Linux-Advocacy Digest #634, Volume #26 Mon, 22 May 00 09:13:07 EDT
Contents:
Re: There is NO reason to use Linux...It just STINX (Donal K. Fellows)
Re: There is NO reason to use Linux...It just STINX (Martijn Bruns)
Re: Microsoft W2K lack of goals. (Perry Pip)
Re: Things Linux can't do! (Perry Pip)
W2K BSOD's documented *not* to be hardware (Was: lack of goals. (Perry Pip)
Re: The future... (Full Name)
Re: Microsoft W2K lack of goals. ("Erik Funkenbusch")
Re: Need ideas for university funded project for linux (Robie Basak)
Re: Top 10 Reasons to use Linux (Darren Winsper)
Re: Desktop use, office apps (Darren Winsper)
Re: Microsoft W2K lack of goals. ("Erik Funkenbusch")
Re: Linux fails - again (Nathaniel Jay Lee)
----------------------------------------------------------------------------
From: [EMAIL PROTECTED] (Donal K. Fellows)
Subject: Re: There is NO reason to use Linux...It just STINX
Date: 22 May 2000 12:03:37 GMT
In article <[EMAIL PROTECTED]>,
Pete Goodwin <[EMAIL PROTECTED]> wrote:
> In America I guess. Here in the UK we still have local charges for
> telephone calls (nothing is free yet) and a max. of 56K modems. ADSL
> and Cable modems aren't here yet.
Not if you're with the right ISP/Telco. <ultrasmug>
Donal.
--
Donal K. Fellows http://www.cs.man.ac.uk/~fellowsd/ [EMAIL PROTECTED]
-- I may seem more arrogant, but I think that's just because you didn't
realize how arrogant I was before. :^)
-- Jeffrey Hobbs <[EMAIL PROTECTED]>
------------------------------
From: Martijn Bruns <[EMAIL PROTECTED]>
Subject: Re: There is NO reason to use Linux...It just STINX
Date: Mon, 22 May 2000 14:39:28 +0200
"Donal K. Fellows" schreef:
>
> In article <[EMAIL PROTECTED]>,
> Pete Goodwin <[EMAIL PROTECTED]> wrote:
> > In America I guess. Here in the UK we still have local charges for
> > telephone calls (nothing is free yet) and a max. of 56K modems. ADSL
> > and Cable modems aren't here yet.
>
> Not if you're with the right ISP/Telco. <ultrasmug>
Hey, that's odd.
I thought i had heard someone say the UK had fiberglass cable and
cable-modemconnections already. It could be he (not the poster)
was lying. :-)
------------------------------
From: [EMAIL PROTECTED] (Perry Pip)
Crossposted-To: comp.os.ms-windows.advocacy,comp.os.ms-windows.nt.advocacy
Subject: Re: Microsoft W2K lack of goals.
Reply-To: [EMAIL PROTECTED]
Date: Mon, 22 May 2000 12:37:07 GMT
On 21 May 2000 06:53:43 GMT,
abraxas <[EMAIL PROTECTED]> wrote:
>In comp.os.linux.advocacy Stephen S. Edwards II <[EMAIL PROTECTED]> wrote:
>
>> I'll say. Windows2000 DataCenter Server can access up to 64GB of
>> RAM, and 16 processors.
>
>So they tell us. Whens that gonna be done again?
>
>> Throw clustering in there, and you have
>> a very powerful and cost-effective high-end solution. Not bad at
>> all for a boxed OS.
>
>If it ever gets released.
>
Of course today as we speek there already is Cplant at Sandia. 628
64-bit alphaprocessors, running Redhat 6:
http://www.cs.sandia.gov/cplant/
which is, btw, #44 on the top 500 list at:
http://www.top500.org/lists/TOP500List.php3?Y=1999&M=11
Yet their is not one NT machine on the top 500. Most are Solaris,
IRIX, AIX and HPUX. NT just can't scale to the big leagues.
Perry
------------------------------
From: [EMAIL PROTECTED] (Perry Pip)
Crossposted-To: comp.os.ms-windows.advocacy,comp.os.ms-windows.nt.advocacy
Subject: Re: Things Linux can't do!
Reply-To: [EMAIL PROTECTED]
Date: Mon, 22 May 2000 12:37:17 GMT
On Sat, 20 May 2000 18:02:37 +1000,
Christopher Smith <[EMAIL PROTECTED]> wrote:
>
>"Perry Pip" <[EMAIL PROTECTED]> wrote in message
>news:[EMAIL PROTECTED]...
>> On Sat, 20 May 2000 02:54:01 +1000,
>> Christopher Smith <[EMAIL PROTECTED]> wrote:
>> >
>> >I've been back up the thread. The context was, IIRC, "problems", and the
>> >specific sentence about BSODs.
>>
>> You really are getting desparate...here it is going back seven posts
>> and their is no use of the word "problems" untill Stephen's use of it,
>> just an ongoing talk about BSOD's and lockups under X:
>>
>> http://x28.deja.com/getdoc.xp?AN=623126285 - reference to "Blue screen"
>>
>> http://x28.deja.com/getdoc.xp?AN=623143491 - reference to Linux crashes.
>>
>> http://x28.deja.com/getdoc.xp?AN=623268665 - reference to "blue screen"
>>
>> http://x28.deja.com/getdoc.xp?AN=623294410 - reference by Stephen to
>> X locking up.
>>
>> http://x28.deja.com/getdoc.xp?AN=623126285 - reference to X locking up.
>
>The above are totally irrelevant. Especially as setting the context since
>it changes from apps crashign to apps bringing down the system to BSODs.
>
>> http://x28.deja.com/getdoc.xp?AN=623637730 - Stephen's original claim
>> "I have seen a lot of BSODs in my time, and in every single
>> instance..."
>
>Thus specifically indicating the BSODs he had experienced.
And thus the context of the discussion: BSODs.
>> http://x28.deja.com/getdoc.xp?AN=623697578 - someone wrote "MS doesn't
>> aggree with you stephen. In a survey _they published_, 40 percent of
>> BSODs were attributed to 'internal NT components'. Hardware, apps,
>> etc. got the rest."
>
>Once again, irrelevant to the point at hand.
Huh?? The context of this post was also BSOD's. The context of the
discussion has not changes.
>> http://x28.deja.com/getdoc.xp?AN=623940112 - Stephen's story changed
>> to "most of the problems I've encountered, and witnessed..."
>
>Thus changing the context.
Wrong. "Problems" is a generic term whose meaning is defined by the
context. The context of the last two posts was about BSOD's and in
direct response to the last post Stephen used "problems" in that
context. So in that context: "problems" == "BSODs". That's common
everyday English language. My apologies if I have erronously assumed
that English is your first language.
>> Now that's five looks you've had "Christopher" and you're still
> ^^^^^^^^^^^^^
>I'm wondering why that's in ""s.
Keep wondering.....
>> playing your little game of evasion. I've provided the evidence, it's
>> up to you to have some reading comprehension.
>
>I have read, comprehended, and your claim of Stephen's lying is, at best,
>tenuous.
That's your (heavily biased) opinion and you are entitled to it.
------------------------------
From: [EMAIL PROTECTED] (Perry Pip)
Crossposted-To: comp.os.ms-windows.advocacy,comp.os.ms-windows.nt.advocacy
Subject: W2K BSOD's documented *not* to be hardware (Was: lack of goals.
Reply-To: [EMAIL PROTECTED]
Date: Mon, 22 May 2000 12:39:50 GMT
On Sat, 20 May 2000 20:31:18 -0500,
Erik Funkenbusch <[EMAIL PROTECTED]> wrote:
>Charlie Ebert <[EMAIL PROTECTED]> wrote in message
>news:[EMAIL PROTECTED]...
>> W2K took something like 1.5 to 2 years
>> over-run to get out the door.
>
>Yes, similar to Linux Kernel 2.4's being a year overdue.
Could you provide me some documention that 2.4 is a year overdue. That
would mean is was due last May.
>> Also reported goals were to eliminate the
>> problem of blue screening as well as
>> maintain performance levels in multitasking.
>
>No. It's impossible to eliminate such things, just like it's impossible to
>eliminate kernel panics from Linux. If you could eliminate them, they
>wouldn't need to be there.
>
>Things like faulty hardware and to some extent background radiation *WILL*
>cause glitches on non-military spec systems.
This is an evasion, Eric. Yes, faulty hardware and background
radiation can crash an OS. However I don't live across the street from
Three Mile Island and faulty hardware can always be replaced. There
there are other ways an OS can crash.
Here is an example where W2K can BSOD by allowing an application to
misuse resources:
http://support.microsoft.com/support/kb/articles/Q195/8/57.ASP?LN=EN-US&SD=gn&F
R=0
Note in the above URL under resolution it says: "To resolve this
problem, the application has to be modified to ..." So MS is clearly
blaming the problem on the Application, not the OS . Here is another
example:
http://support.microsoft.com/support/kb/articles/Q245/1/12.ASP?LN=EN-US&SD=gn&F
R=0
Agian, MS blames the problem on the application vendor. In the
following example, a bad network application can BSOD W2K, but
at least MS takes responsibility:
http://support.microsoft.com/support/kb/articles/q260/9/56.ASP?LN=EN-US&SD=gn&F
R=0
Here are some additional examples of OS bugs acknowledged by MS to
cause BSOD's:
http://support.microsoft.com/support/kb/articles/q257/8/13.asp?LN=EN-US&SD=gn&F
R=0
http://support.microsoft.com/support/kb/articles/Q232/9/48.ASP?LN=EN-US&SD=gn&F
R=0
http://support.microsoft.com/support/kb/articles/Q254/6/11.ASP?LN=EN-US&SD=gn&F
R=0
And then there is the whole issue of bad device drivers. For one
thing, it is not necessary for most device drivers to have access to
kernel address space. The proper way to avoid that is to use a
microkerenel architecture, which neither Linux or NT does. There are
some areas though, where Linux does a better job of keeping device
drivers out of the kernel. For example, in the following case an HP
printer driver BSOD's W2K when printing to a network printer:
http://support.microsoft.com/support/kb/articles/Q199/1/18.ASP?LN=EN-US&SD=gn&F
R=0
It is extremely unlikely this will ever happen to Linux, becuase the
"printer drivers" are filters provided by ghostscript, a process
called by lpd, neither of which has access to kernel address
space. The worst a bad printer filter can do is send garbage to the
printer or tie up some resources untill the admin kill -9's it. It is
utterly unecessary to give a printer driver, which is nothing more
than a data filter, access to kernel address space.
Similarily, we all know that because W2K has GDI in the kernel,
bad video drivers can also cause a BSOD. In contract, an X server on
linux does not have access to kernel address space, and so X crashes
only lock up the console. Of course, if you are running a desktop, you
work is lost anyways so it doesn't matter. If you are running a server
however, it makes all the difference.
And in the case where device drivers have access to kernel address
space, they are effectively part of the OS and should be treated as
such. Nearly all open source Linux drivers are distributred with the
kernel and developed under the same open source model as the kernel
itself. I have never had a problem with a device driver distributed
with the kernel. Will MS has started signing drivers. alot of stuff is
not supported in that way.
It is no wonder with these design issues, we are constantly hearing
more BSOD stories than 'kernel panic' issues. And I have used Linux
long enough to see a few unexplainable kernel panics, at least under
1.0, 1.2, and 2.0, but none so far under 2.2 (fingers crossed).
>> totally failed at goal #2.
>
>Only because you are listening to "reported goals" by people who don't know
>what they're talking about.
Microsft's goal is to make W2K as stable as UNIX. IMHO the jury is
still out on that, but the initial evidence does not look good.
>It's (W2K is) significantly faster than NT if it has it's minimum
>requirements. 10% faster on average.
Proof please!!
Perry
------------------------------
From: [EMAIL PROTECTED] (Full Name)
Subject: Re: The future...
Date: Mon, 22 May 2000 12:50:05 GMT
On Thu, 18 May 2000 21:57:59 -0400, mlw <[EMAIL PROTECTED]> wrote:
>
>The Server market distinct from the Workstation is gone. Desktop PCs
>will either get smaller in the direction of thin-clients, or be
>indistinguishable from servers.
>
Users want thick clients. They more or less demand them - it makes
them feel secure.
I don't know if you've notice but the world has moved away from thin
clients over the last 20 years. The only way this will change is if
thin clients appear more or less identical to thick clients from the
user perspective.
Even Linux users prefer thick clients. That's why they keep bleating
about Linux as a desktop OS. They are too stupid to realise the power
of a Unix system is its ability to support thin clients.
>I think the NOS market is gone. Novel and whom ever is pursuing it is
>wasting their time. All real OS's will just do it right.
>
"A real OS" - this is a meaningless statement.
>Windows is going to die. Not because of MS, exactly, but because the
>world is going towards standards. While UNIX is not a majority player,
>it is a standards based multi-vendor platform. MS will bluster about
>being the "defacto-standard" but more and more IT people are realizing
>that public standards are better than ubiquitous proprietary standards.
>
Standards in Unix?
Where is the standard location to install third party software?
Where is the standard location for the shadow password file?
What is the standard GIU for a Unix box?
What is the standard name of the mounted file system table file?
Why is there a /usr/etc and a /etc on many Unix installs? Why is
/usr/etc sometimes a symbolic link to /etc?
Were is the standard location to keep users files on a Unix system?
What is the standard shell on a Unix system? Why is it traditional to
write scripts in Bourne shell?
Where is the standard place to assign the path variable for sh?
cshell? bash? korn shell? trusted cshell?
Why does Oracle 7 place it's listener configuration files in
/var/adm/oracle? What is the difference between /usr/adm and
/var/adm? Why is /usr/adm sometimes a symbolic link to /var/adm?
Why is it when I telnet from one Unix system to another I have to
issue a 'stty erase' command so I can delete characters?
Why is /usr/boot a symbolic link to /usr/kvm/boot on some Unix
systems?
Why does the SunOS box I just logged into have 1741 symbolic links in
/etc, /usr and /bin?
Why was my password just broadcast in clear text on at least two
subnets?
What is the standard name of the kernel executable on a Unix box?
Why is Solaris so different to SunOS?
Why does Linux have so many distributions? What is the 'standard'
Linux distribution?
How old are you? Have you ever actually managed any real (sic) Unix
operating systems before? Is installing Red Hat on your PC at home
the only Unix experience you've had?
------------------------------
From: "Erik Funkenbusch" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.ms-windows.advocacy,comp.os.ms-windows.nt.advocacy
Subject: Re: Microsoft W2K lack of goals.
Date: Mon, 22 May 2000 08:04:04 -0500
David Goldstein <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> Every report that I have read on the speed of W2K vs NT4 is that NT4
> is faster. Perhaps that is on identical hardwarem, since W2K consumes
> even more resources than NT4 does, it would stand to reason that I need
> to do a major hardware upgrade to see the improvements; of course, NT4
> would run that much faster, too :)
Guess you don't read many reports. For instance:
http://www.zdnet.com/pcmag/stories/reviews/0,6755,2426071,00.html
"Windows 2000 Server's performance surpasses that of Windows NT 4 Server.
Our benchmark tests showed across-the-board improvement for several
network-server functions including file and print, application service, and
serving up Web pages. Architectural enhancements in network communication,
multiprocessor scaling, and file-system components account for the overall
boost."
Then throw in the fact that Windows 2000 has been setting benchmark records
in places like www.tpc.org
Perhaps you can post some URL's to reviews that show NT4 being significantly
faster.
> > Depends what you mean by improvements. Fadout menus (can Linux do
that?),
> > fadout Windows (now part of the API).
>
> Wow, a major innovation on MS' part. Fadeout menus! What will they
> think of next?
There's a lot to be said for ergonomic engineering.
------------------------------
From: [EMAIL PROTECTED] (Robie Basak)
Crossposted-To:
comp.os.linux.development.apps,comp.os.linux.development.system,comp.os.linux.misc,comp.os.linux.setup
Subject: Re: Need ideas for university funded project for linux
Reply-To: [EMAIL PROTECTED]
Date: 17 May 2000 22:59:17 GMT
On Wed, 17 May 2000 16:52:41 GMT, eyez said:
>quoting <[EMAIL PROTECTED]>:
>[snip]
>>Mongoose <[EMAIL PROTECTED]> writes:
>>[snip]
>>can put my programming hours where my mouth is. (Reason I'm not using
>>it now? The fucking Aureal Vortex 2 drivers are (a) non-free; and (b)
>>unusably poor.)
Do you mean non-free as in beer non-free or RMS non-free? If
you mean the former, check out http://linux.aureal.com
They're not brilliant, but the sound in Quake 3 works, and mp3s play
fine :-)
Robie.
--
------------------------------
From: [EMAIL PROTECTED] (Darren Winsper)
Subject: Re: Top 10 Reasons to use Linux
Date: 18 May 2000 06:07:30 GMT
On Wed, 17 May 2000 11:01:49 GMT, Full Name <[EMAIL PROTECTED]> wrote:
> 9. You have no friends and no life, so spending all day building
> kernels is actually a step up.
Isn't this ironic coming from someone who has nothing better to do
than troll COLA. What's wrong, have you run out of valid reasons to
bash Linux?
--
Darren Winsper (El Capitano) - ICQ #8899775
Stellar Legacy project member - http://www.stellarlegacy.tsx.org
DVD boycotts. Are you doing your bit?
This message was typed before a live studio audience.
------------------------------
From: [EMAIL PROTECTED] (Darren Winsper)
Subject: Re: Desktop use, office apps
Date: 18 May 2000 06:07:36 GMT
On Tue, 16 May 2000 21:42:10 GMT, R. Christopher Harshman
<[EMAIL PROTECTED]> wrote:
> About 4.5MB/sec sustained, according to benchmarks.
Ouch, that is slow:
stargazer:~# hdparm -t /dev/hdc
/dev/hdc:
Timing buffered disk reads: 64 MB in 6.23 seconds = 10.27 MB/sec
stargazer:~# hdparm -t /dev/hda
/dev/hda:
Timing buffered disk reads: 64 MB in 6.18 seconds = 10.36 MB/sec
(Those are sustainable speeds without the filesystem overhead)
/dev/hda is a mid-range 7200rpm IDE drive. /dev/hdc is an el-cheapo
5200rpm drive from 1998. If those can sustain 10MB/sec then your
drives must either be really old or misconfigured.
I have this sort of command in one of my bootup scripts:
hdparm -q -k 1 -c 3 -m 32 /dev/hdc
man hdparm for more info.
--
Darren Winsper (El Capitano) - ICQ #8899775
Stellar Legacy project member - http://www.stellarlegacy.tsx.org
DVD boycotts. Are you doing your bit?
This message was typed before a live studio audience.
------------------------------
From: "Erik Funkenbusch" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.ms-windows.advocacy,comp.os.ms-windows.nt.advocacy
Subject: Re: Microsoft W2K lack of goals.
Date: Mon, 22 May 2000 08:10:15 -0500
Perry Pip <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> Of course today as we speek there already is Cplant at Sandia. 628
> 64-bit alphaprocessors, running Redhat 6:
>
> http://www.cs.sandia.gov/cplant/
>
> which is, btw, #44 on the top 500 list at:
>
> http://www.top500.org/lists/TOP500List.php3?Y=1999&M=11
>
> Yet their is not one NT machine on the top 500. Most are Solaris,
> IRIX, AIX and HPUX. NT just can't scale to the big leagues.
NT had scalability limits, like most OS's. For instance, Solaris has a
maximum of 64 processors IIRC. NT was limited to in commercial designs.
There have been custom HAL's written for NT 4 that gave it much better
scaleability.
Windows 2000 has much fewer limitations and has commercial products capable
of up to 32 processors. I imagine that 64 bit NT will be able to scale even
further.
------------------------------
From: Nathaniel Jay Lee <[EMAIL PROTECTED]>
Subject: Re: Linux fails - again
Date: Mon, 22 May 2000 08:05:35 -0500
mlw wrote:
>
> Full Name wrote:
> [long ridiculous story snipped]
>
> First of all the story you tell indicates one thing, you must have put
> an nfs drive in fstab. So when the Linux box was booting, it was looking
> for the nfs server. (Which, by your own account, was not operational at
> the time) The one line that makes this clear is: "It then came up OK
> since the Ultra was at that stage fully functional."
>
> When you put an entry in fstab, and direct that it should be mounted as
> a local file system, the machine is configured to treat that file system
> as critical and not boot without it. Booting without it would have be an
> error. Linux treated the situation EXACTLY as you had configured it too.
>
> If you want to treat the file system as "mount on demand" take a look at
> autofs.
> --
> Mohawk Software
> Windows 9x, Windows NT, UNIX, Linux. Applications, drivers, support.
> Visit http://www.mohawksoft.com
> Have you noticed the way people's intelligence capabilities decline
> sharply the minute they start waving guns around?
The other thing to look into is the way some distributions treat NFS
mounts that are in fstab. Some of them use a script to look for NFS
mounts in fstab and if they don't come up in a certain time period
(configurable, default of 2 minutes I believe on SuSE) the machine
continues to boot). This is the method that all of my machines use.
Yes, the 2 minute wait is annoying, but it does complete the boot-up
eventually.
Nathaniel Jay Lee
[EMAIL PROTECTED]
------------------------------
** 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
******************************
