some of the arguments
that need to be satisfied.

Followups set to alt.os.development

 

[Richard Bos]

Check. Some news server transport is playing up, and it's something at
ntli.net.

Richard