#341 From: "Lynn H. Maxson" <lmaxson@...>
Date: Wed Feb 27, 2002 8:24 am
Subject: Re: A gift horse... lynnmaxson
Offline Offline
Send Email Send Email
Invite to Yahoo! 360° Invite to Yahoo! 360°
ShadoW writes:
"Perhaps he is using an editor that assembles phrases which are
stored in a database."
Part 12 - Feb 26 2002
Re: Part 12
#342 From: J Christopher Kennedy <kb7nmu@...>
Date: Thu Feb 28, 2002 11:07 am
Subject: Re: Re: Licence for this project. kb7nmu
Offline Offline
Send Email Send Email
Invite to Yahoo! 360° Invite to Yahoo! 360°
In <a5e33g+ofnl@eGroups.com>, on 02/25/2002
at 07:25 PM, "drittervonfuenf" <3rdof5@...> said:
>The Copyright of a sourcefile stays always with the author(s). You can't
>give away your copyright, thats not legaly allowed. You can put source
>under GPL, and make it free to use etc, but thats a different. Even if you
>write code for your
>employer, you have the copyright, the employer "only" the IP.
Um, not true. Copyright includes the right to assign or sell all rights
whatsoever. You are still the author, but you can give up all rights. Just
look at all the early Science Fiction that is copyrighted by Street & Smith
instead of by the Authors. Not to mention that Copyright Law of 1978 allows
for permanent sale of rights. (Lets just ignore the hell that is DMCA. No
really, ignore it.)
Of course the above paragraph only applies to the USA.
I recommend that a paragraph be added that disclaims all rights granted by
the DMCA. Standard copyright didn't need no fixing, at least not in the
USA.
--
-----------------------------------------------------------
J Christopher Kennedy <kb7nmu@...>
http://www.dragonbard.com/
-----------------------------------------------------------
Date: Thu Feb 28, 2002 11:07 am
Subject: Re: Re: Licence for this project. kb7nmu
Offline Offline
Send Email Send Email
Invite to Yahoo! 360° Invite to Yahoo! 360°
In <a5e33g+ofnl@eGroups.com>, on 02/25/2002
at 07:25 PM, "drittervonfuenf" <3rdof5@...> said:
>The Copyright of a sourcefile stays always with the author(s). You can't
>give away your copyright, thats not legaly allowed. You can put source
>under GPL, and make it free to use etc, but thats a different. Even if you
>write code for your
>employer, you have the copyright, the employer "only" the IP.
Um, not true. Copyright includes the right to assign or sell all rights
whatsoever. You are still the author, but you can give up all rights. Just
look at all the early Science Fiction that is copyrighted by Street & Smith
instead of by the Authors. Not to mention that Copyright Law of 1978 allows
for permanent sale of rights. (Lets just ignore the hell that is DMCA. No
really, ignore it.)
Of course the above paragraph only applies to the USA.
I recommend that a paragraph be added that disclaims all rights granted by
the DMCA. Standard copyright didn't need no fixing, at least not in the
USA.
--
-----------------------------------------------------------
J Christopher Kennedy <kb7nmu@...>
http://www.dragonbard.com/
-----------------------------------------------------------
Re: Part 12
#343 From: "tomleem7659" <jersey@...>
Date: Thu Feb 28, 2002 9:42 pm
Subject: Re: osFree and FreeOS? tomleem7659
Offline Offline
Send Email Send Email
Invite to Yahoo! 360° Invite to Yahoo! 360°
> The FreeOS project has been going on for many years andnot
> a single line of code has been written.
>
> osFree is about doing things now, not dreaming and talking.
>
> Btw: The freeos.org webpage is a Linux thing and has nothing
> to do with the FreeOS yahoogroup.
>
>
>
>
> Sincerely
>
> JMA
> Development and Consulting
I have a feeling the 'freeos' group isn't sure what
they want to do; create a version of linux that supports
OS/2 or create a free version of OS/2.
TomLeeM
Date: Thu Feb 28, 2002 9:42 pm
Subject: Re: osFree and FreeOS? tomleem7659
Offline Offline
Send Email Send Email
Invite to Yahoo! 360° Invite to Yahoo! 360°
> The FreeOS project has been going on for many years andnot
> a single line of code has been written.
>
> osFree is about doing things now, not dreaming and talking.
>
> Btw: The freeos.org webpage is a Linux thing and has nothing
> to do with the FreeOS yahoogroup.
>
>
>
>
> Sincerely
>
> JMA
> Development and Consulting
I have a feeling the 'freeos' group isn't sure what
they want to do; create a version of linux that supports
OS/2 or create a free version of OS/2.
TomLeeM
Re: Part 12
#344 From: "Michal Necasek" <michaln@...>
Date: Thu Feb 28, 2002 9:45 pm
Subject: Re: Re: Licence for this project. michalnec
Offline Offline
Send Email Send Email
Invite to Yahoo! 360° Invite to Yahoo! 360°
On Thu, 28 Feb 2002 00:07:20 -0800, J Christopher Kennedy wrote:
>Um, not true. Copyright includes the right to assign or sell all rights
>whatsoever.
>
Not in European countries AFAIK.
But there's the concept of "work for hire".
Michal
Date: Thu Feb 28, 2002 9:45 pm
Subject: Re: Re: Licence for this project. michalnec
Offline Offline
Send Email Send Email
Invite to Yahoo! 360° Invite to Yahoo! 360°
On Thu, 28 Feb 2002 00:07:20 -0800, J Christopher Kennedy wrote:
>Um, not true. Copyright includes the right to assign or sell all rights
>whatsoever.
>
Not in European countries AFAIK.
But there's the concept of "work for hire".
Michal
Re: Part 12
#345 From: "timur_tabi" <timur@...>
Date: Fri Mar 1, 2002 7:52 pm
Subject: Re: Linux + OS/2 layer timur_tabi
Offline Offline
Send Email Send Email
Invite to Yahoo! 360° Invite to Yahoo! 360°
--- In osFree@y..., "sandervl2000" <sandervl2000@y...> wrote:
> 1) Build a Wine like 'emulator' for OS/2 applications; emulate
> every OS/2/PM api call with standard linux/x11 functions
>
> 2) Use linux kernel + X11 as a base to basically build
> a linux distribution with OS/2 appearance (+ application
interface)
> (similar to Lindows; linux distro meant as a windows replacement)
>
> 3) Only use the linux kernel (ignoring X11) and build your own PM &
> WPS subsystem.
>
> The first one seems pointless as there are not enough OS/2
> applications to justify the work.
I disagree. I use Linux every day, and I curse it every day. There
are more than enough OS/2 applications for me, but I wish I wouldn't
have to "fight" the OS/2 kernel to get better hardware support. And
I'm not just talking driver availability, I'm talking about how well
the OS supports the hardware. Driver functionality is so uneven, that
it's not enough to ask whether OS/2 supports a particular piece of
hardware, but rather whether it supports it well. The only exception
is video support, thanks to SciTech.
>Option 3 doesn't have much appeal
> either.
It appeals a great deal to me.
> The 2nd one seems the most interesting to me.
But what point would it have? The OS/2 look-and-feel is not a strong
selling point by itself. Linux GUI applications are crap. The GUI
elements are ugly and often mis-aligned. Standard CUA concepts are
usually ignored. Linux GUIs and GUI apps look like that've been
cobbled together with bailing wire and chewing gum (to use an American
expression).
> You'll still be stuck with base (kernel) that has an ancient (&
> possibly incompatible) design.
Ancient? You mean the Linux kernel!?!?!?
> Another possible hazard is the unclear future of linux. It might
> be bright and shiny now, but I doubt it can ever compate with MS
> on the desktop. Once people realize that, linux development might
> grind to a halt. (I might be completely wrong though)
Sander, you're definitely a bright guy, which is why this statement
surprises me. Linux isn't about taking the desktop away from
Microsoft. A lot of Linux users want that, but it's quite obvious
that the most don't care enough. It hasn't stopped any of them in the
past, and it certainly won't stop them in the future.
Linux is for Linux users and developers. You don't have to dominate
in order to survive.
> Driver availability is ok (I suppose) for Linux, but having
> a NT/Win2k/XP compatible driver model ensures more and supported
> drivers. (That's a definite plus for ReactOS)
But it's pointless. Microsoft changes the OS interface with
practically every revision. The ReactOS developers won't be able to
keep up.
> If there is a kernel available that does all we need, then we
should
> by all means use it. However, I haven't found one yet.
That's why we should take the closest thing to perfection we have, and
improve it. I've made numerous arguments in favor of Linux (you can
read them on the FreeOS mailing list), but the #1 point I've been
making is that any kernel you choose must be very much "alive" on its
own. That is, development on it must be active and thorough, and must
have reached "critical mass". Maintaining an OS kernel requires A LOT
of work, much more than any single group of developers. Linux has
reached critical mass, because there are A LOT of developers who want
to work on it, and keep it going. Even if a dozen developers drop
out, it won't matter. ReactOS, OS/2, even the *BSDs, don't really
have that advantage.
Linux and Windows are the only OS's (for PCs at least) that qualify.
Windows is not an option for the obvious reasons, so that leaves
Linux. Using Linux as a base allows us to IGNORE kernel development.
Date: Fri Mar 1, 2002 7:52 pm
Subject: Re: Linux + OS/2 layer timur_tabi
Offline Offline
Send Email Send Email
Invite to Yahoo! 360° Invite to Yahoo! 360°
--- In osFree@y..., "sandervl2000" <sandervl2000@y...> wrote:
> 1) Build a Wine like 'emulator' for OS/2 applications; emulate
> every OS/2/PM api call with standard linux/x11 functions
>
> 2) Use linux kernel + X11 as a base to basically build
> a linux distribution with OS/2 appearance (+ application
interface)
> (similar to Lindows; linux distro meant as a windows replacement)
>
> 3) Only use the linux kernel (ignoring X11) and build your own PM &
> WPS subsystem.
>
> The first one seems pointless as there are not enough OS/2
> applications to justify the work.
I disagree. I use Linux every day, and I curse it every day. There
are more than enough OS/2 applications for me, but I wish I wouldn't
have to "fight" the OS/2 kernel to get better hardware support. And
I'm not just talking driver availability, I'm talking about how well
the OS supports the hardware. Driver functionality is so uneven, that
it's not enough to ask whether OS/2 supports a particular piece of
hardware, but rather whether it supports it well. The only exception
is video support, thanks to SciTech.
>Option 3 doesn't have much appeal
> either.
It appeals a great deal to me.
> The 2nd one seems the most interesting to me.
But what point would it have? The OS/2 look-and-feel is not a strong
selling point by itself. Linux GUI applications are crap. The GUI
elements are ugly and often mis-aligned. Standard CUA concepts are
usually ignored. Linux GUIs and GUI apps look like that've been
cobbled together with bailing wire and chewing gum (to use an American
expression).
> You'll still be stuck with base (kernel) that has an ancient (&
> possibly incompatible) design.
Ancient? You mean the Linux kernel!?!?!?
> Another possible hazard is the unclear future of linux. It might
> be bright and shiny now, but I doubt it can ever compate with MS
> on the desktop. Once people realize that, linux development might
> grind to a halt. (I might be completely wrong though)
Sander, you're definitely a bright guy, which is why this statement
surprises me. Linux isn't about taking the desktop away from
Microsoft. A lot of Linux users want that, but it's quite obvious
that the most don't care enough. It hasn't stopped any of them in the
past, and it certainly won't stop them in the future.
Linux is for Linux users and developers. You don't have to dominate
in order to survive.
> Driver availability is ok (I suppose) for Linux, but having
> a NT/Win2k/XP compatible driver model ensures more and supported
> drivers. (That's a definite plus for ReactOS)
But it's pointless. Microsoft changes the OS interface with
practically every revision. The ReactOS developers won't be able to
keep up.
> If there is a kernel available that does all we need, then we
should
> by all means use it. However, I haven't found one yet.
That's why we should take the closest thing to perfection we have, and
improve it. I've made numerous arguments in favor of Linux (you can
read them on the FreeOS mailing list), but the #1 point I've been
making is that any kernel you choose must be very much "alive" on its
own. That is, development on it must be active and thorough, and must
have reached "critical mass". Maintaining an OS kernel requires A LOT
of work, much more than any single group of developers. Linux has
reached critical mass, because there are A LOT of developers who want
to work on it, and keep it going. Even if a dozen developers drop
out, it won't matter. ReactOS, OS/2, even the *BSDs, don't really
have that advantage.
Linux and Windows are the only OS's (for PCs at least) that qualify.
Windows is not an option for the obvious reasons, so that leaves
Linux. Using Linux as a base allows us to IGNORE kernel development.
Re: Part 12
#346 From: "tomleem7659" <jersey@...>
Date: Fri Mar 1, 2002 9:39 pm
Subject: Nudawn and OSFree? tomleem7659
Offline Offline
Send Email Send Email
Invite to Yahoo! 360° Invite to Yahoo! 360°
Is Nudawn at http://nudawn.sourceforge.net -
http://sourceforge.net/projects/nudawn similar
to osFree?
TomLeeM
PS IIRC on the url.
Date: Fri Mar 1, 2002 9:39 pm
Subject: Nudawn and OSFree? tomleem7659
Offline Offline
Send Email Send Email
Invite to Yahoo! 360° Invite to Yahoo! 360°
Is Nudawn at http://nudawn.sourceforge.net -
http://sourceforge.net/projects/nudawn similar
to osFree?
TomLeeM
PS IIRC on the url.
Re: Part 12
#347 From: "JMA" <mail@...>
Date: Fri Mar 1, 2002 10:22 pm
Subject: Re: Nudawn and OSFree? mailjmase
Offline Offline
Send Email Send Email
Invite to Yahoo! 360° Invite to Yahoo! 360°
On Fri, 01 Mar 2002 18:39:29 -0000, tomleem7659 wrote:
>Is Nudawn at http://nudawn.sourceforge.net -
>http://sourceforge.net/projects/nudawn similar
>to osFree?
>
Dont know, it has nothing on the pages except:
"NuDawn is an open source operating system guaranteeing support for DOS, Windows
(Win16/Win32), and OS/2 applications in one simple,
configurable package. Legacy harwdare support is the prime directive of NuDawn,
allowing the installation on XT/286/386/486"
Only one person are there (though registered as two) and there are no source not
docs nothing...
Sincerely
JMA
Development and Consulting
John Martin , jma@...
==================================
Website: http://www.jma.se/
email: mail@...
Phone: 46-(0)70-6278410
==================================
Date: Fri Mar 1, 2002 10:22 pm
Subject: Re: Nudawn and OSFree? mailjmase
Offline Offline
Send Email Send Email
Invite to Yahoo! 360° Invite to Yahoo! 360°
On Fri, 01 Mar 2002 18:39:29 -0000, tomleem7659 wrote:
>Is Nudawn at http://nudawn.sourceforge.net -
>http://sourceforge.net/projects/nudawn similar
>to osFree?
>
Dont know, it has nothing on the pages except:
"NuDawn is an open source operating system guaranteeing support for DOS, Windows
(Win16/Win32), and OS/2 applications in one simple,
configurable package. Legacy harwdare support is the prime directive of NuDawn,
allowing the installation on XT/286/386/486"
Only one person are there (though registered as two) and there are no source not
docs nothing...
Sincerely
JMA
Development and Consulting
John Martin , jma@...
==================================
Website: http://www.jma.se/
email: mail@...
Phone: 46-(0)70-6278410
==================================
Re: Part 12
#348 From: "sandervl2000" <sandervl2000@...>
Date: Sat Mar 2, 2002 1:40 am
Subject: Re: Linux + OS/2 layer sandervl2000
Offline Offline
Send Email Send Email
Invite to Yahoo! 360° Invite to Yahoo! 360°
--- In osFree@y..., "timur_tabi" <timur@t...> wrote:
> > The first one seems pointless as there are not enough OS/2
> > applications to justify the work.
> I disagree. I use Linux every day, and I curse it every day. There
> are more than enough OS/2 applications for me, but I wish I wouldn't
> have to "fight" the OS/2 kernel to get better hardware support. And
> I'm not just talking driver availability, I'm talking about how well
> the OS supports the hardware. Driver functionality is so uneven,
that
> it's not enough to ask whether OS/2 supports a particular piece of
I think you are overreacting. I have drivers for every piece of
hardware
I own. If there's something missing, I write it myself.
Is the driver situation with OS/2 really that much worse than in
Linux?
Linux drivers can be ported back, you know.
> > The 2nd one seems the most interesting to me.
> But what point would it have? The OS/2 look-and-feel is not a
strong
> selling point by itself. Linux GUI applications are crap. The GUI
> elements are ugly and often mis-aligned. Standard CUA concepts are
> usually ignored. Linux GUIs and GUI apps look like that've been
> cobbled together with bailing wire and chewing gum (to use an
American
> expression).
Sure, but what kind of audience are you trying to reach? The small
and shrinking OS/2 userbase or do you wish to attract new users?
Ditching X means you only have the linux kernel + drivers and loose
access to all existing GUI linux apps (although I'm not impressed
with them either so far)
There are far more developers working on linux applications than on
the kernel & drivers.
> > You'll still be stuck with base (kernel) that has an ancient (&
> > possibly incompatible) design.
> Ancient? You mean the Linux kernel!?!?!?
Yes. Ancient. I really dislike the linux kernel. It doesn't even
have an abstract interface for drivers. Everybody can access internal
kernel structures and a lot of drivers do that.
Not to mention the total lack of backwards compatibility for driver
binaries.
When I have to make a list of good kernel designs (that I know), then
the order is: NT, OS/2, Linux, Win9x
> > Driver availability is ok (I suppose) for Linux, but having
> > a NT/Win2k/XP compatible driver model ensures more and supported
> > drivers. (That's a definite plus for ReactOS)
> But it's pointless. Microsoft changes the OS interface with
> practically every revision. The ReactOS developers won't be able to
> keep up.
It's not that bad. Emulating kernel services is no big deal.
> That's why we should take the closest thing to perfection we have,
and
> improve it. I've made numerous arguments in favor of Linux (you can
> read them on the FreeOS mailing list), but the #1 point I've been
> making is that any kernel you choose must be very much "alive" on
its
> own. That is, development on it must be active and thorough, and
must
> have reached "critical mass". Maintaining an OS kernel requires A
LOT
> of work, much more than any single group of developers. Linux has
> reached critical mass, because there are A LOT of developers who
want
> to work on it, and keep it going. Even if a dozen developers drop
> out, it won't matter. ReactOS, OS/2, even the *BSDs, don't really
> have that advantage.
You're right about that. You should however keep your options open
and try to write the OS/2, PM and WPS replacement as portable as
possible. Just to avoid getting stuck with a single solution.
Sander
Date: Sat Mar 2, 2002 1:40 am
Subject: Re: Linux + OS/2 layer sandervl2000
Offline Offline
Send Email Send Email
Invite to Yahoo! 360° Invite to Yahoo! 360°
--- In osFree@y..., "timur_tabi" <timur@t...> wrote:
> > The first one seems pointless as there are not enough OS/2
> > applications to justify the work.
> I disagree. I use Linux every day, and I curse it every day. There
> are more than enough OS/2 applications for me, but I wish I wouldn't
> have to "fight" the OS/2 kernel to get better hardware support. And
> I'm not just talking driver availability, I'm talking about how well
> the OS supports the hardware. Driver functionality is so uneven,
that
> it's not enough to ask whether OS/2 supports a particular piece of
I think you are overreacting. I have drivers for every piece of
hardware
I own. If there's something missing, I write it myself.
Is the driver situation with OS/2 really that much worse than in
Linux?
Linux drivers can be ported back, you know.
> > The 2nd one seems the most interesting to me.
> But what point would it have? The OS/2 look-and-feel is not a
strong
> selling point by itself. Linux GUI applications are crap. The GUI
> elements are ugly and often mis-aligned. Standard CUA concepts are
> usually ignored. Linux GUIs and GUI apps look like that've been
> cobbled together with bailing wire and chewing gum (to use an
American
> expression).
Sure, but what kind of audience are you trying to reach? The small
and shrinking OS/2 userbase or do you wish to attract new users?
Ditching X means you only have the linux kernel + drivers and loose
access to all existing GUI linux apps (although I'm not impressed
with them either so far)
There are far more developers working on linux applications than on
the kernel & drivers.
> > You'll still be stuck with base (kernel) that has an ancient (&
> > possibly incompatible) design.
> Ancient? You mean the Linux kernel!?!?!?
Yes. Ancient. I really dislike the linux kernel. It doesn't even
have an abstract interface for drivers. Everybody can access internal
kernel structures and a lot of drivers do that.
Not to mention the total lack of backwards compatibility for driver
binaries.
When I have to make a list of good kernel designs (that I know), then
the order is: NT, OS/2, Linux, Win9x
> > Driver availability is ok (I suppose) for Linux, but having
> > a NT/Win2k/XP compatible driver model ensures more and supported
> > drivers. (That's a definite plus for ReactOS)
> But it's pointless. Microsoft changes the OS interface with
> practically every revision. The ReactOS developers won't be able to
> keep up.
It's not that bad. Emulating kernel services is no big deal.
> That's why we should take the closest thing to perfection we have,
and
> improve it. I've made numerous arguments in favor of Linux (you can
> read them on the FreeOS mailing list), but the #1 point I've been
> making is that any kernel you choose must be very much "alive" on
its
> own. That is, development on it must be active and thorough, and
must
> have reached "critical mass". Maintaining an OS kernel requires A
LOT
> of work, much more than any single group of developers. Linux has
> reached critical mass, because there are A LOT of developers who
want
> to work on it, and keep it going. Even if a dozen developers drop
> out, it won't matter. ReactOS, OS/2, even the *BSDs, don't really
> have that advantage.
You're right about that. You should however keep your options open
and try to write the OS/2, PM and WPS replacement as portable as
possible. Just to avoid getting stuck with a single solution.
Sander
Re: Part 12
#349 From: "Michal Necasek" <michaln@...>
Date: Sat Mar 2, 2002 4:10 am
Subject: Re: Re: Linux + OS/2 layer michalnec
Offline Offline
Send Email Send Email
Invite to Yahoo! 360° Invite to Yahoo! 360°
On Fri, 01 Mar 2002 22:40:23 -0000, sandervl2000 wrote:
>When I have to make a list of good kernel designs (that I know), then
>the order is: NT, OS/2, Linux, Win9x
>
Hmm, I thought Win9x was in the "horrible design" category<g>?
But other than that I agree with your ordering. The only fault
of the NT kernel is that it's maybe a little overengineered.
Michal
Date: Sat Mar 2, 2002 4:10 am
Subject: Re: Re: Linux + OS/2 layer michalnec
Offline Offline
Send Email Send Email
Invite to Yahoo! 360° Invite to Yahoo! 360°
On Fri, 01 Mar 2002 22:40:23 -0000, sandervl2000 wrote:
>When I have to make a list of good kernel designs (that I know), then
>the order is: NT, OS/2, Linux, Win9x
>
Hmm, I thought Win9x was in the "horrible design" category<g>?
But other than that I agree with your ordering. The only fault
of the NT kernel is that it's maybe a little overengineered.
Michal
Re: Part 12
#350 From: Bert Thomas <bert@...>
Date: Sat Mar 2, 2002 10:29 am
Subject: Re: Re: Linux + OS/2 layer bertbrothom
Offline Offline
Send Email Send Email
Invite to Yahoo! 360° Invite to Yahoo! 360°
sandervl2000 schreef:
>
> --- In osFree@y..., "timur_tabi" <timur@t...> wrote:
> > > You'll still be stuck with base (kernel) that has an ancient (&
> > > possibly incompatible) design.
> > Ancient? You mean the Linux kernel!?!?!?
> Yes. Ancient. I really dislike the linux kernel. It doesn't even
> have an abstract interface for drivers. Everybody can access internal
> kernel structures and a lot of drivers do that.
> Not to mention the total lack of backwards compatibility for driver
> binaries.
Interesting, I never looked at it that way. You're right, they added
something to the module loading mechanism to detect kernel
version/driver version mismatches and prevent the driver from loading. I
think however that in most cases it will be enough to re'make' the
driver again to get it working.
> When I have to make a list of good kernel designs (that I know), then
> the order is: NT, OS/2, Linux, Win9x
>
What about the trick to put parts of Win32 subsystem in ring 0, just for
performance reasons? That's not very elegant I think...?
What about BeOS? It was object oriented designed from the ground up, and
it performes very well. I once had several movies running on my desktop
without any glitch, BeOS mixing the audiostreams in realtime. With
Windows NT, sliding the scroll bar of Netscape is enough to get just an
audio stream disturbed. If someone keeps 'ping'ing my machine, it just
gets locked up for me, apparantly busy handling the interrupts of the
NIC. And what about the priority inversion bug in Windows NT? This is
the reason why OS/2 has soft-realtime support and NT hasn't.
Bert
Date: Sat Mar 2, 2002 10:29 am
Subject: Re: Re: Linux + OS/2 layer bertbrothom
Offline Offline
Send Email Send Email
Invite to Yahoo! 360° Invite to Yahoo! 360°
sandervl2000 schreef:
>
> --- In osFree@y..., "timur_tabi" <timur@t...> wrote:
> > > You'll still be stuck with base (kernel) that has an ancient (&
> > > possibly incompatible) design.
> > Ancient? You mean the Linux kernel!?!?!?
> Yes. Ancient. I really dislike the linux kernel. It doesn't even
> have an abstract interface for drivers. Everybody can access internal
> kernel structures and a lot of drivers do that.
> Not to mention the total lack of backwards compatibility for driver
> binaries.
Interesting, I never looked at it that way. You're right, they added
something to the module loading mechanism to detect kernel
version/driver version mismatches and prevent the driver from loading. I
think however that in most cases it will be enough to re'make' the
driver again to get it working.
> When I have to make a list of good kernel designs (that I know), then
> the order is: NT, OS/2, Linux, Win9x
>
What about the trick to put parts of Win32 subsystem in ring 0, just for
performance reasons? That's not very elegant I think...?
What about BeOS? It was object oriented designed from the ground up, and
it performes very well. I once had several movies running on my desktop
without any glitch, BeOS mixing the audiostreams in realtime. With
Windows NT, sliding the scroll bar of Netscape is enough to get just an
audio stream disturbed. If someone keeps 'ping'ing my machine, it just
gets locked up for me, apparantly busy handling the interrupts of the
NIC. And what about the priority inversion bug in Windows NT? This is
the reason why OS/2 has soft-realtime support and NT hasn't.
Bert