Darren New wrote:
> Warp wrote:
>>   Being a JavaScript interpreter, it doesn't have too much choice than to
>> support FP, does it?-)
> 
> If the FP code failed silently, and most javascript didn't use FP, then
> most javascript wouldn't fail. :-)

AFAIK, Javascript has no integers. An interpreter may use integers for
numbers when it can, and automatically switch to floats if they are needed;
but that'd be just an optimization.

Post a reply to this message

Warp wrote:
> clipka <nomail@nomail> wrote:
>> But maybe what you mean is trigonometric functions. These, indeed, may be
>> not so commonplace.
> 
>   Thunderbird seemed to at least use logarithms...

Wow, any idea what for?

Post a reply to this message

In povray.off-topic Nicolas Alvarez <nic### [at] gmailcom> wrote:
> Warp wrote:
> > clipka <nomail@nomail> wrote:
> >> But maybe what you mean is trigonometric functions. These, indeed, may be
> >> not so commonplace.
> > 
> >   Thunderbird seemed to at least use logarithms...

> Wow, any idea what for?

  I didn't study it closely, and I already deleted the sources...

-- 
                                                          - Warp

Post a reply to this message

Darren New <dne### [at] sanrrcom> wrote:
>  From WikiPedia:

> """
> An operating system (commonly abbreviated OS and O/S) is the infrastructure 
> software component of a computer system; it is responsible for the 
> management and coordination of activities and the sharing of the limited 
> resources of the computer. The operating system acts as a host for 
> applications that are run on the machine. As a host, one of the purposes of 
> an operating system is to handle the details of the operation of the 
> hardware. This relieves application programs from having to manage these 
> details and makes it easier to write applications.
> """

> Sure sounds like MS-DOS to me.

  MS-DOS sounds like "to handle the details of the operation of the
hardware. This relieves application programs from having to manage these
details and makes it easier to write applications." ?

  Haha.

> Unless you don't count memory, serial and parallel ports, keyboard, disk 
> space, and network connections as "limited resources".

  Yeah, MS-DOS had great support for managing memory, network connections
and the like. In your dreams maybe.

  MS-DOS was nothing but an application launcher, which kept some routines
in memory for the application to call if it wanted. After the application
launched, it had absolute control of the machine. Basically the application
became the de-facto "operating system", if we can call it that.

  Calling MS-DOS an operating system is akin to calling grub an operating
system.

-- 
                                                          - Warp

Post a reply to this message

(Redirected from a discussion on povray.beta-test)

Warp wrote:
>   MS-DOS sounds like "to handle the details of the operation of the
> hardware. This relieves application programs from having to manage these
> details and makes it easier to write applications." ?

ANSI.SYS
COM1:
LPT:
Floppy disks
Hard disks
ASCII keyboard input
HIMEM.SYS
TSR programs
NetBIOS

I could open files on floppies, hard drives, and CD-ROMs all with the same 
API under MS-DOS. In what was isn't that handling the details of the 
operation of that hardware?

>   Yeah, MS-DOS had great support for managing memory, network connections
> and the like. In your dreams maybe.

I didn't say it had great support. It had support: TSR programs for managing 
memory, NetBIOS for managing network connections.

>   MS-DOS was nothing but an application launcher, which kept some routines
> in memory for the application to call if it wanted. After the application
> launched, it had absolute control of the machine.

So? That doesn't mean it didn't manage the hardware and share resources. It 
just means there was no protection to keep you from bypassing the OS. No, 
sorry, a machine without an OS is one where you dial in the codes on the 
front panel switches to load your application off paper tape and then branch 
to it.

Did *you* ever write an MS-DOS program that had to avoid accidentally 
stomping on other people's files on the same hard drive?

Do you think that the applications in your cell phone don't have complete 
control over the hardware? Do you think your cell phone doesn't have an OS 
in it?

Do you think the AmgiaOS wasn't actually an OS?

Other than scoffing, what do you think, specifically, an OS has to do that 
MS-DOS doesn't do at least in a primitive way?

> Basically the application
> became the de-facto "operating system", if we can call it that.

You don't understand what an "operating system" is, I fear. Or, rather, your 
definition is too narrow. There are all kinds of operating systems out there 
that don't prevent you from crashing the device if you want.

-- 
   Darren New, San Diego CA, USA (PST)
   The NFL should go international. I'd pay to
   see the Detroit Lions vs the Roman Catholics.

Post a reply to this message

Darren New <dne### [at] sanrrcom> wrote:
> I could open files on floppies, hard drives, and CD-ROMs all with the same 
> API under MS-DOS. In what was isn't that handling the details of the 
> operation of that hardware?

  Grub can do that too. Heck, grub can even load a RAM disk image from a
disk (which uses basically any file system known to man), load it in memory
and boot from it. Grub also supports keyboards and mice.

  I suppose that makes grub an OS as well.

> >   MS-DOS was nothing but an application launcher, which kept some routines
> > in memory for the application to call if it wanted. After the application
> > launched, it had absolute control of the machine.

> So? That doesn't mean it didn't manage the hardware and share resources.

  Then grub is an OS, and bios is an OS.

  Ever been to the bios setup screen? Lots of support for hardware there.
And bios can launch programs too.

  Too bad the "os" part of "bios" does not stand for "operating system",
but it could well do.

> It 
> just means there was no protection to keep you from bypassing the OS. No, 
> sorry, a machine without an OS is one where you dial in the codes on the 
> front panel switches to load your application off paper tape and then branch 
> to it.

  That's BS, and you know it.

> Did *you* ever write an MS-DOS program that had to avoid accidentally 
> stomping on other people's files on the same hard drive?

  I have written a MS-DOS application which accidentally made the HD
inaccessible because of a bug. It almost caused me tons of trouble too.

> Do you think that the applications in your cell phone don't have complete 
> control over the hardware?

  Actually I do. The apps in my phone are extremely restricted on what they
can do.

> Do you think your cell phone doesn't have an OS in it?

  It does, which is precisely why those apps don't have free access to
everything.

> Do you think the AmgiaOS wasn't actually an OS?

  I don't know the specifics of AmigaOS to tell.

> Other than scoffing, what do you think, specifically, an OS has to do that 
> MS-DOS doesn't do at least in a primitive way?

  Maybe I could answer with a question: If grub is not an OS and MS-DOS is,
then where exactly is the line?

  Your definition of "operating system" seems to be "if the author decided
to call it an operating system, it is one".

-- 
                                                          - Warp

Post a reply to this message

Warp wrote:
>   I suppose that makes grub an OS as well.

I suppose if you consider GRUB an OS for loading Linux, that might be the 
case. The fact that GRUB is no longer around once it has loaded the program 
means it's probably not actually an operating system. You can't take one 
thing an OS does and say "my program does that too, hence I'm an OS." Basic 
logic, ya know.  Not all men are Socrates.

>> So? That doesn't mean it didn't manage the hardware and share resources.
>   Then grub is an OS, and bios is an OS.

Grub doesn't manage hardware for an application, nor does it share resources 
for an application. GRUB isn't even around in memory (AFAIK) by the time 
Linux is finished booting, so I can't imagine how you think it's managing 
resources for its application.

>   Ever been to the bios setup screen? Lots of support for hardware there.
> And bios can launch programs too.

The BIOS can certainly be seen as a form of OS. Even more primitive than 
DOS. But it certainly shares some of the same attributes.

>> It 
>> just means there was no protection to keep you from bypassing the OS. No, 
>> sorry, a machine without an OS is one where you dial in the codes on the 
>> front panel switches to load your application off paper tape and then branch 
>> to it.
> 
>   That's BS, and you know it.

I don't follow. An OS that isn't there once the application program starts 
isn't an OS. I'm not sure why you think it's BS that a machine without an OS 
needs all the device drivers loaded into each application. Maybe you've just 
never worked on a machine without an OS.

In my experience, when someone simply answers "That's BS and you know it", 
it often means "Good point, but shut up about it."

>> Did *you* ever write an MS-DOS program that had to avoid accidentally 
>> stomping on other people's files on the same hard drive?
> 
>   I have written a MS-DOS application which accidentally made the HD
> inaccessible because of a bug. It almost caused me tons of trouble too.

Yes. So? Again, just because the OS doesn't protect you from accessing the 
hardware directly doesn't mean it isn't an OS. If that's what you're 
disagreeing with, then say so, and we can end this discussion.

>> Do you think that the applications in your cell phone don't have complete 
>> control over the hardware?
> 
>   Actually I do. The apps in my phone are extremely restricted on what they
> can do.

The ones written outside the phone, yes. I'm talking about the apps that 
come with the phone. You know, like the address book and such? Response to 
voice recognition? The bit that lets you pick which image from the camera to 
send via MMS?

>> Do you think your cell phone doesn't have an OS in it?
> 
>   It does, which is precisely why those apps don't have free access to
> everything.

Sorry. *your* apps might not have free access. Certainly most of the 
supplied apps (including the interpreter) has access.

>> Do you think the AmgiaOS wasn't actually an OS?
>   I don't know the specifics of AmigaOS to tell.

It ran on a 68000, which means no memory protection and no protected mode 
opcodes. Does that make it "not an OS"?

How about a B5000, which also ran on a machine with no hardware memory 
protection and no hardware protected mode, but which nevertheless was a 
multiuser system with restrictions to keep people from stomping on other 
peoples memory? (Not unlike your phone apps, most likely.)

How about BREW, the cell-phone 3rd-party app library that runs raw C code in 
the phone, but for which there's an administrative human review to make sure 
your code is safe to run? Is it an OS if the protections are enforced by 
human beings?

>> Other than scoffing, what do you think, specifically, an OS has to do that 
>> MS-DOS doesn't do at least in a primitive way?
> 
>   Maybe I could answer with a question: If grub is not an OS and MS-DOS is,
> then where exactly is the line?

MS-DOS is still around after your application finishes running. It's 
providing services to multiple (sometimes even concurrent) applications. It 
manages resource allocations between different applications.

I'd still like you to answer the question I'm asking. Because it sounds like 
you're saying that any OS where you can bypass the OS and talk straight to 
hardware isn't an OS, and I suspect you'll find that the expression 
"operating system" was coined before machines had protected modes.

>   Your definition of "operating system" seems to be "if the author decided
> to call it an operating system, it is one".

Firstly, no, it's wikipedia's definition. Secondly, it seems pretty clear to 
me that if the "OS" isn't around and providing services while the 
application is running, then it isn't an OS.

-- 
   Darren New, San Diego CA, USA (PST)
   The NFL should go international. I'd pay to
   see the Detroit Lions vs the Roman Catholics.

Post a reply to this message

Darren New wrote:
> Yes. So? Again, just because the OS doesn't protect you from accessing 
> the hardware directly doesn't mean it isn't an OS. If that's what you're 
> disagreeing with, then say so, and we can end this discussion.

BTW, Microsoft Xenix also ran on machines with no hardware protection, as 
did early Macintoshes. If you want to say neither Xenix nor MacOS were 
actual operating systems because they didn't keep you from crashing the 
machine, then I'll be forced to conclude that C++ isn't an actual 
programming language. :-)

-- 
   Darren New, San Diego CA, USA (PST)
   The NFL should go international. I'd pay to
   see the Detroit Lions vs the Roman Catholics.

Post a reply to this message

Darren New <dne### [at] sanrrcom> wrote:
> The fact that GRUB is no longer around once it has loaded the program 
> means it's probably not actually an operating system.

  Sound pretty much like MS-DOS to me.

  If your definition of OS is a bunch of utility routines in memory a
program can call for diverse tasks (eg. accessing a disk), then I suppose
MS-DOS, is an OS, bios is an OS, and basically anything is an OS.

  Bios is, in fact, a perfect example of "operating system" by this
definition.

> In my experience, when someone simply answers "That's BS and you know it", 
> it often means "Good point, but shut up about it."

  Now you are insulting me.

> >> Do you think that the applications in your cell phone don't have complete 
> >> control over the hardware?
> > 
> >   Actually I do. The apps in my phone are extremely restricted on what they
> > can do.

> The ones written outside the phone, yes. I'm talking about the apps that 
> come with the phone. You know, like the address book and such? Response to 
> voice recognition? The bit that lets you pick which image from the camera to 
> send via MMS?

  Funny that you don't even know which phone I have.

  Since you are talking in such an expert tone of voice about cellphone
applications, I assume you have written programs for Symbian and know
perfectly what you are talking about?

> >> Other than scoffing, what do you think, specifically, an OS has to do that 
> >> MS-DOS doesn't do at least in a primitive way?
> > 
> >   Maybe I could answer with a question: If grub is not an OS and MS-DOS is,
> > then where exactly is the line?

> MS-DOS is still around after your application finishes running. It's 
> providing services to multiple (sometimes even concurrent) applications. It 
> manages resource allocations between different applications.

  Then bios is also an OS. I suppose that settles it.

-- 
                                                          - Warp

Post a reply to this message

Warp <war### [at] tagpovrayorg> wrote:
>   MS-DOS sounds like "to handle the details of the operation of the
> hardware. This relieves application programs from having to manage these
> details and makes it easier to write applications." ?
>
>   Haha.
>   MS-DOS was nothing but an application launcher, which kept some routines
> in memory for the application to call if it wanted. After the application
> launched, it had absolute control of the machine. Basically the application
> became the de-facto "operating system", if we can call it that.
>
>   Calling MS-DOS an operating system is akin to calling grub an operating
> system.

Indeed.

Post a reply to this message

Warp wrote:
> Darren New <dne### [at] sanrrcom> wrote:
>> The fact that GRUB is no longer around once it has loaded the program 
>> means it's probably not actually an operating system.
> 
>   Sound pretty much like MS-DOS to me.

Err, no. MS-DOS is still around, supporting your application, once your 
application is running. That's why main() can call exit() and you get a 
prompt back.

>   Bios is, in fact, a perfect example of "operating system" by this
> definition.

Early operating systems were very primitive.

>> In my experience, when someone simply answers "That's BS and you know it", 
>> it often means "Good point, but shut up about it."
> 
>   Now you are insulting me.

Then stop accusing me of attempting to BS you. :-) If you're going to take 
that statement personally, then so will I.

Seriously, if you'd answer the question I asked, it's pretty easy to have a 
conversation. So far, you've done everything you can do to argue with me 
that MS-DOS *isn't* an operating system without actually telling me *why* 
you think that. *That* sounds like BS to me.

Given that the manufacturer and Wikipedia say that Xenix, MS-DOS, and MCP 
are operating systems, I would think it's incumbent upon you if you disagree 
to at least say why you disagree instead of saying they're all BSing.

>   Funny that you don't even know which phone I have.

I'm guessing it probably has a camera and a contact list on it.

>   Since you are talking in such an expert tone of voice about cellphone
> applications, I assume you have written programs for Symbian and know
> perfectly what you are talking about?

Yes. Not in cell phones, mind, but in credit card terminals.

In addition, my wife programs cell phones, so I have a pretty good idea of 
how the code is organized inside at least some of them, and how it's 
distributed to the OEMs, and so on.

>>>> Other than scoffing, what do you think, specifically, an OS has to do that 
>>>> MS-DOS doesn't do at least in a primitive way?
>>>   Maybe I could answer with a question: If grub is not an OS and MS-DOS is,
>>> then where exactly is the line?
> 
>> MS-DOS is still around after your application finishes running. It's 
>> providing services to multiple (sometimes even concurrent) applications. It 
>> manages resource allocations between different applications.
> 
>   Then bios is also an OS. I suppose that settles it.

I'll notice that you've still managed to avoid answering the question. I 
guess those batch processing monitors on the old mainframes weren't OSes 
either, then.

But, since you'd prefer to argue on an insulting and emotionally-driven 
level rather than actually discuss something that might lead to actually 
giving someone something to think about (or heaven forbid changing someone's 
opinion), I'll let it drop with a sad shake of the head.

-- 
   Darren New, San Diego CA, USA (PST)
   Why is there a chainsaw in DOOM?
   There aren't any trees on Mars.

Post a reply to this message

Darren New <dne### [at] sanrrcom> wrote:
> >   Then bios is also an OS. I suppose that settles it.

> I'll notice that you've still managed to avoid answering the question. I 
> guess those batch processing monitors on the old mainframes weren't OSes 
> either, then.

> But, since you'd prefer to argue on an insulting and emotionally-driven 
> level rather than actually discuss something that might lead to actually 
> giving someone something to think about (or heaven forbid changing someone's 
> opinion), I'll let it drop with a sad shake of the head.

  As I said, that settles it. You made your point of view clear and there's
no misunderstanding. I don't agree with it, but continuing arguing about it
would be useless.

-- 
                                                          - Warp

Post a reply to this message

Warp wrote:
>   As I said, that settles it.

Actually, thinking on it, the BIOS isn't an operating system because it 
doesn't actually manage any resources for the applications. It provides 
device drivers somewhat more abstract than the hardware, but it doesn't 
manage it in a way that makes it possible to share amongst independent 
applications.

> I don't agree with it, but continuing arguing about it
> would be useless.

Probably. It might be nice to know what part you don't agree with, tho. 
Rather than just saying "No it isn't!"

-- 
   Darren New, San Diego CA, USA (PST)
   Why is there a chainsaw in DOOM?
   There aren't any trees on Mars.

Post a reply to this message

Darren New <dne### [at] sanrrcom> wrote:
> > I don't agree with it, but continuing arguing about it
> > would be useless.

> Probably. It might be nice to know what part you don't agree with, tho. 
> Rather than just saying "No it isn't!"

  I remember we having this same conversation years ago. I remember
discussing it in more detail. We probably didn't agree back then either.

-- 
                                                          - Warp

Post a reply to this message

Warp wrote:
>   I remember we having this same conversation years ago. 

Nice dodge!  With the number of statements you've made disclaiming any need 
to actually answer the question, you could have answered the question. That 
said, I'm going to keep going, because I feel like it. You don't have to 
believe that anything in this post is addressing you or anything you've 
said. Feel free to also get annoyed if you think I'm lying about the 
immediately preceding sentence also.



There's a logical falacy wherein you use argument ad absurdum, where one 
says "If X, then Y, and Y is clearly not acceptable, so not X." Except for 
that to work, one actually needs that logical implication there. It doesn't 
work to say "If it were raining, then everyone would want to be dead, which 
is clearly untrue, so it must not be raining."  This flaw is usually pretty 
obvious in religious arguments, as well, along the lines of "If there were 
no God, then nobody would have a reason to be good or even want to live, so 
obviously there must be a God." It does of course work in some places, like 
the proof that every lossless compressor must increase the size of some 
inputs. It also seems to work in some cases but fails due to implicit 
assumptions, like Aristotle's proof that the universe is infinite, when what 
he actually proved was that the universe is unbounded.



Getting back to operating systems, after thinking about it a bit (because 
what else am I going to do while stairmastering?), I think the determinant 
is whether the "OS" manages resources between different applications (either 
concurrently or consecutively).

If there's only one possible application, and it runs without maintaining 
any state between runs, then whatever it's on top of isn't an OS, but just a 
library (or BIOS, which is a special case). Disregarding the built-in wall 
clock, my microwave oven is back to the same state it shipped from the 
factory every time it finishes cooking something. There's nothing carrying 
over from "run" to "run", so it has one application, and that application 
never interacts with anything else, including later executions of itself.

In this sense, neither GRUB nor a BIOS is an operating system. Every time 
you run GRUB, it starts from scratch without regard for what happened 
before, and it interacts with nothing except its own data (the OS it's 
loading). The BIOS also doesn't share resources between applications - it 
runs one application at a time (such as Windows or MS-DOS or Linux) and it 
gives that OS complete control over the hardware without providing any 
mechanisms for (for example) reserving space on a floppy for one boot and 
not the other.

The partition tables on a HD are the very first step of turning a boot 
loader into a "operating system". However, again, there's nothing "managed". 
The BIOS doesn't use the partition table, nor does it create or enforce it. 
The boot record stored on the hard disk uses the partition table, and the 
operating systems by convention respect it, but those aren't part of the BIOS.

I also think it doesn't make sense to talk about an operating system 
"managing resources for an application" if it's not around while the 
application is running. Hence, things like GRUB, which are overwritten as 
part of the boot process and are gone until you reset the machine, don't 
count as operating systems.

If the "OS" is linked into every application, which is loaded into memory 
and then overwritten when the next application is run, that may or may not 
be an "operating system" depending on what it does, IMO. That is, of course, 
more open to interpretation. I don't think FORTH counts as an operating 
system, as there's no "application" outside the FORTH system that manages 
the sharing of any resources - instead, there are conventions where one 
manually notes which blocks of disk space belong to which files, for 
example. This is about as close to an "operating system" as you can get 
without actually being one, I think.

MS-DOS (and Classic Mac, Amiga OS, MS Xenix, CP/M, TRS-DOS, etc) would all 
be operating systems because they manage the disk space for applications (if 
nothing else). Two independent applications can have the space they used 
managed without interference, even if they can't run at the same time.

Given that people are calling things "operating systems" in 1956 in the IBM 
704, long before a computer was even capable of running two programs at 
once, I think it's pretty obvious that most of the world (including those 
who invented the term) think that "operating system" doesn't require 
concurrent execution, preemptive scheduling, or support of multiple users. 
Otherwise, we wouldn't have the more specialized terms "batch processing 
OS", "multi-user OS", "timeshare OS", or "preemptive multitasking OS".

-- 
   Darren New, San Diego CA, USA (PST)
   Why is there a chainsaw in DOOM?
   There aren't any trees on Mars.

Post a reply to this message

Darren New <dne### [at] sanrrcom> wrote:
> Nice dodge!

  Dude, you got even more dogmatic than me about this. And I'm not talking
about the "definition of OS" thing. I suppose I got a bit under your skin,
even though I didn't really do it purposefully.

  Basically what I wanted to say is that I'm not fully agreeing with you,
but I'm tired of going again through all this conversation. If you want to
see it as irrational and stubborn behavior, then by all means. Maybe I am.
If you want to think that I ran out of rational arguments and that's why
I'm "dodging" your questions, then by all means. Maybe I am doing exactly
that. If you think I'm doing this because I know I'm wrong but too proud
to admit it, then so be it.

  I just got tired of continuing this conversation, that's all.

  This definition of "operating system" is not mine. It's what I was
taught. Did they teach me wrongly? Maybe. It didn't sound irrational
to me, though. I'm just not up to write an essay here about the subject.
I don't have the motivation nor all the facts memorized.

-- 
                                                          - Warp

Post a reply to this message

Warp wrote:
>   Basically what I wanted to say is that I'm not fully agreeing with you,
> but I'm tired of going again through all this conversation.

That's fine. I was just curious. I wasn't trying to change your mind about 
anything. I was just wondering what part of the definition you were 
contesting. :-)

I know there are a bunch of people who don't count something like MS-DOS as 
a "real" operating system. I was just curious what feature it was missing 
that you thought was necessary.

>   I just got tired of continuing this conversation, that's all.

I have no problem with that. As I said, feel free not to follow up. I just 
felt like rambling. Every once in a while, I try to structure a bunch of 
thoughts and write them down, just to keep in practice.

>   This definition of "operating system" is not mine. It's what I was
> taught. Did they teach me wrongly? Maybe. It didn't sound irrational
> to me, though. I'm just not up to write an essay here about the subject.
> I don't have the motivation nor all the facts memorized.

Again, that's fine. I was just looking at the definition on Wiki, and 
stretching my brain. As I said, feel free not to follow up. I'm not arguing 
with you. I'm just expressing opinions.  Sometimes I get in the mood to talk 
about something nobody else is interested in also. :-)

-- 
   Darren New, San Diego CA, USA (PST)
   Why is there a chainsaw in DOOM?
   There aren't any trees on Mars.

Post a reply to this message

Darren New wrote:
> MS-DOS is still around after your application finishes running. It's 
> providing services to multiple (sometimes even concurrent) applications. 
> It manages resource allocations between different applications.
> 
> I'd still like you to answer the question I'm asking. Because it sounds 
> like you're saying that any OS where you can bypass the OS and talk 
> straight to hardware isn't an OS, and I suspect you'll find that the 
> expression "operating system" was coined before machines had protected 
> modes.

Yeah. His definition would make Apple DOS and Prodos "not operating 
systems", because it was possible, in both, to talk directly to Apple 
hardware, and even take full control of the reads and writes from the 
disk (like in copy programs, which used there "own" code to do that, 
since it let them control timing and sector numbers, etc.)

For that matter, it wouldn't even be a valid argument to claim that an 
OS needs to provide resources to "multiple" apps, which is a possible 
argument, since while Microsoft abandoned further work on DOS when they 
put out Windows, some companies, like one called Concurrent Systems, 
made multiuser DOS implementations that supported normal DOS 
applications, but used task switching methods to run them for multiple 
users.

-- 
void main () {
   If Schrödingers_cat is alive or version > 98 {
     if version = "Vista" {
       call slow_by_half();
       call DRM_everything();
     }
     call functional_code();
   }
   else
     call crash_windows();
}

<A HREF='http://www.daz3d.com/index.php?refid=16130551'>Get 3D Models, 
3D Content, and 3D Software at DAZ3D!</A>

Post a reply to this message

Warp wrote:
>   This definition of "operating system" is not mine. It's what I was
> taught. Did they teach me wrongly? Maybe.

I think.. Bios "could be", but just having some way to get into it, so 
you can change some things, but not truly "use" any of your hardware, 
makes it "not one". Its not like there is a clear dividing line, except 
that, if you want to make one, an OS would have to have these attributes:

1. It allows user interaction on some level.
2. Ability to remain in memory and run a new application, once the 
current one terminates (bios can't do that, short of restarting the 
entire machine).
3. It sits between applications and a lower level interface.
4. Its not a script engine.

#4 is there *solely* on the basis that while DOS is an OS, something 
like DOSBOX isn't. Why? Because the later *pretends* to have access to a 
lower level interface, while *pretending* to provide all the same 
features of the OS it simulates. So, its one of the fuzzy ones. It is, 
and it isn't. Now, if you ran something that simulated the hardware, 
then you ran DOS *on top* of that simulated hardware, then it would be. 
Same with script engines. What you are dealing with is not access to the 
low level interfaces of the machine itself, but to a simulation of those 
interfaces that are needed to allow the applications to run in it.

Point being. Its gotten damn blurry. Now you can have a physical 
machine, running a simple OS, which runs an application that simulates a 
second machine, running a complex OS, which is running a sim of a much 
simpler machine, which is running a very simple OS, which is running 
DOSBOX, which simulates "both" the machine *and* the OS. In theory, you 
could then run Windows 3.1 on that, and run some ancient copy of DOSBOX 
inside that, and... Its gets damn insane. Ten years ago, people could 
have given you a clear and concise, "Yep, that is, and that isn't." 
Now... Who the hell knows in some cases. lol

-- 
void main () {
   If Schrödingers_cat is alive or version > 98 {
     if version = "Vista" {
       call slow_by_half();
       call DRM_everything();
     }
     call functional_code();
   }
   else
     call crash_windows();
}

<A HREF='http://www.daz3d.com/index.php?refid=16130551'>Get 3D Models, 
3D Content, and 3D Software at DAZ3D!</A>

Post a reply to this message

Patrick Elliott wrote:
> made multiuser DOS implementations that supported normal DOS 
> applications, but used task switching methods to run them for multiple 
> users.

I even programmed a version of CP/M that supported multiple Z-80 CPUs and 
would run one "session" on each CPU, back in the first half of the '80s or 
so. That was truly weird.

-- 
   Darren New, San Diego CA, USA (PST)
   Why is there a chainsaw in DOOM?
   There aren't any trees on Mars.

Post a reply to this message

Patrick Elliott <sel### [at] npgcablecom> wrote:
> Yeah. His definition would make Apple DOS and Prodos "not operating 
> systems", because it was possible, in both, to talk directly to Apple 
> hardware

  You don't even know what "my" definition is (which isn't mine, as I have
said several times). You just assume from what Darren seems to assume.

-- 
                                                          - Warp

Post a reply to this message

Warp <war### [at] tagpovrayorg> wrote:
> Patrick Elliott <sel### [at] npgcablecom> wrote:
> > Yeah. His definition would make Apple DOS and Prodos "not operating
> > systems", because it was possible, in both, to talk directly to Apple
> > hardware
>
>   You don't even know what "my" definition is (which isn't mine, as I have
> said several times). You just assume from what Darren seems to assume.

Your definition is the Tanenbaum one, right?

Post a reply to this message

nemesis <nam### [at] gmailcom> wrote:
> Your definition is the Tanenbaum one, right?

  I tried to google material about that, but I couldn't find anything
relevant. Do you know of any page about that subject? I would be interested
(if for nothing else, to be able to answer your question).

-- 
                                                          - Warp

Post a reply to this message

Patrick Elliott wrote:
> 
> Point being. Its gotten damn blurry. Now you can have a physical
> machine, running a simple OS, which runs an application that simulates a
> second machine, running a complex OS, which is running a sim of a much
> simpler machine, which is running a very simple OS, which is running
> DOSBOX, which simulates "both" the machine *and* the OS. In theory, you
> could then run Windows 3.1 on that, and run some ancient copy of DOSBOX
> inside that, and... Its gets damn insane. Ten years ago, people could
> have given you a clear and concise, "Yep, that is, and that isn't."
> Now... Who the hell knows in some cases. lol
> 

And to make it even more unclear, you can run bunch of these "piles"
simultaneously on that single physical machine. Just add virtualization
layer, such as Xen/HyperV/ESX :).

-Aero

Post a reply to this message

Warp <war### [at] tagpovrayorg> wrote:
> nemesis <nam### [at] gmailcom> wrote:
> > Your definition is the Tanenbaum one, right?
>
>   I tried to google material about that, but I couldn't find anything
> relevant. Do you know of any page about that subject? I would be interested
> (if for nothing else, to be able to answer your question).

Oh, please!  Andrew Tanenbaum!  You sure know the man!

http://en.wikipedia.org/wiki/Andrew_S._Tanenbaum

Author of MINIX, many influential books on OS design:

http://www.prenhall.com/tanenbaum/

(I've read at least 2)

and a memorable flamewar with Linus Torvalds in the beginning of Linux over bad
design choices:

http://groups.google.com/group/comp.os.minix/browse_thread/thread/c25870d7a41696d2/f447530d082cd95d?tvc=2

Post a reply to this message

Eero Ahonen wrote:
> And to make it even more unclear, you can run bunch of these "piles"
> simultaneously on that single physical machine. Just add virtualization
> layer, such as Xen/HyperV/ESX :).

That's kind of what I was thinking in terms of BIOS vs OS.  Xen is an OS, 
because it can run multiple applications, even tho those applications are 
OSes. A BIOS isn't an OS, even tho it runs an application that's an OS, 
because it isn't helping share resources in any way.

-- 
   Darren New, San Diego CA, USA (PST)
   Why is there a chainsaw in DOOM?
   There aren't any trees on Mars.

Post a reply to this message

nemesis wrote:
> and a memorable flamewar with Linus Torvalds in the beginning of Linux over bad
> design choices:

It's always fun to go back and read what influential people wrote a long 
time ago about the future of things that have lasted for a decade or two.

-- 
   Darren New, San Diego CA, USA (PST)
   Why is there a chainsaw in DOOM?
   There aren't any trees on Mars.

Post a reply to this message

Warp wrote:
>   You don't even know what "my" definition is (which isn't mine, as I have
> said several times). You just assume from what Darren seems to assume.

It would be lots easier to avoid assuming if you were less coy about what 
you meant. :-)  But I'm not assuming anything. I'm guessing and asking. And 
I thought you were tired of discussing it? ;-)

-- 
   Darren New, San Diego CA, USA (PST)
   Why is there a chainsaw in DOOM?
   There aren't any trees on Mars.

Post a reply to this message

Darren New wrote:
> 
> It's always fun to go back and read what influential people wrote a long
> time ago about the future of things that have lasted for a decade or two.
> 

Yep. The disgussion nemesis linked is very entertaining to read, thanks
of that for him :).

-Aero

Post a reply to this message

nemesis <nam### [at] gmailcom> wrote:
> [-- text/plain, encoding 8bit, charset: iso-8859-1, 25 lines --]

> Warp <war### [at] tagpovrayorg> wrote:
> > nemesis <nam### [at] gmailcom> wrote:
> > > Your definition is the Tanenbaum one, right?
> >
> >   I tried to google material about that, but I couldn't find anything
> > relevant. Do you know of any page about that subject? I would be interested
> > (if for nothing else, to be able to answer your question).

> Oh, please!  Andrew Tanenbaum!  You sure know the man!

  I didn't ask who Andrew Tanenbaum is.

-- 
                                                          - Warp

Post a reply to this message

Darren New <dne### [at] sanrrcom> wrote:
> nemesis wrote:
> > and a memorable flamewar with Linus Torvalds in the beginning of Linux over bad
> > design choices:
>
> It's always fun to go back and read what influential people wrote a long
> time ago about the future of things that have lasted for a decade or two.

Indeed, it's fun to watch arguments about obsolete techs flying around with so
much positive vibe. :)

BTW, according to Tanenbaum in that very thread, MS-DOS is definitely an OS.  So
I guess Tanenbaum is definitely not the basis for Warp's OS definition...

Post a reply to this message

Eero Ahonen <aer### [at] removethiszbxtnetinvalid> wrote:
> Darren New wrote:
> >
> > It's always fun to go back and read what influential people wrote a long
> > time ago about the future of things that have lasted for a decade or two.
> >
>
> Yep. The disgussion nemesis linked is very entertaining to read, thanks
> of that for him :).

heh, I thought you all knew about it. ;)

Post a reply to this message

nemesis wrote:
> BTW, according to Tanenbaum in that very thread, MS-DOS is definitely an OS.  So
> I guess Tanenbaum is definitely not the basis for Warp's OS definition...

Yeah, I was kind of curious what definition Warp was using, given that 
Merriam-Webster, Wikipedia, the people who coined the term "operating 
sytem", and the people who wrote MS-DOS, all seem to agree that MS-DOS would 
be an operating system.

Of course, we all know what it does and doesn't do, so exactly what it's 
called isn't very relevant, methinks.

-- 
   Darren New, San Diego CA, USA (PST)
   Why is there a chainsaw in DOOM?
   There aren't any trees on Mars.

Post a reply to this message

nemesis wrote:
> 
> heh, I thought you all knew about it. ;)
> 

We all aren't *that* old in our nerd-age :p.

-Aero

Post a reply to this message

Eero Ahonen <aer### [at] removethiszbxtnetinvalid> wrote:
> nemesis wrote:
> >
> > heh, I thought you all knew about it. ;)
> >
>
> We all aren't *that* old in our nerd-age :p.

By 1992, I didn't have a computer and was off playing a Super Nintendo. :)

But it's such a memorable piece of Linux lore that I heard about it many and
many times even when I first got into it by 2001...

Post a reply to this message

nemesis wrote:
> 
> By 1992, I didn't have a computer 

I^HMy parents actually had one, but no internet connection. Nor any
other connection - IIRC my mom bought a modem for us somewhere around
1995. Anyway, me being 13 at the time didn't make me too much intrested
in coding.

> and was off playing a Super Nintendo. :)

I bought my SNES in 2008. 15 euros, containing 2 games. Much cheaper
than in 1992.

OTOH I also bought my Wii in 2008. ~400 euros, containing 2 "real"
games, Wii Play, 2 controllers and a loading station for 2 controllers.

> But it's such a memorable piece of Linux lore that I heard about it many and
> many times even when I first got into it by 2001...
> 

It surely is. I can actually wonder how no-one has linked it for me
before...

-Aero

Post a reply to this message

On Sat, 03 Jan 2009 13:24:46 -0500, Warp wrote:

>  Yeah, MS-DOS had great support for managing memory, network connections
> and the like. In your dreams maybe.

INT2F redirection provided resources for redirection needed for network 
connections.  And it included memory management routines.  It didn't do 
memory protection well, but that's more a sign of the times than anything.

>   MS-DOS was nothing but an application launcher, which kept some
>   routines
> in memory for the application to call if it wanted. After the
> application launched, it had absolute control of the machine. Basically
> the application became the de-facto "operating system", if we can call
> it that.

Nonsense.  The kernel itself included a large amount of functionality in 
the ISRs used for accessing all aspects of the machine.  It wasn't a 
preemptive multitasking operating system, but the ability to multitask 
isn't something that defines an operating system - or didn't at the time.

One could argue that TSR programs actually gave the system the ability to 
multitask, but it was more like task switching than multitasking.  You 
could conceivably stretch (but it woudl be a stretch) to nonpreemptive 
multitasking, but I don't think I'd go that far.

The fact that I've got several books on DOS operating system programming, 
interrupts, and functionality would seem to counter your argument that 
DOS is just an "application launcher" or anything like GRUB.

Jim

Post a reply to this message

On Sat, 03 Jan 2009 15:36:06 -0500, Warp wrote:

>> The fact that GRUB is no longer around once it has loaded the program
>> means it's probably not actually an operating system.
> 
>   Sound pretty much like MS-DOS to me.

Then you don't understand how MS-DOS works, because it doesn't terminate 
from memory when you launch WordPerfect.  It stays resident.  GRUB 
doesn't stay resident.  At all.

Jim

Post a reply to this message

On Sun, 04 Jan 2009 10:10:29 -0800, Darren New wrote:

> Xen is an OS

I don't know that I'd agree with that.  A hypervisor is a hardware 
abstraction layer, messing about with the various memory rings in the 
system in order to make the OS running within the hypervisor believe it's 
running in Ring 0 when it's not in order to take advantage of memory 
protection features in the hardware without tripping the native hardware 
up when a child domain bombs out.

Jim

Post a reply to this message

Jim Henderson wrote:
> On Sun, 04 Jan 2009 10:10:29 -0800, Darren New wrote:
> 
>> Xen is an OS
> 
> I don't know that I'd agree with that.

Is VM/CMS an OS? I bet IBM would say so. :-)

> A hypervisor is a hardware 
> abstraction layer, messing about with the various memory rings in the 
> system in order to make the OS running within the hypervisor believe it's 
> running in Ring 0 when it's not in order to take advantage of memory 
> protection features in the hardware without tripping the native hardware 
> up when a child domain bombs out.

Sure. And Linux is a hardware abstraction layer, making your application 
think it has a linear RAM address space to mess with and providing linear 
sequences of bytes on disk to read and write (organized as a single tree of 
names), when none of those are actually true. One of the jobs of an OS is 
necessarily abstracting the hardware in *some* way, if you accept that 
maintaining state and managing resources between independent applications is 
part of the definition of an OS.

Of course you could argue either way.  I was more contrasting Xen against a 
BIOS than arguing that Xen must necessarily be considered an OS.

-- 
   Darren New, San Diego CA, USA (PST)
   Why is there a chainsaw in DOOM?
   There aren't any trees on Mars.

Post a reply to this message

On Mon, 05 Jan 2009 14:43:09 -0800, Darren New wrote:

> Jim Henderson wrote:
>> On Sun, 04 Jan 2009 10:10:29 -0800, Darren New wrote:
>> 
>>> Xen is an OS
>> 
>> I don't know that I'd agree with that.
> 
> Is VM/CMS an OS? I bet IBM would say so. :-)

It's also probably a bit more involved than XEN is.  I suppose I could 
ask my local XEN expert, was just recalling from a presentation he gave a 
few weeks ago at a conference I attended about how XEN actually works.  
It's pretty fascinating, OS or not.

> 
>> A hypervisor is a hardware
>> abstraction layer, messing about with the various memory rings in the
>> system in order to make the OS running within the hypervisor believe
>> it's running in Ring 0 when it's not in order to take advantage of
>> memory protection features in the hardware without tripping the native
>> hardware up when a child domain bombs out.
> 
> Sure. And Linux is a hardware abstraction layer, making your application
> think it has a linear RAM address space to mess with and providing
> linear sequences of bytes on disk to read and write (organized as a
> single tree of names), when none of those are actually true. One of the
> jobs of an OS is necessarily abstracting the hardware in *some* way, if
> you accept that maintaining state and managing resources between
> independent applications is part of the definition of an OS.
> 
> Of course you could argue either way.  I was more contrasting Xen
> against a BIOS than arguing that Xen must necessarily be considered an
> OS.

True.  I just wanted to make the point. ;-)

Jim

Post a reply to this message

On Mon, 05 Jan 2009 17:23:17 -0500, Jim Henderson wrote:

> INT2F redirection provided resources for redirection needed for network
> connections.  And it included memory management routines

Just to clarify, "it" in the second sentence here means "MS-DOS" not 
"INT2F", though I think some of the DOS extender functions probably were 
accessed through 2F.

Jim

Post a reply to this message

nemesis <nam### [at] gmailcom> wrote:
> Your definition is the Tanenbaum one, right?

  I found this:

http://www.cs.rpi.edu/~hollingd/opsys/notes/Chapter1/Chapter1.pdf

  I'm not exactly sure what it's supposed to be, but looks like some kind
of presentation about Tanenbaum's book.

  Especially page 5 seems interesting.

-- 
                                                          - Warp

Post a reply to this message

Jim Henderson wrote:
> It's also probably a bit more involved than XEN is. 

I'm not sure. I used it only vaguely. Note that the "VM" part means "virtual 
machine", and "CMS" was the single-user OS that ran under it without change, 
so I think the differences between that and XEN are probably pretty monor.

-- 
   Darren New, San Diego CA, USA (PST)
   Why is there a chainsaw in DOOM?
   There aren't any trees on Mars.

Post a reply to this message

On Mon, 05 Jan 2009 15:54:14 -0800, Darren New wrote:

> Jim Henderson wrote:
>> It's also probably a bit more involved than XEN is.
> 
> I'm not sure. I used it only vaguely. Note that the "VM" part means
> "virtual machine", and "CMS" was the single-user OS that ran under it
> without change, so I think the differences between that and XEN are
> probably pretty monor.

Maybe, I'd have to talk to someone who knows VM/CMS better.  Used to be a 
guy I worked with who thought of VMware in those terms.

Jim

Post a reply to this message

Warp wrote:
>   Especially page 5 seems interesting.

Considering it's named "modern operating systems", and I wouldn't exactly 
count MS-DOS as a "modern operating system," I'm not surprised :-)  The 
contention of this paper (on page 5 at least) seems to be that no machine 
running unsafe code without hardware protection can possibly be running an 
operating system, nor can a machine like the Palm Pilot be running an OS, 
since only one program runs at a time. I find this counter-intuitive, given 
the existence of terms like "single-user operating system", "batch operating 
system", and so on.

So, if the *compiler* enforces the protection (like Singularity or the 
B5000) and not the hardware, does the code that regulates the sharing of 
resources and runs threads outside the address space of any particular 
applications count as an operating system? :-)

Page 22 seems to also contradict page 5, listing "handheld operating 
systems" and "smart card operating systems", for example, neither of which 
would seem to be allowing multiple programs to run at the same time, manage 
or protect memory, or multiple resources in space.

I just think claiming that stuff like MS-DOS or WinCE isn't an operating 
system, or that AppleDOS wasn't an OS, is going to confuse whoever one tries 
to talk to. Again, it's like claiming C isn't a real programming language 
because there are undefined programs that will compile and run.

-- 
   Darren New, San Diego CA, USA (PST)
   Why is there a chainsaw in DOOM?
   There aren't any trees on Mars.

Post a reply to this message

On Tue, 06 Jan 2009 01:07:31 +0100, Darren New <dne### [at] sanrrcom> wrote:
> Warp wrote:
>>   Especially page 5 seems interesting.
>
> Considering it's named "modern operating systems", and I wouldn't  
> exactly count MS-DOS as a "modern operating system," I'm not surprised  
> :-)

Note that in the actual book Tanenbaum explicitly calls MS-DOS an  
"operating system" and in fact dedicates an entire chapter to it. "Modern"  
in this context basically means the 1980s.



-- 
FE

Post a reply to this message

Darren New <dne### [at] sanrrcom> wrote:
> Warp wrote:
> >   Especially page 5 seems interesting.

> Considering it's named "modern operating systems", and I wouldn't exactly 
> count MS-DOS as a "modern operating system," I'm not surprised :-)

  Would you call Unix a "modern operating system"? It was originally developed
in 1969, and as wikipedia puts it, "Unix was designed to be portable,
multi-tasking and multi-user in a time-sharing configuration."

  But if you like, we could settle with: MS-DOS is not an operating system
by modern standards.

-- 
                                                          - Warp

Post a reply to this message

Warp wrote:
>   Would you call Unix a "modern operating system"?

Only in the sense that there haven't been a whole lot of fundamental 
improvements in OS concepts since then.  At least, not that have caught on. 
It's possible to be both old and modern: see plumbing, for example.

>   But if you like, we could settle with: MS-DOS is not an operating system
> by modern standards.

I'm happy with "MS-DOS is a very limited OS that I wouldn't want to use for 
general purpose computing these days." :-) I don't think we actually have to 
settle into agreement here. The discussion was interesting, tho.

-- 
   Darren New, San Diego CA, USA (PST)
   Why is there a chainsaw in DOOM?
   There aren't any trees on Mars.

Post a reply to this message