🚀 go-pugleaf

> Somebody asked her about computers and she told us all that she uses a
> Mac because the OS is based on Linux (yes, Linux)

A variant of BSD Unix:
http://www.apple.com/macosx/features/unix/

(Lots of people probably don't know or care about the different innards
and origins of the various *ux operating systems, even if, directly or
via some additional user interface, they use one...  She *is* correct
about the broad general idea, in an as-seen-from-space sort of way.)

> while Windows machines still have 16-bit MS-DOS at the core.

AFAIK their last DOS based operating system was Windows ME.   As of
Windows XP/2000,  there was some ability to run DOS in emulation so
them what needs it can  see if it works, but no more actual DOS in
there.

--Joe, owner of a fairly stable Windows ME machine, believe it or not...

Le Trôle wrote:
> No.
> No.
> Yes.
> No.
> Yes.

Not that I was asking for debunkage or verification since this is a
folkolore group, but OS X has BSD as its guts, not Linux. Any ability
to run Linux applications is due to API compatibility. Such a thing
exists in the Unix/Linux world.

RFM

"Fritz M" <nospam@masoner.net> wrote in message
news:1131311440.963767.217460@o13g2000cwo.googlegroups.com...
> I met a journalist on the airport shuttle bus today. She's a great
> storyteller and everybody on the bus seemed to be entertained by her
> tales. She's a stringer for outdoor and travel type publications. This
> journalist vectored a number of wonderful whoppers.
>
> 1. Drug dogs are all addicted to drugs.

No.

> 2. A big mass of condoms is floating in the middle of the Pacific
> Ocean.

No.

> 3. Some large percentage of $100 bills have traces of cocaine.

Yes.

> 4. A janitor unplugged life support machines and killed several
> patients.

No.

> [Treading on BoR]
> Somebody asked her about computers and she told us all that she uses a
> Mac because the OS is based on Linux (yes, Linux),

Yes.

> while Windows
> machines still have 16-bit MS-DOS at the core.

No.

On 2005-11-06, Le Trôle <letrole@hotmail.com> wrote:
> "Fritz M" <nospam@masoner.net> wrote in message
> news:1131311440.963767.217460@o13g2000cwo.googlegroups.com...
>>
>> [Treading on BoR]
>> Somebody asked her about computers and she told us all that she uses a
>> Mac because the OS is based on Linux (yes, Linux),
>
> Yes.

No.

>> while Windows machines still have 16-bit MS-DOS at the core.
>
> No.

Of what vintage? It was true as late as Windows ME [despite attempts by
MS to hide this]

"Le Tr?le" <letrole@hotmail.com> wrote:
> "Fritz M" <nospam@masoner.net> wrote in message
> news:1131311440.963767.217460@o13g2000cwo.googlegroups.com...
>> I met a journalist on the airport shuttle bus today. She's a great
>> storyteller and everybody on the bus seemed to be entertained by her
>> tales. She's a stringer for outdoor and travel type publications. This
>> journalist vectored a number of wonderful whoppers.
>>
>> [Treading on BoR]
>> Somebody asked her about computers and she told us all that she uses a
>> Mac because the OS is based on Linux (yes, Linux),

> Yes.

Actually it's based on BSD.


--
David Griffith
dgriffi@cs.csbuak.edu  <-- Switch the 'b' and 'u'

"Jordan Abel" <jmabel@purdue.edu> wrote in message
news:slrndmsu66.1dr9.jmabel@random.yi.org...
> On 2005-11-06, Le Trôle <letrole@hotmail.com> wrote:
> > "Fritz M" <nospam@masoner.net> wrote in message
> > news:1131311440.963767.217460@o13g2000cwo.googlegroups.com...
> >>
> >> [Treading on BoR]
> >> Somebody asked her about computers and she told us all that she uses a
> >> Mac because the OS is based on Linux (yes, Linux),
> >
> > Yes.
>
> No.

It's FreeBSD with a Linux abstraction layer. The code is there,
Linux applications run without recompiling, but Apple doesn't
support it officially. (Fear of GPL virus that plagues Linux)

> >> while Windows machines still have 16-bit MS-DOS at the core.
> >
> > No.
>
> Of what vintage?

2000

> It was true as late as Windows ME [despite attempts by MS to hide this]

Asked and Answered.

On 2005-11-06, Le Trôle <letrole@hotmail.com> wrote:
> "Jordan Abel" <jmabel@purdue.edu> wrote in message
> news:slrndmsu66.1dr9.jmabel@random.yi.org...
>> On 2005-11-06, Le Trôle <letrole@hotmail.com> wrote:
>> > "Fritz M" <nospam@masoner.net> wrote in message
>> > news:1131311440.963767.217460@o13g2000cwo.googlegroups.com...
>> >>
>> >> [Treading on BoR]
>> >> Somebody asked her about computers and she told us all that she uses a
>> >> Mac because the OS is based on Linux (yes, Linux),
>> >
>> > Yes.
>>
>> No.
>
> It's FreeBSD with a Linux abstraction layer. The code is there,
> Linux applications run without recompiling, but Apple doesn't
> support it officially. (Fear of GPL virus that plagues Linux)

That hardly makes it "based on linux".

"Jordan Abel" <jmabel@purdue.edu> wrote in message
news:slrndmt5io.2du1.jmabel@random.yi.org...
> On 2005-11-06, Le Trôle <letrole@hotmail.com> wrote:
> > "Jordan Abel" <jmabel@purdue.edu> wrote in message
> > news:slrndmsu66.1dr9.jmabel@random.yi.org...
> >> On 2005-11-06, Le Trôle <letrole@hotmail.com> wrote:
> >> > "Fritz M" <nospam@masoner.net> wrote in message
> >> > news:1131311440.963767.217460@o13g2000cwo.googlegroups.com...
> >> >>
> >> >> [Treading on BoR]
> >> >> Somebody asked her about computers and she told us all that she uses
a
> >> >> Mac because the OS is based on Linux (yes, Linux),
> >> >
> >> > Yes.
> >>
> >> No.
> >
> > It's FreeBSD with a Linux abstraction layer. The code is there,
> > Linux applications run without recompiling, but Apple doesn't
> > support it officially. (Fear of GPL virus that plagues Linux)
>
> That hardly makes it "based on linux".

OS X was built with this compatibility in mind. The abstraction layer
for Linux was intended as a way of extending the applications that
were available for use on a Mac. But, the GPL issue put a stop to it.

An abstraction layer is of little use if the application and the os don't
have equivalent functions. FreeBSD and Linux are similar, but not
identical. Changes were made to OS X to take into account specific
calls that could be made by Linux, but without a FreeBSD equivalent.

That's why Linux code, but not just any other Unix such as SCO,
can run without recompiling specifically for OS X.

On 2005-11-07, Le Trôle <letrole@hotmail.com> wrote:
> "Jordan Abel" <jmabel@purdue.edu> wrote in message
> news:slrndmt5io.2du1.jmabel@random.yi.org...
>> On 2005-11-06, Le Trôle <letrole@hotmail.com> wrote:
>> > "Jordan Abel" <jmabel@purdue.edu> wrote in message
>> > news:slrndmsu66.1dr9.jmabel@random.yi.org...
>> >> On 2005-11-06, Le Trôle <letrole@hotmail.com> wrote:
>> >> > "Fritz M" <nospam@masoner.net> wrote in message
>> >> > news:1131311440.963767.217460@o13g2000cwo.googlegroups.com...
>> >> >>
>> >> >> [Treading on BoR]
>> >> >> Somebody asked her about computers and she told us all that she uses
> a
>> >> >> Mac because the OS is based on Linux (yes, Linux),
>> >> >
>> >> > Yes.
>> >>
>> >> No.
>> >
>> > It's FreeBSD with a Linux abstraction layer. The code is there,
>> > Linux applications run without recompiling, but Apple doesn't
>> > support it officially. (Fear of GPL virus that plagues Linux)
>>
>> That hardly makes it "based on linux".
>
> OS X was built with this compatibility in mind. The abstraction layer
> for Linux was intended as a way of extending the applications that
> were available for use on a Mac. But, the GPL issue put a stop to it.
>
> An abstraction layer is of little use if the application and the os don't
> have equivalent functions. FreeBSD and Linux are similar, but not
> identical. Changes were made to OS X to take into account specific
> calls that could be made by Linux, but without a FreeBSD equivalent.
>
> That's why Linux code, but not just any other Unix such as SCO,
> can run without recompiling specifically for OS X.

And is it likely that "based on linux" means that, when it's not even a
reasonable interpretation for "based on linux", or is it more likely
that the person who thought it was "based on linux" was just confusing
freebsd with linux, or thouht that all unix was linux, or whatever?

"Le Tr?le" <letrole@hotmail.com> wrote:

:OS X was built with this compatibility in mind. The abstraction layer
:for Linux was intended as a way of extending the applications that
:were available for use on a Mac. But, the GPL issue put a stop to it.

:An abstraction layer is of little use if the application and the os don't
:have equivalent functions. FreeBSD and Linux are similar, but not
:identical. Changes were made to OS X to take into account specific
:calls that could be made by Linux, but without a FreeBSD equivalent.

:That's why Linux code, but not just any other Unix such as SCO,
:can run without recompiling specifically for OS X.

Even if this were strictly true (HINT: it's not), how would this be
based on Linux?

Why don't you admit you're one of the people who thinks "Open source
UNIX-like means Linux", despite it not being so?  It's okay.  Plenty of
otherwise intelligent people think this.  Most of them even manage to
tie their shoes in the morning.


David

"Jordan Abel" <jmabel@purdue.edu> wrote in message
news:slrndmt81o.2du1.jmabel@random.yi.org...
> On 2005-11-07, Le Trôle <letrole@hotmail.com> wrote:
> > "Jordan Abel" <jmabel@purdue.edu> wrote in message
> > news:slrndmt5io.2du1.jmabel@random.yi.org...
> >> On 2005-11-06, Le Trôle <letrole@hotmail.com> wrote:
> >> > "Jordan Abel" <jmabel@purdue.edu> wrote in message
> >> > news:slrndmsu66.1dr9.jmabel@random.yi.org...
> >> >> On 2005-11-06, Le Trôle <letrole@hotmail.com> wrote:
> >> >> > "Fritz M" <nospam@masoner.net> wrote in message
> >> >> > news:1131311440.963767.217460@o13g2000cwo.googlegroups.com...
> >> >> >>
> >> >> >> [Treading on BoR]
> >> >> >> Somebody asked her about computers and she told us all that she
uses
> > a
> >> >> >> Mac because the OS is based on Linux (yes, Linux),
> >> >> >
> >> >> > Yes.
> >> >>
> >> >> No.
> >> >
> >> > It's FreeBSD with a Linux abstraction layer. The code is there,
> >> > Linux applications run without recompiling, but Apple doesn't
> >> > support it officially. (Fear of GPL virus that plagues Linux)
> >>
> >> That hardly makes it "based on linux".
> >
> > OS X was built with this compatibility in mind. The abstraction layer
> > for Linux was intended as a way of extending the applications that
> > were available for use on a Mac. But, the GPL issue put a stop to it.
> >
> > An abstraction layer is of little use if the application and the os
don't
> > have equivalent functions. FreeBSD and Linux are similar, but not
> > identical. Changes were made to OS X to take into account specific
> > calls that could be made by Linux, but without a FreeBSD equivalent.
> >
> > That's why Linux code, but not just any other Unix such as SCO,
> > can run without recompiling specifically for OS X.
>
> And is it likely that "based on linux" means that, when it's not even a
> reasonable interpretation for "based on linux", or is it more likely
> that the person who thought it was "based on linux" was just confusing
> freebsd with linux, or thouht that all unix was linux, or whatever?

It depends. If you're a regular user, you would see the derivation of NT
from 95,
whereas NT is actually closer to VMS, except for a few familiar
applications.
If the journalist was using applications that came from Linux, in a
roundabout
way, she was seeing the basis of OS X on Linux, the same way that the
shell of NT is based on 95. Journalists use some weird software for layout
and so forth, so she may be familiar with some bespoke Linux stuff.

"David Scheidt" <dscheidt@panix.com> wrote in message
news:dkm7t1$oht$1@reader2.panix.com...
> "Le Tr?le" <letrole@hotmail.com> wrote:
>
> :OS X was built with this compatibility in mind. The abstraction layer
> :for Linux was intended as a way of extending the applications that
> :were available for use on a Mac. But, the GPL issue put a stop to it.
>
> :An abstraction layer is of little use if the application and the os don't
> :have equivalent functions. FreeBSD and Linux are similar, but not
> :identical. Changes were made to OS X to take into account specific
> :calls that could be made by Linux, but without a FreeBSD equivalent.
>
> :That's why Linux code, but not just any other Unix such as SCO,
> :can run without recompiling specifically for OS X.
>
> Even if this were strictly true (HINT: it's not), how would this be
> based on Linux?
>
> Why don't you admit you're one of the people who thinks "Open source
> UNIX-like means Linux", despite it not being so?  It's okay.  Plenty of
> otherwise intelligent people think this.  Most of them even manage to
> tie their shoes in the morning.

But that is not the case. I'm not referring to the guts of the os in
question,
simply the operational intent that had to be quashed after all the legal
issues
of the GPL came up around the time that it was released. OS X was intended
to work with Linux, because that's where the market was seen to be. And,
this would be the appeal for the journalist, since it's likely that she
would
see crossover. Just as Apple originally intended.

Shorter Version.
You make cookies. People drink coffee. People like to dunk cookies
in their coffee. You avoid fruity flavours in your cookies and stick to
rich caramels and gooey chocolate. You base your end product on
the flavour of coffee, yet you have no actual coffee in you cookies.

Your cookies are based on coffee.

"Le Tr?le" <letrole@hotmail.com> wrote:
:"David Scheidt" <dscheidt@panix.com> wrote in message
:news:dkm7t1$oht$1@reader2.panix.com...
:> "Le Tr?le" <letrole@hotmail.com> wrote:
:>
:> :OS X was built with this compatibility in mind. The abstraction layer
:> :for Linux was intended as a way of extending the applications that
:> :were available for use on a Mac. But, the GPL issue put a stop to it.
:>
:> :An abstraction layer is of little use if the application and the os don't
:> :have equivalent functions. FreeBSD and Linux are similar, but not
:> :identical. Changes were made to OS X to take into account specific
:> :calls that could be made by Linux, but without a FreeBSD equivalent.
:>
:> :That's why Linux code, but not just any other Unix such as SCO,
:> :can run without recompiling specifically for OS X.
:>
:> Even if this were strictly true (HINT: it's not), how would this be
:> based on Linux?
:>
:> Why don't you admit you're one of the people who thinks "Open source
:> UNIX-like means Linux", despite it not being so?  It's okay.  Plenty of
:> otherwise intelligent people think this.  Most of them even manage to
:> tie their shoes in the morning.

:But that is not the case. I'm not referring to the guts of the os in
:question,
:simply the operational intent that had to be quashed after all the legal
:issues
:of the GPL came up around the time that it was released. OS X was intended

Legal issues around the GPL didn't "come up" around the time of the OS X
release.  They were well understood, before Apple even publicly
started talking about a 'NIX based OS, by people who care about such
things (like, say, the *BSD and NeXT people Apple hired to make OS X
happen).  Even the CEO would have been well aware of GNU issues,
having had to deal with them at his PPOE, a company that sold a
BSD-derived core, but with a user land dependent on stuff from GNU.
And, of course, FreeBSD has had a linux emulation layer for a long,
long time.


:Shorter Version.
:You make cookies. People drink coffee. People like to dunk cookies
:in their coffee. You avoid fruity flavours in your cookies and stick to
:rich caramels and gooey chocolate. You base your end product on
:the flavour of coffee, yet you have no actual coffee in you cookies.

:Your cookies are based on coffee.

By extension, Linus Tovalds sees this nifty OS called UNIX, and
decides to copy it.  No UNIX, but, clearly, based on it.

"David Scheidt" <dscheidt@panix.com> wrote in message
news:dkmdt1$j46$1@reader2.panix.com...
> "Le Tr?le" <letrole@hotmail.com> wrote:
> :"David Scheidt" <dscheidt@panix.com> wrote in message
> :news:dkm7t1$oht$1@reader2.panix.com...
> :> "Le Tr?le" <letrole@hotmail.com> wrote:
> :>
> :> :OS X was built with this compatibility in mind. The abstraction layer
> :> :for Linux was intended as a way of extending the applications that
> :> :were available for use on a Mac. But, the GPL issue put a stop to it.
> :>
> :> :An abstraction layer is of little use if the application and the os
don't
> :> :have equivalent functions. FreeBSD and Linux are similar, but not
> :> :identical. Changes were made to OS X to take into account specific
> :> :calls that could be made by Linux, but without a FreeBSD equivalent.
> :>
> :> :That's why Linux code, but not just any other Unix such as SCO,
> :> :can run without recompiling specifically for OS X.
> :>
> :> Even if this were strictly true (HINT: it's not), how would this be
> :> based on Linux?
> :>
> :> Why don't you admit you're one of the people who thinks "Open source
> :> UNIX-like means Linux", despite it not being so?  It's okay.  Plenty of
> :> otherwise intelligent people think this.  Most of them even manage to
> :> tie their shoes in the morning.
>
> :But that is not the case. I'm not referring to the guts of the os in
> :question,
> :simply the operational intent that had to be quashed after all the legal
> :issues
> :of the GPL came up around the time that it was released. OS X was
intended
>
> Legal issues around the GPL didn't "come up" around the time of the OS X
> release.  They were well understood,

but not considered to be a threat at the time.

> before Apple even publicly
> started talking about a 'NIX based OS, by people who care about such
> things (like, say, the *BSD and NeXT people Apple hired to make OS X
> happen).  Even the CEO would have been well aware of GNU issues,

Side note: We had to basically gut all GIF functionality in a finished £1.1m
project back in 2002 because Unisys threw a wobbly over LWZ patents,
and they were going after anyone big enough to make a difference but small
enough not to be able to fight back. We knew, we gambled, we lost.
GIFs? They're everywhere. WTF?

We bet on JPG and Forgent, and so far, we've won, because they're
jackasses with no real claim other than a tenuous connection.

Not that I place our legal team on par with Apple's, of course.

> having had to deal with them at his PPOE, a company that sold a
> BSD-derived core, but with a user land dependent on stuff from GNU.
> And, of course, FreeBSD has had a linux emulation layer for a long,
> long time.

That layer would have to be actively incorporated into OS X for a
specific reason, unless, of course, the people doing OS X were just
willy-nilly adding code for no other reason than the fact that it came
along as part of the agreement. Actually, it would possibly answer the
question I have raised if it could be said that they merely used the
existing BSD code, or if they did indeed do additional coding to
make it current with Linux as it would be seen in the wild.


> :Shorter Version.
> :You make cookies. People drink coffee. People like to dunk cookies
> :in their coffee. You avoid fruity flavours in your cookies and stick to
> :rich caramels and gooey chocolate. You base your end product on
> :the flavour of coffee, yet you have no actual coffee in you cookies.
>
> :Your cookies are based on coffee.
>
> By extension, Linus Tovalds sees this nifty OS called UNIX, and
> decides to copy it.  No UNIX, but, clearly, based on it.

I'd say that's a fair comment.

In <1131327144.171233.165570@o13g2000cwo.googlegroups.com>,
on 6 Nov 2005 17:32:24 -0800, Ad absurdum per aspera wrote:
>> Somebody asked her about computers and she told us all that she uses a
>> Mac because the OS is based on Linux (yes, Linux)
>
> A variant of BSD Unix:
> http://www.apple.com/macosx/features/unix/
>
> (Lots of people probably don't know or care about the different innards
> and origins of the various *ux operating systems, even if, directly or
> via some additional user interface, they use one...

  I tried to 'ssh -Y' from my father's MacOSX machine
to my FreeBSD machine.  Didn't happen.  Then tried
'ssh -X' which also didn't happen... 'ssh' worked
fine, for whatever that's worth... but that'd work
even from Redmondware (I imagine... I try never to
touch the stuff).

  I didn't have time to discover *why* the graphical
interface wouldn't work, but I was disappointed that
Mac didn't work like I'm used to my FreeBSD and Linux
machines working.
  With more time and a broadband connection I might
have tried VNC, but over dialup?  I don't think so.

  Yes, the MacOSX CLI is somewhat similar to 'nix, but
it's not like you can drop a 'nix user in front of
a Mac and have him pretend it's 'nix.  It just isn't.

> She *is* correct
> about the broad general idea, in an as-seen-from-space sort of way.)

  At least from Mars.  Certainly not from the moon.

Fritz M <nospam@masoner.net> wrote:
:Le Tr?le wrote:
:> No.
:> No.
:> Yes.
:> No.
:> Yes.

:Not that I was asking for debunkage or verification since this is a
:folkolore group, but OS X has BSD as its guts, not Linux. Any ability
:to run Linux applications is due to API compatibility. Such a thing
:exists in the Unix/Linux world.

Not to pick a nit, or point out the BOA, but it's ABI[1]
compatibility that makes applications able to run without
re-compilation.  API[2] compatibility is what makes it possible to
compile a program to run natively on the machine without changing the
source.

David

[1]  Application Binary Interface
[2]              Programmer

"Le Trôle" <letrole@hotmail.com> wrote:

>You make cookies. People drink coffee. People like to dunk cookies
>in their coffee. You avoid fruity flavours in your cookies and stick to
>rich caramels and gooey chocolate. You base your end product on
>the flavour of coffee, yet you have no actual coffee in you cookies.
>
>Your cookies are based on coffee.

That has to be one of the most amazing leaps of illogic I've ever
seen.

D.
--
Touch-twice life. Eat. Drink. Laugh.

-Resolved: To be more temperate in my postings.
Oct 5th, 2004 JDL

Le Trôle proclaimed:

> But that is not the case. I'm not referring to the guts of the os in
> question,
> simply the operational intent that had to be quashed after all the legal
> issues
> of the GPL came up around the time that it was released. OS X was intended
> to work with Linux, because that's where the market was seen to be. And,
> this would be the appeal for the journalist, since it's likely that she
> would
> see crossover. Just as Apple originally intended.

OK, you got a cite for this?

Or is this just more of your hand waving and egregiously uninformed
blather?


Apple wanted excellent networking client support, support that could
integrate with all of the big Unix based [not Linux] server operating
systems of the early 90's.  Plus they had already decided to go RISC,
helping form the PowerPC alliance with IBM and Motorola.  They also
wanted preemptive multitasking SMP, VM with protection, and networking
running on the kernel.  At this point if you still think Linux, you know
sod all about computers.

Apple considered Win/NT, Solaris, SVR4, BeOS, NeXT, etc. with Apple
acquiring NeXT with its Mach [free] kernel.   The Mach kernel became the
innards.

Yes, Apple did play with Linux on Mach, machts nichts.

The real roots were: OpenStep + FreeBSD4.4 tossed in with a dash of Blue
Box, FreeBSD, NetBSD, Mach 3 [more for the server side] and out popped
OS/X.

Don't recall seeing a lot of Linux compatibility testing.  Mostly AIX,
Solaris, HP-UX, etc.   It is a great Unix server style client, a class
which in my non-humble opinion will include Linux when they learn a bit
more about the wisdom of sometimes asking yourself why all those bright
people elsewhere may have done things just a tad differently.



> Your cookies are based on coffee.

   Your cookies are based on road apples.

David Scheidt proclaimed:

> "Le Tr?le" <letrole@hotmail.com> wrote:
> :"David Scheidt" <dscheidt@panix.com> wrote in message
> :news:dkm7t1$oht$1@reader2.panix.com...
> :> "Le Tr?le" <letrole@hotmail.com> wrote:
> :>
> :> :OS X was built with this compatibility in mind. The abstraction layer
> :> :for Linux was intended as a way of extending the applications that
> :> :were available for use on a Mac. But, the GPL issue put a stop to it.
> :>
> :> :An abstraction layer is of little use if the application and the os don't
> :> :have equivalent functions. FreeBSD and Linux are similar, but not
> :> :identical. Changes were made to OS X to take into account specific
> :> :calls that could be made by Linux, but without a FreeBSD equivalent.
> :>
> :> :That's why Linux code, but not just any other Unix such as SCO,
> :> :can run without recompiling specifically for OS X.
> :>
> :> Even if this were strictly true (HINT: it's not), how would this be
> :> based on Linux?
> :>
> :> Why don't you admit you're one of the people who thinks "Open source
> :> UNIX-like means Linux", despite it not being so?  It's okay.  Plenty of
> :> otherwise intelligent people think this.  Most of them even manage to
> :> tie their shoes in the morning.
>
> :But that is not the case. I'm not referring to the guts of the os in
> :question,
> :simply the operational intent that had to be quashed after all the legal
> :issues
> :of the GPL came up around the time that it was released. OS X was intended
>
> Legal issues around the GPL didn't "come up" around the time of the OS X
> release.  They were well understood, before Apple even publicly
> started talking about a 'NIX based OS, by people who care about such
> things (like, say, the *BSD and NeXT people Apple hired to make OS X
> happen).  Even the CEO would have been well aware of GNU issues,
> having had to deal with them at his PPOE, a company that sold a
> BSD-derived core, but with a user land dependent on stuff from GNU.
> And, of course, FreeBSD has had a linux emulation layer for a long,
> long time.

I don't recall GPL being an issue, the Mach kernel in OS/X was from CMU
which I thought was free, and unrestricted licensing.
>
>
> :Shorter Version.
> :You make cookies. People drink coffee. People like to dunk cookies
> :in their coffee. You avoid fruity flavours in your cookies and stick to
> :rich caramels and gooey chocolate. You base your end product on
> :the flavour of coffee, yet you have no actual coffee in you cookies.
>
> :Your cookies are based on coffee.
>
> By extension, Linus Tovalds sees this nifty OS called UNIX, and
> decides to copy it.  No UNIX, but, clearly, based on it.

Or a university sees this nice code AT&T has developed, and decides to
render a free version of it, originally for academic use, but then it
turns out the BSD folks knew more about networking and scaling than AT&T
did....

Derek Lyons proclaimed:

> "Le Trôle" <letrole@hotmail.com> wrote:
>
>
>>You make cookies. People drink coffee. People like to dunk cookies
>>in their coffee. You avoid fruity flavours in your cookies and stick to
>>rich caramels and gooey chocolate. You base your end product on
>>the flavour of coffee, yet you have no actual coffee in you cookies.
>>
>>Your cookies are based on coffee.
>
>
> That has to be one of the most amazing leaps of illogic I've ever
> seen.

   No, the Linux one was even more amazing...

Ad absurdum per aspera proclaimed:

>>Somebody asked her about computers and she told us all that she uses a
>>Mac because the OS is based on Linux (yes, Linux)
>
>
> A variant of BSD Unix:
> http://www.apple.com/macosx/features/unix/
>
> (Lots of people probably don't know or care about the different innards
> and origins of the various *ux operating systems, even if, directly or
> via some additional user interface, they use one...  She *is* correct
> about the broad general idea, in an as-seen-from-space sort of way.)
>
>
>>while Windows machines still have 16-bit MS-DOS at the core.
>
>
> AFAIK their last DOS based operating system was Windows ME.   As of
> Windows XP/2000,  there was some ability to run DOS in emulation so
> them what needs it can  see if it works, but no more actual DOS in
> there.
>
> --Joe, owner of a fairly stable Windows ME machine, believe it or not...
>
   I have one too.  The Windose Me box is laying flat on its face, out
   cold.  You don't get much more stable than that.

> Somebody asked her about computers and she told us all that she uses a
> Mac because the OS is based on Linux (yes, Linux), while Windows
> machines still have 16-bit MS-DOS at the core.

Since there seems to be a fair amount of misconception in this thread on at
least the latter of these two statements (several other posters have
correctly noted that OS X is based on BSD UNIX) I will throw in my .02.

Microsoft MS-DOS is a 16-bit real-mode operating system, and a pretty good
example of the breed. It was designed to run on an architecture with a
20-bit address bus, thus the 1 MB address space and source of the legendary
Bill Gates comment "640K should be enough for anyone." (I wish the Amiga
people had listened to that - they divided their 1 MB address space down the
middle and gave users only 512K). MS-DOS does not support multitasking,
multiprocessing, or virtual memory. However, as the hardware advanced a
number of various hacks and patches were applied to allow some limited
pseudo-multitasking and access to memory beyond 1 MB. Examples of the latter
is the well-known "DOS Extender" that came with DOOM and the EMM386 / HIMEM
/ MEMMAKER utilities that allowed one to move some DOS drivers above 1 MB.

Microsoft Windows versions 1 and 2 were little more than graphical shells on
a savagely sliced-and-diced MS-DOS kernel. Windows 3.0 did take some
advantage of the primitive multitasking and memory features of the Intel 286
and was a sort of 16 / 24 / 32 bit hybrid. However, Microsoft Windows 3.1
running in 386 enhanced mode is a 32-bit, multitasking, virtual memory
operating system - albeit one that uses lots of 16-bit chunks of DOS as
services. Typing "win" at the DOS prompt runs the Windows loader, which
unloads DOS and establishes 32-bit Windows as the OS. From Windows 3.1,
File -Exit unloads Windows and reloads DOS. Before the screaming begins, one
should note that several distros of Linux could play this same trick: you
can run their loader from the DOS prompt and start Linux, then "exit" Linux
back into DOS. No one in their right mind would claim that Linux is a DOS
program; neither is Windows 3.1.

The multitasking and virtual memory support in Win 3.1, while it was there
and did work after a fashion, was primitive at best. Windows 95 introduced
significant improvements and the Windows VMM system was changed quite a bit
by Win 95 OSR2 and then pretty well established by Windows 98 (thus the many
programs that require Win 98 or better). In all of these operating systems,
the old DOS prompt can be executed as a service under 32-bit Windows, and
you can run as many of them as you wish, concurrently, and each is tricked
by Windows into thinking it owns its own real-mode computer. Windows 9x also
uses a number of old 16-bit DOS functions for things like dealing with the
BIOS and system clock (mainly because there is little point in re-writing
them as 32-bit versions). In addition, Windows 95 and later DO still carry
some DOS architectural baggage (such as a requirement that process control
blocks are restricted to a particular area in physical memory, which leads
to occasional difficulties but makes some backwards-compatibility issues
easier to resolve) but to say that Windows 95 = DOS is nonsense. Indeed, the
differences between Windows 3.0 and Windows 3.1 (which appear to be the same
to the casual user) are MUCH more significant than are the differences
between Win 3.1 and Win 95 (which look quite a bit different but are quite
similar under the hood).

Microsoft's new-from-the-ground-up completely-independent-of-DOS 32-bit OS
with support for multitasking, multiprocessing, and virtual memory was
Windows NT. It was written in large part by former VAX folks appropriated
from DEC, and it's claimed to be coincidental that Windows NT (WNT) shares a
similar relationship with VMS as HAL does to IBM (one step behind rather
than one ahead, perhaps?). The first version released was called NT 3.0 so
that it would have the same version number as their existing current version
of Windows; NT 1 and NT 2 did not exist. Oddly enough, NT 3 was raked over
the coals for NOT having much backwards compatibility with existing DOS and
Windows programs, and NT 4 retreated a bit from the somewhat idealistic
architecture of NT 3 and threw in some backwards-compatibility features to
placate the masses. NT 4 was a smashing success (for one thing, it runs on
surprisingly limited hardware) and still very much around today (with the
latest service pack it looks just like 95/98). NT 5 was marketed as Windows
2000; NT 5.1 is better known as Windows XP. XP in particular is marketed as
both a home and office OS, and (like NT 4) sacrifices some desired
theoretical purity for better compatibility with older programs and for
general improvement of performance (achieved, for instance, by doing some
rather unclean things with the graphics subsystem). I am to this day
astonished that some of my old DOS games, with compile dates in the early
1980s, run just fine in an XP console. On the other hand, my 1987 Amiga 2000
wouldn't run most of the software that I had bought for my 1985 Amiga 1000.
However, I digress: Windows NT, of any version, is most definitely not DOS.

One can then say that we have three fairly distinct lines of OS products
from Microsoft:

16-bit line

MS-DOS 1.0 through 6.22, Windows 1 and 2, arguably big chunks of Windows 3.0

32-bit line #1, uses some 16-bit DOS components as services and erroneously
said to be "built on DOS"

Windows 3.1, 95, 95 OSR2, 98, 98 2nd, ME (best of breed is generally
regarded to be 98 2nd)

32-bit line #2, developed independently of DOS but got some DOS-like
functionality grafted in later for compatibility reasons

Windows NT 3, 4, 5 (aka 2000), 5.1 (aka XP)

Do you want to know more?
"Unauthorized Windows 95 Developers Resource Kit" by Andrew Schulman is a
good start; 500+ pages of detailed analysis of the inner workings of DOS,
Win 3, and Win 95 (including lots of multipage listings in C and assembler),
plus a CD of goodies for poking and prodding (and occasionally crashing)
your Windows system.

--
John Goulden

>>> Linux code, but not just any other Unix such as SCO,
>>> can run without recompiling specifically for OS X.

Example?  I've never seen a binary on the web that was advertised as
capable of running on both OS X or PPC Linux.


> OS X was intended to work with Linux,

IBM MVS "works with" Linux since you can FTP files between them.  It's
so vague as to be meaningless.


> because that's where the market was seen to be.

There was almost no market for Linux on Apple hardware when OS X was
developed - PPC Linux implementations were geeky crap that no ordinary
user would have the patience to get running (I'd been a Unix sysadmin
and I gave up).  There was nothing there for Apple to compete with, and
nothing worth co-existing with either.

==============  j-c  ======  @  ======  purr . demon . co . uk  ==============
Jack Campin:  11 Third St, Newtongrange EH22 4PU, Scotland | tel 0131 660 4760
<http://www.purr.demon.co.uk/jack/>   for CD-ROMs and free | fax 0870 0554 975
stuff: Scottish music, food intolerance, & Mac logic fonts | mob 07800 739 557

Nick Spalding proclaimed:

> Lon wrote, in <uu-dnX251ooYNfLeRVn-jw@comcast.com>
>  on Mon, 07 Nov 2005 11:18:26 -0800:
>
>
>>I don't recall GPL being an issue, the Mach kernel in OS/X was from CMU
>>which I thought was free, and unrestricted licensing.
>
>
> I hereby invoke the BOA.  What the hell are all these acronyms that are
> being bandied around in this thread?

Mach is not an acronym, although it is allegedly a mispronunciation of
the original, MUCK, about which I frankly don't recall.

CMU is Carnegie Mellon University.

OS/X to my knowledge is a trademark and not an acronym although you
could interpret it as Operating System Slash 10 if you insist on a
finely picked BOA...

Nick Spalding wrote:

> I hereby invoke the BOA.  What the hell are all these acronyms that are
> being bandied around in this thread?

We have more of a BOR problem than a BOA problem.  To borrow a simile,
it's like calling English Advanced Phoenician, because they use the
sam alphabet.
--
Burroughs "WinXP is based on the Difference Engine" Guy
Vaguer memories available upon request

It ran with motherboard video for a while, then got a cheap 3D video
card; both were alright.

Threw decent-quality RAM into it for a total of 384 MB, bought a
program called MemTurbo, and in good time went from a cheap hardwired
NIC to an 802.11b WiFi card.
I eplaced the hard drive and I did a clean install of XP while I was at
it.

Note that I said *fairly* stable.   I haven't kept statistics that
would withstand scientific scrutiny, but my impression is that has
never been worse than a good solid Windows 98 machine.    It is a good
bit flakier in all ways (more crashes, more failures to boot up or shut
down properly, more tries needed for more installer programs) than
we've gotten accustomed to with XP/2000, and far less stable than my OS
10.3 and 10.4 Macs.  But it is quite usable as a homework and
video-games engine and emergency backup for word processing.

My sysadmin friends from both the Unix and the Windows sides of the
game think I should sell tickets to this spectacle; all I can say is
that it works okay for me.

--Joe

Lon wrote, in <uu-dnX251ooYNfLeRVn-jw@comcast.com>
 on Mon, 07 Nov 2005 11:18:26 -0800:

> I don't recall GPL being an issue, the Mach kernel in OS/X was from CMU
> which I thought was free, and unrestricted licensing.

I hereby invoke the BOA.  What the hell are all these acronyms that are
being bandied around in this thread?
--
Nick Spalding

Nick Spalding <spalding@iol.ie> wrote:
:Lon wrote, in <uu-dnX251ooYNfLeRVn-jw@comcast.com>
: on Mon, 07 Nov 2005 11:18:26 -0800:

:> I don't recall GPL being an issue, the Mach kernel in OS/X was from CMU
:> which I thought was free, and unrestricted licensing.

:I hereby invoke the BOA.  What the hell are all these acronyms that are
:being bandied around in this thread?

GPL -- GNU[1] Public license, a "free" software license from GNU[1]
CMU -- Carnagie Mellon University
OS/x  -- the macintosh operating system, version ten
BSD  -- Berkeley Software mumble, a UNIX like operating system from
the CSRG[2] at UCB[3]

[1]  GNU --- GNU[1] is not UNIX
[2]  CSRG -- Computer Science Research Group
[3]  UCB  -- university of California, Berkeley.

On 2005-11-07, David Scheidt <dscheidt@panix.com> wrote:
> Nick Spalding <spalding@iol.ie> wrote:
>:Lon wrote, in <uu-dnX251ooYNfLeRVn-jw@comcast.com>
>: on Mon, 07 Nov 2005 11:18:26 -0800:
>
>:> I don't recall GPL being an issue, the Mach kernel in OS/X was from CMU
>:> which I thought was free, and unrestricted licensing.
>
>:I hereby invoke the BOA.  What the hell are all these acronyms that are
>:being bandied around in this thread?
>
> GPL -- GNU[1] Public license, a "free" software license from GNU[1]
General Public License
> CMU -- Carnagie Mellon University
> OS/x  -- the macintosh operating system, version ten
> BSD  -- Berkeley Software mumble, a UNIX like operating system from
> the CSRG[2] at UCB[3]
>
> [1]  GNU --- GNU[1] is not UNIX
> [2]  CSRG -- Computer Science Research Group
> [3]  UCB  -- university of California, Berkeley.

You forgot UNIX[4]

[4] UNIX -- By analogy with MULTICS[5]
[5] MULTICS -- Mutliplexed Information & Computing Services

In article <slrndmvid1.o5s.jmabel@random.yi.org>, Jordan Abel
<jmabel@purdue.edu> wrote:

>On 2005-11-07, David Scheidt <dscheidt@panix.com> wrote:
>> Nick Spalding <spalding@iol.ie> wrote:
>>:Lon wrote, in <uu-dnX251ooYNfLeRVn-jw@comcast.com>
>>: on Mon, 07 Nov 2005 11:18:26 -0800:
>>
>>:> I don't recall GPL being an issue, the Mach kernel in OS/X was from CMU
>>:> which I thought was free, and unrestricted licensing.
>>
>>:I hereby invoke the BOA.  What the hell are all these acronyms that are
>>:being bandied around in this thread?
>>
>> GPL -- GNU[1] Public license, a "free" software license from GNU[1]
>General Public License
>> CMU -- Carnagie Mellon University
>> OS/x  -- the macintosh operating system, version ten
>> BSD  -- Berkeley Software mumble, a UNIX like operating system from
>> the CSRG[2] at UCB[3]
>>
>> [1]  GNU --- GNU[1] is not UNIX
>> [2]  CSRG -- Computer Science Research Group
>> [3]  UCB  -- university of California, Berkeley.
>
>You forgot UNIX[4]
>
>[4] UNIX -- By analogy with MULTICS[5]
>[5] MULTICS -- Mutliplexed Information & Computing Services

All of this is fine, but what's BOA?

charles, TLA?

"Ad absurdum per aspera" <jtchew@california.com> wrote in
news:1131327144.171233.165570@o13g2000cwo.googlegroups.com:
X-Antivirus: avast! (VPS 0545-0, 11/07/2005), Outbound message
X-Antivirus-Status: Clean

> --Joe, owner of a fairly stable Windows ME machine, believe
> it or not...
>

I had one, but head geek Chet hated ME and put XP Pro on my
system one night.  He sweetened the deal by giving me a brand
new hard drive, double the old one in size, but I had gotten
used to being the almost only crash-free ME user in the world.

Tell me - did you install a non-standard, or new, video card in
your machine ?  I claimed that was why my ME machine never acted
out on me.  The new card confused it.

--
TeaLady (mari)

"The principal of Race is meant to embody and express the utter
negation of human freedom, the denial of equal rights, a
challenge in the face of mankind." A. Kolnai

In article <3KydncQ7e_rGTfLenZ2dnUVZ_tadnZ2d@comcast.com>,
 Lon <lon.stowell@comcast.net> wrote:

> Nick Spalding proclaimed:
>
> > Lon wrote, in <uu-dnX251ooYNfLeRVn-jw@comcast.com>
> >  on Mon, 07 Nov 2005 11:18:26 -0800:
> >
> >
> >>I don't recall GPL being an issue, the Mach kernel in OS/X was from CMU
> >>which I thought was free, and unrestricted licensing.
> >
> >
> > I hereby invoke the BOA.  What the hell are all these acronyms that are
> > being bandied around in this thread?
>
> Mach is not an acronym, although it is allegedly a mispronunciation of
> the original, MUCK, about which I frankly don't recall.
>
> CMU is Carnegie Mellon University.
>
> OS/X to my knowledge is a trademark and not an acronym although you
> could interpret it as Operating System Slash 10 if you insist on a
> finely picked BOA...

Apple calls it "Mac OS X", and OS does stand for Operating System.
X stands for ten, but is also meant to evoke the aura of Unix.

Ad absurdum per aspera proclaimed:
> It ran with motherboard video for a while, then got a cheap 3D video
> card; both were alright.

I've had more than one experience where a cheap or even free card would
work far better or more reliably than a much more expensive, but older
design, "name" brand.  Ethernet cards tend to be the worst offenders.

>
> Threw decent-quality RAM into it for a total of 384 MB, bought a
> program called MemTurbo, and in good time went from a cheap hardwired
> NIC to an 802.11b WiFi card.
> I eplaced the hard drive and I did a clean install of XP while I was at
> it.

With 512 Meg and up, WinMe could almost be characterized as
non-lethargic.  The Missus had a Me machine that couldn't even run
checkdisk on the original harddrive unless it was partitioned...which it
was not as shipped.

The XP was a good move.

> Note that I said *fairly* stable.   I haven't kept statistics that
> would withstand scientific scrutiny, but my impression is that has
> never been worse than a good solid Windows 98 machine.

Yeah, I guess if you compare a slug to an oxy moron, it might survive.

Win/98SE with full service packs wasn't bad.  Not in the class of
Win/NT, but better than even Win/95b with TCP/IP 1.2.    Ran 98SE well
past its useful life, until the notorious GDI overflow of the entire 9x
lot just ticked me off once too many times in multimedia editing.


> It is a good
> bit flakier in all ways (more crashes, more failures to boot up or shut
> down properly, more tries needed for more installer programs) than
> we've gotten accustomed to with XP/2000, and far less stable than my OS
> 10.3 and 10.4 Macs.  But it is quite usable as a homework and
> video-games engine and emergency backup for word processing.

I normally allow roughly 1 crash every few months before I boot the
operating system... with a Wolverine.   It must come from being from an
environment where uptimes of years were not at all unusual.
>
> My sysadmin friends from both the Unix and the Windows sides of the
> game think I should sell tickets to this spectacle; all I can say is
> that it works okay for me.

Yes, if you have a working Me box, you should sell tickets...    Even
Microsoft wouldn't serve that turkey at a free employee/microserf
thanksgiving lunch.

Derek Lyons proclaimed:

> "Ad absurdum per aspera" <jtchew@california.com> wrote:
>
>
>>Note that I said *fairly* stable.   I haven't kept statistics that
>>would withstand scientific scrutiny, but my impression is that has
>>never been worse than a good solid Windows 98 machine.    It is a good
>>bit flakier in all ways (more crashes, more failures to boot up or shut
>>down properly, more tries needed for more installer programs) than
>>we've gotten accustomed to with XP/2000, and far less stable than my OS
>>10.3 and 10.4 Macs.  But it is quite usable as a homework and
>>video-games engine and emergency backup for word processing.
>
>
> <nods>  My machine (which is a Win 98 machine) is about the same.  I
> can depend on it to lock up about twice a day - almost always in a
> game or reading Usenet, but never in Office.
>
>
>>My sysadmin friends from both the Unix and the Windows sides of the
>>game think I should sell tickets to this spectacle; all I can say is
>>that it works okay for me.
>
>
> There has arisen this feeling that PC's should boot up flawlessly out
> of the box and run with nary a problem, crash, or lockup until there
> is a power outage or you unplug the system prior to running it down to
> the Salvation Army to donate.
>
> Balderdash.
>
> Nuclear power plant controls, aircraft flight control computers, these
> and things like them should be held to such a standard - but to expect
> a commodity machine to perform so is ludicrous.  There is no need
> beyond delusion that such a performance level should be required.

I've had Win/NT boxes running some extremely heavy duty applications go
for months without a single hiccup.  The local power company typically
had more to do with reboots than any unhappiness.

As for cheap commodity boxen not being expected to be reliable,
balderbalderdash and bullshitbalderdash.   That type of thinking is why
there is no American Auto industry any more.

John D. Goulden wrote:

> (I wish the Amiga people had listened to that - they divided their
> 1 MB address space down the middle and gave users only 512K).

The Amiga did not have a 1 MB address space, nor was it divided in the
manner of MS-DOS or segmented as it was in the Intel 8088.

The Amiga OS was always 32-bit and multitasking.  The limit on the
original Amiga's memory capacity was the address bus of the Motorola
68000, which was 24-bit (16 MB).

- Ernie                                  http://home.comcast.net/~erniew

steve@SNIP-THIS.twoloonscoffee.com (Steve Ackman)  wrote on 07.11.05 in <slrndmtk37.2ka8.steve@wizard.dyndns.org>:

> In <1131327144.171233.165570@o13g2000cwo.googlegroups.com>,
> on 6 Nov 2005 17:32:24 -0800, Ad absurdum per aspera wrote:
> >> Somebody asked her about computers and she told us all that she uses a
> >> Mac because the OS is based on Linux (yes, Linux)
> >
> > A variant of BSD Unix:
> > http://www.apple.com/macosx/features/unix/
> >
> > (Lots of people probably don't know or care about the different innards
> > and origins of the various *ux operating systems, even if, directly or
> > via some additional user interface, they use one...
>
>   I tried to 'ssh -Y' from my father's MacOSX machine
> to my FreeBSD machine.  Didn't happen.  Then tried
> 'ssh -X' which also didn't happen... 'ssh' worked
> fine, for whatever that's worth... but that'd work
> even from Redmondware (I imagine... I try never to
> touch the stuff).
>
>   I didn't have time to discover *why* the graphical
> interface wouldn't work, but I was disappointed that
> Mac didn't work like I'm used to my FreeBSD and Linux
> machines working.

Because OS X doesn't use the X Windowing System, of course. No X server,
no ssh -X or -Y.

Well, you *can* get X servers for OS X (same as for Windows), and then I'd
expect ssh -X would work. I do know it works from Windows when I have an X
server running (though I much prefer doing it on Linux; for one, the
Cygwin X server isn't all that stable or fast).

>   Yes, the MacOSX CLI is somewhat similar to 'nix, but

The same way that women are somewhat similar to humans.

> it's not like you can drop a 'nix user in front of
> a Mac and have him pretend it's 'nix.  It just isn't.

You can. Don't confuse X with 'nix. You might as well claim something
isn't 'nix because it doesn't have Samba.

> > She *is* correct
> > about the broad general idea, in an as-seen-from-space sort of way.)
>
>   At least from Mars.  Certainly not from the moon.

Depends on how important you see the difference between Linux and *BSD.
They're certainly closer than, say, AIX.

To be more precise, MacOS X is what you get when you take the Mach (I
believe it's 3.0 these days) so-called microkernel, put a heavily hacked
*BSD on top, and have a userspace that is mostly FreeBSD (with selected
stuff from OpenBSD and maybe also NetBSD, and of course some Apple-
specific stuff as well) - this part is called Darwin - and on top of that,
you run a GUI based on Quartz (a Display PDF server) instead of X.

Kai
--
http://www.westfalen.de/private/khms/
"... by God I *KNOW* what this network is for, and you can't have it."
  - Russ Allbery (rra@stanford.edu)

lon.stowell@comcast.net (Lon)  wrote on 07.11.05 in <pKidnYFfeYnFOvLeRVn-uw@comcast.com>:

> Apple considered Win/NT, Solaris, SVR4, BeOS, NeXT, etc. with Apple
> acquiring NeXT with its Mach [free] kernel.   The Mach kernel became the
> innards.

Actually, NeXT already had a BSD kernel and userland, just not a modern
one, which I imagine was rather more important in deciding on BSD than any
other argument.

> Don't recall seeing a lot of Linux compatibility testing.  Mostly AIX,
> Solaris, HP-UX, etc.   It is a great Unix server style client, a class
> which in my non-humble opinion will include Linux when they learn a bit
> more about the wisdom of sometimes asking yourself why all those bright
> people elsewhere may have done things just a tad differently.

Well, some of these bright people seem to have asked themselves since why
they didn't do it like Linux. For one, quite a number of things about
Linux seem seriously faster (not NFS, though), and SMP scalability seems
excellent these days as well.

Kai
--
http://www.westfalen.de/private/khms/
"... by God I *KNOW* what this network is for, and you can't have it."
  - Russ Allbery (rra@stanford.edu)

letrole@hotmail.com (Le Trôle)  wrote on 06.11.05 in <dkm47d$ngk$1@nwrdmz02.dmz.ncs.ea.ibs-infra.bt.com>:

> "Jordan Abel" <jmabel@purdue.edu> wrote in message
> news:slrndmsu66.1dr9.jmabel@random.yi.org...
> > On 2005-11-06, Le Trôle <letrole@hotmail.com> wrote:
> > > "Fritz M" <nospam@masoner.net> wrote in message
> > > news:1131311440.963767.217460@o13g2000cwo.googlegroups.com...
> > >>
> > >> [Treading on BoR]
> > >> Somebody asked her about computers and she told us all that she uses a
> > >> Mac because the OS is based on Linux (yes, Linux),
> > >
> > > Yes.
> >
> > No.
>
> It's FreeBSD with a Linux abstraction layer. The code is there,
> Linux applications run without recompiling, but Apple doesn't
> support it officially. (Fear of GPL virus that plagues Linux)

If Apple was afraid of GPL, they wouldn't use so much GPL'd software.
Their whole development environment is based on the GNU tools, for
example, and the default shell on OS X these days is bash. Oh, and it
seems Apple is in fact investing in improving these tools, and in
contributing their changes back to the Free Software Foundation.

And keep the religious rethoric down, please ("virus", "plagues").

> > >> while Windows machines still have 16-bit MS-DOS at the core.
> > >
> > > No.
> >
> > Of what vintage?
>
> 2000

False, if by that you mean Windows 2000 and not any odd Windows version in
use in 2000. Windows 2000 is based on the NT (originally stood for "New
Technology"), not on MS-DOS, a completely different core that was 32 bit
from the beginning. (If I rwecall correctly, 2000 actually identifies
itself as NT version 5.0.)

> > It was true as late as Windows ME [despite attempts by MS to hide this]
>
> Asked and Answered.

Mainly showing evidence of FUD[1], though, not of actually understanding
the issues.

[1] Fear, Uncertainty and Doubt. Originally named after an infamous IBM
marketing technique; these days typically employed by The SCO Group ("All
your Unix are belong to us!") and Microsoft (see for example the so-called
Halloween-documents).

Kai
--
http://www.westfalen.de/private/khms/
"... by God I *KNOW* what this network is for, and you can't have it."
  - Russ Allbery (rra@stanford.edu)

jgoulden_news@goulden.org (John D. Goulden)  wrote on 07.11.05 in <dko4a802vt3@news3.newsguy.com>:

> and was a sort of 16 / 24 / 32 bit hybrid. However, Microsoft Windows 3.1
> running in 386 enhanced mode is a 32-bit, multitasking, virtual memory
> operating system - albeit one that uses lots of 16-bit chunks of DOS as
> services. Typing "win" at the DOS prompt runs the Windows loader, which
> unloads DOS and establishes 32-bit Windows as the OS. From Windows 3.1,
> File -Exit unloads Windows and reloads DOS. Before the screaming begins, one

That isn't actually true, as far as I know; DOS is not "unloaded" in any
meaningful sense, and the parts of DOS that Windows 3.1 uses are part of
that DOS from which win.com was started. You can wreck both that DOS and
the Wndows running on top of it by mishandling the relevant parts of
memory.

> should note that several distros of Linux could play this same trick: you
> can run their loader from the DOS prompt and start Linux, then "exit" Linux
> back into DOS. No one in their right mind would claim that Linux is a DOS
> program; neither is Windows 3.1.

I'm not aware of any version of Linux which allowed you to exit back to
DOS.

> The multitasking and virtual memory support in Win 3.1, while it was there
> and did work after a fashion, was primitive at best. Windows 95 introduced
> significant improvements and the Windows VMM system was changed quite a bit
> by Win 95 OSR2 and then pretty well established by Windows 98 (thus the many
> programs that require Win 98 or better). In all of these operating systems,
> the old DOS prompt can be executed as a service under 32-bit Windows, and
> you can run as many of them as you wish, concurrently, and each is tricked
> by Windows into thinking it owns its own real-mode computer. Windows 9x also

That is true, but it is also true that these versions still depend on the
DOS used for booting much like Windows 3.1 did. And in fact, they are also
still started from DOS with "win" and can exit back there.

> uses a number of old 16-bit DOS functions for things like dealing with the
> BIOS and system clock (mainly because there is little point in re-writing
> them as 32-bit versions). In addition, Windows 95 and later DO still carry
> some DOS architectural baggage (such as a requirement that process control
> blocks are restricted to a particular area in physical memory, which leads
> to occasional difficulties but makes some backwards-compatibility issues
> easier to resolve) but to say that Windows 95 = DOS is nonsense. Indeed, the

Of course: it's running on top of DOS, it isn't itself DOS. Or you might
say that DOS is a part of it (as these DOS versions identify themselves as
"Windows" even when running DOS-only). Incidentally, that's versions 7.x
of DOS.

> differences between Windows 3.0 and Windows 3.1 (which appear to be the same
> to the casual user) are MUCH more significant than are the differences
> between Win 3.1 and Win 95 (which look quite a bit different but are quite
> similar under the hood).

Win 95/98/ME have a 32 bit interface for user programs added. There was
also a backport for Win 3.1, called Win32s. Apart from the interface for
programs, they're not much different internally.

> Microsoft's new-from-the-ground-up completely-independent-of-DOS 32-bit OS
> with support for multitasking, multiprocessing, and virtual memory was
> Windows NT. It was written in large part by former VAX folks appropriated
> from DEC, and it's claimed to be coincidental that Windows NT (WNT) shares a
> similar relationship with VMS as HAL does to IBM (one step behind rather
> than one ahead, perhaps?).

Incidentally, the chief architect for NT has a history of trying to write
single-user versions of VMS. Somewhat ironic in view of the later
development, where Microsoft backfitted NT with multi-user capabilities
...

>The first version released was called NT 3.0 so
> that it would have the same version number as their existing current version
> of Windows; NT 1 and NT 2 did not exist. Oddly enough, NT 3 was raked over
> the coals for NOT having much backwards compatibility with existing DOS and
> Windows programs, and NT 4 retreated a bit from the somewhat idealistic
> architecture of NT 3 and threw in some backwards-compatibility features to
> placate the masses. NT 4 was a smashing success (for one thing, it runs on
> surprisingly limited hardware) and still very much around today (with the
> latest service pack it looks just like 95/98).

NT 4 has *always* looked like the 9x line. The last NT to look like
Windows 3.x was NT 3.51, possibly the most stable version of NT ever.

I believe the main difference for NT 4, apart from the new user interface,
was putting the video drivers into the kernel, making the system vastly
less stable but faster.

>NT 5 was marketed as Windows
> 2000; NT 5.1 is better known as Windows XP. XP in particular is marketed as
> both a home and office OS, and (like NT 4) sacrifices some desired
> theoretical purity for better compatibility with older programs and for
> general improvement of performance (achieved, for instance, by doing some
> rather unclean things with the graphics subsystem). I am to this day

No, no, no. The graphics bit, as explained above, was NT 4.

XP, I believe, was the one where Microsoft went all-out to be able to
support pretty much everything supported in 9x, so they could finally kill
that line off. The first move in that direction was 2000, but it didn't go
far enough - 2000 wasn't a very good platform for games especially.

> One can then say that we have three fairly distinct lines of OS products
> from Microsoft:
>
> 16-bit line
>
> MS-DOS 1.0 through 6.22, Windows 1 and 2, arguably big chunks of Windows 3.0
>
> 32-bit line #1, uses some 16-bit DOS components as services and erroneously
> said to be "built on DOS"

Nothing erroneous about that.

> Windows 3.1, 95, 95 OSR2, 98, 98 2nd, ME (best of breed is generally
> regarded to be 98 2nd)
>
> 32-bit line #2, developed independently of DOS but got some DOS-like
> functionality grafted in later for compatibility reasons

Not really grafted in later; I believe the DOS emulator was there from day
1.

> Windows NT 3, 4, 5 (aka 2000), 5.1 (aka XP)

One might also note that the 32 bit program interface exists on both the
Windows 9.x and NT line (see above for Win32s), and the 16 bit interface
is, I believe (though I've never seen Win 1.x) present on Windows 1.x
through the 9.x line and then emulated on NT, and the modern DOS program
interface (mostly stolen from Xenix) from DOS 2.0 through the 9x line and
emulated starting on 3.1 and ever since (yes, that means 3.1 and the 9x
line had both emulated and unemulated variants), and the original DOS
program interface, mostly stolen from CP/M, has been there since DOS 1.0
(actually, since QD-DOS which became MS-DOS 1.0 when Microsoft bought it)
and been handled the same as the modern one ever since that was created
with DOS 2.0.

Ok, maybe you prefer "inspired by" instead of "stolen from".

Kai
--
http://www.westfalen.de/private/khms/
"... by God I *KNOW* what this network is for, and you can't have it."
  - Russ Allbery (rra@stanford.edu)

> There has arisen this feeling that PC's should boot up flawlessly out
> of the box and run with nary a problem, crash, or lockup until there
> is a power outage or you unplug the system prior to running it down to
> the Salvation Army to donate.
>
> Balderdash.

I wonder if anybody has ever done even a back of the envelope
calculation of the cost in man-hours (not to mention the frustration)
from what is at best a fledgling technology.

With a nod to Frederick Brooks and the lack of silver bullets and the
abundance of werewolves, I won't even delve into functions that don't
work as you expect or that invoke the Law of Unexpected Conequences in
their interactions with one another, in complicated software -- I'm
just talking about severe bugs.

Consumer PCs are about where cars were in the 20s:  starting to get
fairly fast and able to get you places, but you still have to crank 'em
by hand and know how to adjust the spark and whatnot and be prepared to
spend a lot of time changing tires.


> to expect a commodity machine to perform so is ludicrous.  There is no need
> beyond delusion that such a performance level should be required.

I'm not comparing this particular commodity to aerospace or biomedical
or other high-consequence scenarios; I'm comparing it to readily
available and increasingly viable alternatives.  Some of those
alternatives give me uptime measurable in months (just how many months
I dunno, since with recent editions of Mac OS 10, the alternative with
which I'm most familiar, I  have to terminate the experiment  for an
upgrade, or electrical maintenance on the building, or somesuch, not
because of an OS crash or an application crash that takes down the OS).


Cheers,
--Joe

"Ad absurdum per aspera" <jtchew@california.com> wrote:

>Note that I said *fairly* stable.   I haven't kept statistics that
>would withstand scientific scrutiny, but my impression is that has
>never been worse than a good solid Windows 98 machine.    It is a good
>bit flakier in all ways (more crashes, more failures to boot up or shut
>down properly, more tries needed for more installer programs) than
>we've gotten accustomed to with XP/2000, and far less stable than my OS
>10.3 and 10.4 Macs.  But it is quite usable as a homework and
>video-games engine and emergency backup for word processing.

<nods>  My machine (which is a Win 98 machine) is about the same.  I
can depend on it to lock up about twice a day - almost always in a
game or reading Usenet, but never in Office.

>My sysadmin friends from both the Unix and the Windows sides of the
>game think I should sell tickets to this spectacle; all I can say is
>that it works okay for me.

There has arisen this feeling that PC's should boot up flawlessly out
of the box and run with nary a problem, crash, or lockup until there
is a power outage or you unplug the system prior to running it down to
the Salvation Army to donate.

Balderdash.

Nuclear power plant controls, aircraft flight control computers, these
and things like them should be held to such a standard - but to expect
a commodity machine to perform so is ludicrous.  There is no need
beyond delusion that such a performance level should be required.

D.
--
Touch-twice life. Eat. Drink. Laugh.

-Resolved: To be more temperate in my postings.
Oct 5th, 2004 JDL

"Derek Lyons" <fairwater@gmail.com> wrote in message
news:437236ee.4398492@news.supernews.com...
> Lon <lon.stowell@comcast.net> wrote:
>
>>As for cheap commodity boxen not being expected to be reliable,
>>balderbalderdash and bullshitbalderdash.
>
> Ah.  You feel that an assertion made without support is sufficient.
>
>>That type of thinking is why there is no American Auto industry any more.
>
> Except of course - for the American Auto Industry that employs many
> thousands of Americans in manufacturing.
>
Your response was to a hyperbole.  Our auto industry is still in existence
but it is losing out to the competition from Europe and Japan, possibly due
to the fact that they (the competition) haven't bought into the "planned
obsolescence" that our industry seems to pride itself on.

Don
--
Ever had one of those days where you just felt like:
http://cosmoslair.com/BadDay.html ?

"Derek Lyons" <fairwater@gmail.com> wrote in message
news:43733755.4501796@news.supernews.com...
> "Ad absurdum per aspera" <jtchew@california.com> wrote:
>>
>>> to expect a commodity machine to perform so is ludicrous.  There is no
>>> need
>>> beyond delusion that such a performance level should be required.
>>
>>I'm not comparing this particular commodity to aerospace or biomedical
>>or other high-consequence scenarios; I'm comparing it to readily
>>available and increasingly viable alternatives.  Some of those
>>alternatives give me uptime measurable in months
>
> So?  What does that give you beyond a warm and fuzzy feeling?
>
Reliability.

> So?  What does that give you beyond a warm and fuzzy feeling?

It's the difference between spending my time repairing my productivity
tools and spending it being productive.
Cheers,
--Joe

Derek Lyons proclaimed:

> Lon <lon.stowell@comcast.net> wrote:
>
>
>>As for cheap commodity boxen not being expected to be reliable,
>>balderbalderdash and bullshitbalderdash.
>
>
> Ah.  You feel that an assertion made without support is sufficient.

   Moi?  I think perhaps you made one and got a bullshit called on you.
   Take essentially the same box.  Change the operating system. Uptimes
   in years.  Not at all rare, odd, or unexpected.

>>That type of thinking is why there is no American Auto industry any more.
>
>
> Except of course - for the American Auto Industry that employs many
> thousands of Americans in manufacturing.

   Yes, as opposed to the many millions it used to before someone created
   cars clearly denoting the distinction between "cheap" and
   "inexpensive".

   Most are now employed by the overseas auto industry who happen to
   locate factories in the United States to sell foreign branded autos
   here.  Granted, there are a few locations in Europe that build for
   American automobile companies, but those are offset rather handily
   by the foreign automobile companies building american branded
   vehicles here...or in Mexico.

Don Freeman proclaimed:

> "Derek Lyons" <fairwater@gmail.com> wrote in message
> news:437236ee.4398492@news.supernews.com...
>
>>Lon <lon.stowell@comcast.net> wrote:
>>
>>
>>>As for cheap commodity boxen not being expected to be reliable,
>>>balderbalderdash and bullshitbalderdash.
>>
>>Ah.  You feel that an assertion made without support is sufficient.
>>
>>
>>>That type of thinking is why there is no American Auto industry any more.
>>
>>Except of course - for the American Auto Industry that employs many
>>thousands of Americans in manufacturing.
>>
>
> Your response was to a hyperbole.  Our auto industry is still in existence
> but it is losing out to the competition from Europe and Japan, possibly due
> to the fact that they (the competition) haven't bought into the "planned
> obsolescence" that our industry seems to pride itself on.

   Oddly enough, there are American auto companies that exceed the
   reliability ratings of such staid old lines as Mercedes.  With
   Porsche also scrambling to recover their quality image in the
   face of some heavy competition moving in.  This statement is
   not hyperbole, easily checked, and does not strain the boundaries
   by including Volvo as an American company--which it could.  This
   set of changes is so amazing that not only are *some* American
   models more reliable than some European, even Jaguar and Land
   Rover have learned how to spell the term without being spotted
   3 vowels, 3 consonants, and a free Lucas generator.

Lon <lon.stowell@comcast.net> wrote:

>As for cheap commodity boxen not being expected to be reliable,
>balderbalderdash and bullshitbalderdash.

Ah.  You feel that an assertion made without support is sufficient.

>That type of thinking is why there is no American Auto industry any more.

Except of course - for the American Auto Industry that employs many
thousands of Americans in manufacturing.

D.
--
Touch-twice life. Eat. Drink. Laugh.

-Resolved: To be more temperate in my postings.
Oct 5th, 2004 JDL

"Ad absurdum per aspera" <jtchew@california.com> wrote:
>
>> to expect a commodity machine to perform so is ludicrous.  There is no need
>> beyond delusion that such a performance level should be required.
>
>I'm not comparing this particular commodity to aerospace or biomedical
>or other high-consequence scenarios; I'm comparing it to readily
>available and increasingly viable alternatives.  Some of those
>alternatives give me uptime measurable in months

So?  What does that give you beyond a warm and fuzzy feeling?

D.
--
Touch-twice life. Eat. Drink. Laugh.

-Resolved: To be more temperate in my postings.
Oct 5th, 2004 JDL

On Mon, 7 Nov 2005 04:21:44 UTC, Steve Ackman
<steve@SNIP-THIS.twoloonscoffee.com> wrote:

>   I tried to 'ssh -Y' from my father's MacOSX machine
> to my FreeBSD machine.

New to Mac and OS/X here.  How did you get a command line?

--
   John Varela
   Trade OLD lamps for NEW for email.

In article <4370c364.1159776@news.supernews.com>,
Derek Lyons <fairwater@gmail.com> wrote:

snip--

>There has arisen this feeling that PC's should boot up flawlessly out
>of the box and run with nary a problem, crash, or lockup until there
>is a power outage or you unplug the system prior to running it down to
>the Salvation Army to donate.
>

i have no such expectations

but with some small effort (typically less than 4 hours per box)
i do obtain the behaviour described above...

most of the effort involves replacing the supplied OS from redmond

>Balderdash.
>
>Nuclear power plant controls, aircraft flight control computers, these
>and things like them should be held to such a standard - but to expect
>a commodity machine to perform so is ludicrous.  There is no need
>beyond delusion that such a performance level should be required.
>

i have some friends who have lost a great deal of time and work to
crashes who might disagree with you

while they do not desire or need flawless real time performance
they would greatly appreciate not losing their work at random times

i have had some success in weaning them away from microsoft,
but not as much as i would like

sidd