Linux-Advocacy Digest #883, Volume #25           Thu, 30 Mar 00 18:13:08 EST

Contents:
  Re: Giving up on NT (Wolfgang Weisselberg)
  Re: Windows 2000: nothing worse (mlw)
  Re: Opensource article first chapter draft for criticism (Ron Reeder)
  Re: Windows 2000: nothing worse ("Christopher Smith")
  xfs is out! ("horst")
  Re: Windows 2000: nothing worse (Gary Hallock)
  Re: Windows 2000: nothing worse (mlw)
  Re: Windows 2000: nothing worse (JEDIDIAH)
  Re: Windows 2000: nothing worse ("Christopher Smith")
  Re: Why did we even need NT in the first place?

----------------------------------------------------------------------------

From: [EMAIL PROTECTED] (Wolfgang Weisselberg)
Crossposted-To: 
comp.os.ms-windows.nt.advocacy,comp.os.os2.advocacy,comp.sys.mac.advocacy
Subject: Re: Giving up on NT
Date: 30 Mar 2000 21:17:08 GMT
Reply-To: [EMAIL PROTECTED]

On Tue, 28 Mar 2000 22:39:28 GMT,
        Chris Wenham <[EMAIL PROTECTED]> wrote:

> On 3/28/00, 1:03:55 PM, [EMAIL PROTECTED] (Wolfgang 
> Weisselberg) wrote regarding Re: Giving up on NT:

> > Run fvwm-95 on it, if you need 'similar to M$-Windows'[1].  Or
> > KDE.  Or any other 'tuned' WM.  Actually, try a different WM.
> > Your users might like that.  Andf if not ... well, let them choose
> > a different one.

> > Remember that you can run almost any window manager under X.  X
> > itself is -- as you probably know -- not much different from a
> > driver in function: It gives an open, standardized, usable,
> > network-transparent, ... interface to accessing the screen.  And
> > yes, X runs under most unixes, you can use it from your Win-PC
> > (MIX, eXceed, ...) and as you might have guessed: there are quite
> > a few implementations (including X-Terminals).  At least one of
> > these (XFree86 -- for 80x86) is free (GPL).

>  I think the idea that a window manager can give you the Windows user 
> interface is deceptive. 

Why don't you say: No, AFAIK you cannot because $X, $Y and $Z?  

"think" is an interesting way to claim something without needing
to give proof.  Especially as you forgot the "*similar to*" part.
If your requirement is 'never ever change the interface, not even
a bit', then stay with your current version.  Forever.

And unless you specify which parts you need, nobody can tell you
it it's being emulated now.

>  You can change a window manager, but you can't change the toolkit 
> that the application was built with*. QT, GTK, Xtoolkit and so-on all 
> look and behave differently. 

> * - Please amuse me by suggesting I ought to hack the source code.

Why should I?  I merely point out that it took just a couple of
days to port the freed sources of Netscape to QT.  By people who
had never seen the code before.  And for you I would advise you to
hire a few programmers from all the license cost you will safe to
write a glue-wrapper for toolkit A, so it uses internally toolkit
B.  Then simply tell your linker to load the wrapper instead of
toolkit A.  No recompiling required.

Now you'll whine about the inconsistent interfaces between
programs and I'll have to point out that that is not different
from Windows, as it stands.  I mean, "Start" to stop is somewhat
consistent, yes?

>  You can't even change the behavior of basic cutting and pasting with 
> your window manager.

Can you change the *basic* behavior of "point to focus/click
to focus" or can you switch off the 'click anywhere to pop-up'
with Windows WM?
Does the Windows-WM understand virtual screens?

Those are things that do belong to a WM.

cut&paste are not functions of the WM, IMHO, they are in X, but
you can (at least partially) access them from the outside.  Want
multible cut buffers?  xcb.  (Pray tell, can you change the C&P
behavior in Windows?  How?)

-Wolfgang

------------------------------

From: mlw <[EMAIL PROTECTED]>
Crossposted-To: comp.os.ms-windows.nt.advocacy
Subject: Re: Windows 2000: nothing worse
Date: Thu, 30 Mar 2000 16:18:31 -0500

Christopher Smith wrote:
> 
> "abraxas" <[EMAIL PROTECTED]> wrote in message
> news:8c06fp$14b0$[EMAIL PROTECTED]...
> > > I have real problems classifying NT as Multiplatform. two or three tops.
> > > Multiplatform is something like netbsd. Linux, although originally
> > > developed for the x86, has been proven to be capable of being
> > > multiplatform. NT has only been ported to alpha, mips, and is Merced out
> > > yet? Mips was dropped. The Alpha's port is on shaky ground (is it on
> > > again or off again this week?)
> 
> It's somewhat unfair to compare a commercial OS (that has to be profitable)
> to free OSes like that.

Why? If they claim it is portable, then porting it to other platforms
should be part of the business model. If the OS is not portable, stop
calling it portable.

You mean to tell me that NT on a power PC wouldn't be an interresting
market? How about NT on Sun hardware? These are targets that should make
marketing people drool. Maybe not for sales of that product, but for the
argument that NT is portable and therefore a viable option. 

As it is now, portability of NT is, at best, a joke.

> 
> >
> > Its off again.
> >
> > They also dropped a PPC port a few years ago.  Just when it was looking
> > quite nice actually.
> 
> Quite nice ?  Just who was making machines that could run it ?  How about
> software ?

Actually SGI was using mips. It was a political move more than anything.

> 
> > Go figure.
> >
> > I guess x86 hardware is "superior" for some reason.
> 
> In some ways.  It's common, cheap and fast, relatively speaking.

x86 hardware is cheap, that is, more or less, its only advantage.

-- 
Mohawk Software
Windows 9x, Windows NT, UNIX, Linux. Applications, drivers, support. 
Visit http://www.mohawksoft.com
"We've got a blind date with destiny, and it looks like she ordered the
lobster"

------------------------------

From: Ron Reeder <[EMAIL PROTECTED]>
Subject: Re: Opensource article first chapter draft for criticism
Date: Thu, 30 Mar 2000 14:38:34 -0700

Tom Steinberg wrote:
> 
> This is a great concept, which is very interesting from an economic point of
> view. Do you know of any examples you could suggest where putting work in
> now has been weighed up against reduced long term costs, like you suggest
> above?
> 
>     thanks again,
>         Tom Steinberg

Ok, you got me. - I don't have any good examples.

(Remember you can get good info on usenet... but signal to noise ratio is 
often well below one) 

Other than the obvious - RedHat, SUSE, and Caldera development on Linux 
kernel and utilities.  These are contrived, because of course these companies
benefit directly - still the code they develope goes into the public domain.
In fact, the days of these major open source projects being written by
vollunteers
has largely gone away already. Most principle contributers either being directly
employeed to do open systems work or supported in it as a sideline effort 
at work - Such as Linus T. himself. (His main job at Transmeta is developing
a version of linux for portables - not a version a distribution.) 

The thought came to me because of a flame-war I read on usenet five years 
ago between
a consultant/programmer and the Company that had hired him to do some
work on some GNU utility (I think it was gcc) They wanted a cross 
compiler for some new specialty processor (some kind of embedded app).
I forget what the flame war was about... but the fact that someone had
hired him specifically to work on opensource stuck in my head.
This is what I was basing my statement on. 


I know that at my company - I and another gentleman are attempting to re-work
the SCSI tape device driver to implement a more complete set of drive commands,
and implement error reporting and real time performance statistics - We are
assuming that our mods may well be accecpted into the general code base...

I know that a lot of the work on ethernet device drivers was done at NASA 
where some of the first Beowulf Clusters where attempted (Beowulf - networking
lots of small computers together via a network to make a super computer) 
The key bit there is networked... the origional linux ethernet device drivers,
where _just good enough_  to work... in a cluster to need them to perform well.
So, a NASA programmer updated them ... but in that case became the curator for 
many of the device drivers. Hmm, bad example.

This goes back to the GPL, if companies want certain functionality added
to a product... and that most other people would too.. then they can pay
to have it developed and know that it will remain in the public domain,
even after someone else begins maintaining it. 


To get a real answer to this you'd have to:

a) Ask the current keepers of the code - if they have absorbed code from
companies or consultants that desired specific functionality.   OR,

b) Review the development threads.  
The mailing lists where the active developers and testers discuss changes and
debugging
(ie.  linux-kernel)  are often sorted into threads  of conversation/topic.  
These are very informative... and part of the "Free" in the "Free Software
Foundation"
That is, not only do you have access to the source... but, you have access to
the discussions the devopers had, on why the programs are implemented as they
are.  

That may not appear amazing (which it is)  but remember, that a lot of people 
have been complaining about MS having secret or non-disclosed interfaces that
their applications programmers can use to make MS code run better/faster on 
MS OS'  - I suspect that it isn't that the API's are secret, but just better
understood within MS. 
 
or C)  Ask in the usenet.... 

Which I'll do by cross - posting this reply...   

I hope you don't mind...  It's generally considered bad manners
to publicly post private e-mail.  But, your goal is info.... 
and I haven't helped. 



-- 

+-------------------------------+-------------------------------+
| Ron Reeder                    | [EMAIL PROTECTED]           |
| Denver Technical Support      | Phone: (303) 389-4408         |
| Western Geophysical Company   | Fax:   (303) 595-0667         |
+-------------------------------+-------------------------------+

------------------------------

From: "Christopher Smith" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.ms-windows.nt.advocacy
Subject: Re: Windows 2000: nothing worse
Date: Fri, 31 Mar 2000 07:49:38 +1000


"mlw" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> Christopher Smith wrote:
> >
> > "abraxas" <[EMAIL PROTECTED]> wrote in message
> > news:8c06fp$14b0$[EMAIL PROTECTED]...
> > > > I have real problems classifying NT as Multiplatform. two or three
tops.
> > > > Multiplatform is something like netbsd. Linux, although originally
> > > > developed for the x86, has been proven to be capable of being
> > > > multiplatform. NT has only been ported to alpha, mips, and is Merced
out
> > > > yet? Mips was dropped. The Alpha's port is on shaky ground (is it on
> > > > again or off again this week?)
> >
> > It's somewhat unfair to compare a commercial OS (that has to be
profitable)
> > to free OSes like that.
>
> Why? If they claim it is portable, then porting it to other platforms
> should be part of the business model.

Why ?  Why would you port to a platform that your product wouldn't sell on ?
How could you _justify_ a port to a platform that isn't really interested ?

> If the OS is not portable, stop
> calling it portable.

Portable does not imply ported.

> You mean to tell me that NT on a power PC wouldn't be an interresting
> market?

Interesting, sure.  Successful ?  Highly doubtful.  The only volume OEM of
affordable PPC based hardware is Apple, and we all know how well they react
to potential competitors.

> How about NT on Sun hardware?

You don't think Sun would have anything to say about that ?

> These are targets that should make
> marketing people drool.

What on earth for ?  No one using SOlaris on Sun is going to have the
slightest interest in NT.  Ditto $UNIX on PPC.

> Maybe not for sales of that product, but for the
> argument that NT is portable and therefore a viable option.

Eh ?  How does being ported to a dozen platforms suddenly make it a "viable
option" ?

> As it is now, portability of NT is, at best, a joke.

How would you know ?  You know nothing of itsportability, merely of its
ports.  Not many commercial OSes run on more than one or two platforms.

> > > Its off again.
> > >
> > > They also dropped a PPC port a few years ago.  Just when it was
looking
> > > quite nice actually.
> >
> > Quite nice ?  Just who was making machines that could run it ?  How
about
> > software ?
>
> Actually SGI was using mips. It was a political move more than anything.

That has what, precisely, to do with PPC ?

> > > Go figure.
> > >
> > > I guess x86 hardware is "superior" for some reason.
> >
> > In some ways.  It's common, cheap and fast, relatively speaking.
>
> x86 hardware is cheap, that is, more or less, its only advantage.

And ubiquitous (which is an advantage from the "it has to sell"
perspective).



------------------------------

From: "horst" <[EMAIL PROTECTED]>
Subject: xfs is out!
Date: Thu, 30 Mar 2000 21:28:06 GMT

Hey, sgi's xfs is finally out there.
That makes three journaling file systems in the works, jfs, xfs and
reiserfs.

That's good news!



------------------------------

Date: Thu, 30 Mar 2000 17:00:02 -0500
From: Gary Hallock <[EMAIL PROTECTED]>
Crossposted-To: comp.os.ms-windows.nt.advocacy
Subject: Re: Windows 2000: nothing worse

Christopher Smith wrote:

> What on earth for ?  No one using SOlaris on Sun is going to have the
> slightest interest in NT.  Ditto $UNIX on PPC.
>

I think you mean $UNIX on Apple, not PPC.   Unix on PPC exists and is popular -
AIX.   I used to have a 4-way SMP PPC (604e) running AIX on my desk at work
until they upgraded me to a Power 3 machine.

Gary



------------------------------

From: mlw <[EMAIL PROTECTED]>
Crossposted-To: comp.os.ms-windows.nt.advocacy
Subject: Re: Windows 2000: nothing worse
Date: Thu, 30 Mar 2000 17:05:24 -0500

Christopher Smith wrote:
> 
> "mlw" <[EMAIL PROTECTED]> wrote in message
> news:[EMAIL PROTECTED]...
> > Christopher Smith wrote:
> > >
> > > "abraxas" <[EMAIL PROTECTED]> wrote in message
> > > news:8c06fp$14b0$[EMAIL PROTECTED]...
> > > > > I have real problems classifying NT as Multiplatform. two or three
> tops.
> > > > > Multiplatform is something like netbsd. Linux, although originally
> > > > > developed for the x86, has been proven to be capable of being
> > > > > multiplatform. NT has only been ported to alpha, mips, and is Merced
> out
> > > > > yet? Mips was dropped. The Alpha's port is on shaky ground (is it on
> > > > > again or off again this week?)
> > >
> > > It's somewhat unfair to compare a commercial OS (that has to be
> profitable)
> > > to free OSes like that.
> >
> > Why? If they claim it is portable, then porting it to other platforms
> > should be part of the business model.
> 
> Why ?  Why would you port to a platform that your product wouldn't sell on ?
> How could you _justify_ a port to a platform that isn't really interested ?
> 
> > If the OS is not portable, stop
> > calling it portable.
> 
> Portable does not imply ported.

Portability is a claim that must be proven. If you can't prove it, you
ought not claim portability.

> 
> > You mean to tell me that NT on a power PC wouldn't be an interresting
> > market?
> 
> Interesting, sure.  Successful ?  Highly doubtful.  The only volume OEM of
> affordable PPC based hardware is Apple, and we all know how well they react
> to potential competitors.

Why wouldn't Apple like this to sell hardware?

> 
> > How about NT on Sun hardware?
> 
> You don't think Sun would have anything to say about that ?

No, of course not. Sun sells hardware.

> 
> > These are targets that should make
> > marketing people drool.
> 
> What on earth for ?  No one using SOlaris on Sun is going to have the
> slightest interest in NT.  Ditto $UNIX on PPC.

Yes, but I/O bandwidth on a Sun blows away the PC. NT on Sun would be
about as fast as NT could get.

> 
> > Maybe not for sales of that product, but for the
> > argument that NT is portable and therefore a viable option.
> 
> Eh ?  How does being ported to a dozen platforms suddenly make it a "viable
> option" ?

Because it would be proven to be portable. By being portable, one type
of scalability can be just better hardware.

> 
> > As it is now, portability of NT is, at best, a joke.
> 
> How would you know ?  You know nothing of itsportability, merely of its
> ports.  Not many commercial OSes run on more than one or two platforms.

Having worked on NT, as a kernel developer, since the original beta, I
think I have a bit of knowledge in this area. Looking at many of the MS
written drivers, I can honestly say, IMHO, that while NT aims to be
portable, many of the big ticket drivers, such as video and networking,
are not coded in a very portable fashion.

While they did try to be good about wrapping things in macros and
abstracted APIs, they seem to have a hard time breaking away from the
architectural paradigms of the PC. A lot of the things done in NT will
not port very well to dissimilar architectures. Some machines have
explicit I/O, some do not. Endian-ness, is also a problem. 

> 
> > > > Its off again.
> > > >
> > > > They also dropped a PPC port a few years ago.  Just when it was
> looking
> > > > quite nice actually.
> > >
> > > Quite nice ?  Just who was making machines that could run it ?  How
> about
> > > software ?
> >
> > Actually SGI was using mips. It was a political move more than anything.
> 
> That has what, precisely, to do with PPC ?

Nothing, it has to do with Mips being canceled.


-- 
Mohawk Software
Windows 9x, Windows NT, UNIX, Linux. Applications, drivers, support. 
Visit http://www.mohawksoft.com
"We've got a blind date with destiny, and it looks like she ordered the
lobster"

------------------------------

From: [EMAIL PROTECTED] (JEDIDIAH)
Crossposted-To: comp.os.ms-windows.nt.advocacy
Subject: Re: Windows 2000: nothing worse
Date: Thu, 30 Mar 2000 22:23:22 GMT

On Fri, 31 Mar 2000 07:49:38 +1000, Christopher Smith <[EMAIL PROTECTED]> wrote:
>
>"mlw" <[EMAIL PROTECTED]> wrote in message
>news:[EMAIL PROTECTED]...
>> Christopher Smith wrote:
>> >
>> > "abraxas" <[EMAIL PROTECTED]> wrote in message
>> > news:8c06fp$14b0$[EMAIL PROTECTED]...
>> > > > I have real problems classifying NT as Multiplatform. two or three
>tops.
>> > > > Multiplatform is something like netbsd. Linux, although originally
>> > > > developed for the x86, has been proven to be capable of being
>> > > > multiplatform. NT has only been ported to alpha, mips, and is Merced
>out
>> > > > yet? Mips was dropped. The Alpha's port is on shaky ground (is it on
>> > > > again or off again this week?)
>> >
>> > It's somewhat unfair to compare a commercial OS (that has to be
>profitable)
>> > to free OSes like that.
>>
>> Why? If they claim it is portable, then porting it to other platforms
>> should be part of the business model.
>
>Why ?  Why would you port to a platform that your product wouldn't sell on ?
>How could you _justify_ a port to a platform that isn't really interested ?

        If it's genuinely portable, then it won't be a problem. Even without
        the expectation of any sales, it shouldn't be prohibitively costly.
        If it's genuinely portable, one of drones in the tech support call
        center could have ported it for fun in their off hours already.

[deletia]

-- 

        It is not the advocates of free love and software
        that are the communists here , but rather those that        |||
        advocate or perpetuate the necessity of only using         / | \
        one option among many, like in some regime where
        product choice is a thing only seen in museums.
        
                                      Need sane PPP docs? Try penguin.lvcm.com.

------------------------------

From: "Christopher Smith" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.ms-windows.nt.advocacy
Subject: Re: Windows 2000: nothing worse
Date: Fri, 31 Mar 2000 08:26:23 +1000


"Gary Hallock" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> Christopher Smith wrote:
>
> > What on earth for ?  No one using SOlaris on Sun is going to have the
> > slightest interest in NT.  Ditto $UNIX on PPC.
> >
>
> I think you mean $UNIX on Apple, not PPC.   Unix on PPC exists and is
popular -
> AIX.   I used to have a 4-way SMP PPC (604e) running AIX on my desk at
work
> until they upgraded me to a Power 3 machine.

What I was saying is that no-one[1] using Unix on a PPC machine is going to
be interested in plonking NT onto that machine, even if they could.

[1] "No-one" being a somewhat exxaggerated way of saying "stuff all people".



------------------------------

From: <[EMAIL PROTECTED]>
Crossposted-To: comp.os.ms-windows.nt.advocacy
Subject: Re: Why did we even need NT in the first place?
Date: Thu, 30 Mar 2000 14:28:33 -0800

>Heh, how lame.


I accept your condition, but I am willing to overlook your tendancy
to illogic and "lameness".

>The problem is he's asking something that he knows is not the way
>NT does things. In effect, he's creating a strawman.


When you think your back is against the wall you change the
subject.  The only strawman in this branch of this thread is your
tendancy to cry strawman.  I am not employing any disinformation
tatics of any kind including strawmen, I hope you are not consiously
using the disinformation tatic of claiming that disinformation tatics
are being used on you and thereby distracting attention away from
a failure or short comming of you position.

>With Un*x, you would do it that way, but with NT, you would do it
>differently, but they're both means to an end, and both have their
>pros and cons.

If it can be done on NT, differently or not, that proves that I was
not using strawmen.  You destory your own foolish claims yourself.

But I didn't say that you had to do it THAT way with NT.  I was
requesting you to provide the method that you would do it in NT.

Since you don't seem to be able to provide a workable respond to
the problems at hand using Windows and since you keep claiming
that I was presenting the sample problems as strawmen.  I will offer
a solution for these problems within the Windows computing
environment.  If my solutions are not as efficient as what you would
consider fair, then please provide us with the more efficient solutions.

First the the ordinary user problem:

1.  Using your text editor of choice create a dos batch file with the
      following contents:

    gzip -dc s:\share\corporate\documents\* s:\share\dictionary\* >fileone
    words <fileone | sort | uniq >s:\share\dictionary\corporate\all.words
    del fileone

2. Open a dos shell and execute the the dos batch file.

3. When the batch file is finished close the dos shell.

4. Write an email to [EMAIL PROTECTED] and attach a copy of
    s:\share\dictionary\corporate\all.words to the email.

Note this solutions assumes:  that drive S: is mapped to a network
shared directory that serves that same purpose as unix's /usr/share;
that you have assess to the dos versions of uniq.exe, words.exe and
the dos port of GNU sort.exe and GNU gzip.exe.  Each of these programs
are readily available free on the net.  GNU's sort is required because
the dos/windows sort program may not be able to handle enough data
for this problem.

Now for the sysadmin problem:

Use a disk partition cloning software to store a mirror image copy of the
partition on to offline media.  Also make a backup copy of the partition
onto tape for your archives.  Send the offline media via a courier to the
remote site on the other side of the nation and arrange to have someone
at that location, use another copy of the same cloning software to copy
the partition from the offline media onto the target computer's harddrive.

A variation of the procedure would be for you to transport the offline
media to the remote computer's location and install the partition yourself.
If the offline media is a tape then you would not have to make the backup
copy on tape you would already have it.

You see?   I was not using strawmen, the job can be done.  Perhaps
you really couldn't come up with these solutions, I would strongly doubt
you compentence with NT as well as with unix.

If you think I presented solutions that were not efficient enough for the
NT environment, please present the more efficient solutions.

>If you wish to continue debating strawmen, please continue doing so
>in COLA, where it's most common and appropriate.

>
>If you want to debate about the merits of the various OS, let's talk
>in terms of how each one accomplishes different tasks.


Different tasks? Comparing one task an apple to a different task
an orange is what you are proposing?  The only valid way to evaluate
the comprable merrits of different operating systems would be to
examine the solutions that each one supports to solve same various
problems.  The same problem would have to be addressed as best
as is possible with each operating system.  Anything would be invalid
and dishonest.




------------------------------


** FOR YOUR REFERENCE **

The service address, to which questions about the list itself and requests
to be added to or deleted from it should be directed, is:

    Internet: [EMAIL PROTECTED]

You can send mail to the entire list (and comp.os.linux.advocacy) via:

    Internet: [EMAIL PROTECTED]

Linux may be obtained via one of these FTP sites:
    ftp.funet.fi                                pub/Linux
    tsx-11.mit.edu                              pub/linux
    sunsite.unc.edu                             pub/Linux

End of Linux-Advocacy Digest
******************************

Reply via email to