Linux-Advocacy Digest #511, Volume #32 Tue, 27 Feb 01 01:13:03 EST
Contents:
Re: NT vs *nix performance ("Erik Funkenbusch")
Re: NT vs *nix performance (J Sloan)
Re: NT vs *nix performance (Amphetamine Bob)
Re: NT vs *nix performance (J Sloan)
Re: SSH vulnerabilities - still waiting [ was Interesting article ] (J Sloan)
Re: NT vs *nix performance (J Sloan)
Re: Information wants to be free, Revisited (Trevor Zion Bauknight)
Re: Is this odd security behaviour by MS? ("Adam Warner")
Re: Mircosoft Tax (Donovan Rebbechi)
Re: Mircosoft Tax (Donovan Rebbechi)
Re: [OT] .sig (Craig Kelley)
Re: NT vs *nix performance (Craig Kelley)
Re: The Windows guy. (Craig Kelley)
Re: Mircosoft Tax (Tim Hanson)
Redhat's CEO wants the Open Source Comm. to address the US Senate ("Adam Warner")
----------------------------------------------------------------------------
From: "Erik Funkenbusch" <[EMAIL PROTECTED]>
Crossposted-To:
alt.destroy.microsoft,alt.linux.sux,comp.os.ms-windows.advocacy,comp.os.ms-windows.nt.advocacy
Subject: Re: NT vs *nix performance
Date: Mon, 26 Feb 2001 22:51:19 -0600
"Chris Ahlstrom" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> Erik Funkenbusch wrote:
> >
> > There are 3 single-tasking graphic servers that run a web server called
> > "boa" under single-user mode FreeBSD. This gives them the ability to
simply
> > server HTTP graphic files (which are completely static and don't require
any
> > multitasking) very fast.
> >
> > Although, it appears that they're starting to phase even these out.
Check
> > out:
> >
> >
http://uptime.netcraft.com/up/graph?site=64.4.18.24&submit.x=72&submit.y=11
> >
> > Then look at the history, you'll see it wobbles between Win2k and
FreeBSD.
> > They may have even completely replaced it with Win2k, since the last
record
> > of a change was a few weeks ago.
>
> OS, Web Server and Hosting History for 64.4.18.24
>
> OS Server Last changed IP address
Netblock
> Owner
> Windows 2000 Microsoft-IIS/5.0 13-Feb-2001 64.4.18.24
MS Hotmail
> FreeBSD Boa/0.93.17.3 12-Feb-2001 64.4.18.24
MS Hotmail
> FreeBSD Microsoft-IIS/5.0 11-Feb-2001 64.4.18.24
MS Hotmail
> FreeBSD Boa/0.93.17.3 10-Feb-2001 64.4.18.24
MS Hotmail
> FreeBSD Microsoft-IIS/5.0 9-Feb-2001 64.4.18.24
MS Hotmail
> Windows 2000 Microsoft-IIS/5.0 7-Feb-2001 64.4.18.24
MS Hotmail
> FreeBSD Boa/0.93.17.3 5-Feb-2001 64.4.18.24
MS Hotmail
> Windows 2000 Boa/0.93.17.3 4-Feb-2001 64.4.18.24
MS Hotmail
>
> It looks to me like Netcraft can't tell what the hell this address is
running.
No, it was load balanced between FreeBSD and Win2k systems. It now appears
to be fully Win2k.
------------------------------
From: J Sloan <[EMAIL PROTECTED]>
Crossposted-To:
alt.destroy.microsoft,alt.linux.sux,comp.os.ms-windows.advocacy,comp.os.ms-windows.nt.advocacy
Subject: Re: NT vs *nix performance
Date: Tue, 27 Feb 2001 04:42:58 GMT
Jon Johanson wrote:
> "Rex Ballard" <[EMAIL PROTECTED]> wrote in message
>
> > Microsoft will not publish industry standard benchmarks between
> > comparably equipped
> > Microsoft Windows systems and Linux systems.
>
> Perhaps because there are none And why hasn't any linux distributor ever
> done the same? Hmmm? You'd think Red Hat would send their systems out to be
> independantly tested and then tout these impressive wins to everyone.
Actually, IBM and Dell have done so, it's called the
specweb benchmarks - also, SAP has done some
similar testing in fact a Linux/Solaris combination holds
the new SAP benchmark record IIRC.
However, 2.4-based distros are just now beginning to
ship, so give it a few months.
> Gee, and IBM has the money and has done TPC before and yet they don't have a
> benchmark using linux. In fact, they use windows 2K even when running their
> own database.
Yes, IBM supports the windows pc line, it is one of their
supported OSes, but I think you are somewhat deceived
if you think windows is their only concern, Yes, IBM wants
to see windows pcs, there's money in it. However, they
also want to sell mainframes, RS/6000s and AS/400s,
none of which have anything to do with ms windows.
Cheers,
jjs
------------------------------
From: Amphetamine Bob <[EMAIL PROTECTED]>
Reply-To: [EMAIL PROTECTED]
Crossposted-To:
alt.destroy.microsoft,comp.os.ms-windows.advocacy,comp.os.ms-windows.nt.advocacy
Subject: Re: NT vs *nix performance
Date: Mon, 26 Feb 2001 20:52:34 -0800
jjs wrote:
> Jon Johanson wrote:
>
> Gee, and IBM has the money and has done TPC before and yet they
don't have a
> > benchmark using linux. In fact, they use windows 2K even when running their
> > own database.
No, actually IBM has banned Windows 2000 for all internal use! Yes,
it is true!
>
> Yes, IBM supports the windows pc line, it is one of their
> supported OSes, but I think you are somewhat deceived
> if you think windows is their only concern, Yes, IBM wants
> to see windows pcs, there's money in it.
It is interesting to note that the PC Division has been losing $1
billion a year for several years now while OS/2 brings in about a $1
billion a year. Guess which line IBM tries to kill?
However, they
> also want to sell mainframes, RS/6000s and AS/400s,
Darn right they do. This is where IBM really makes a lot of money.
And I mean a lot of money.
--
Bob - shooting the bozo bit at 550 MHZ :). Wheeeeee! ;)
Microsoft "Tech Support".
1) Re-boot
2) Re-boot
3) Re-install all your software
4) Buy the new release (again)
5) Go to 1
------------------------------
From: J Sloan <[EMAIL PROTECTED]>
Crossposted-To:
alt.destroy.microsoft,alt.linux.sux,comp.os.ms-windows.advocacy,comp.os.ms-windows.nt.advocacy
Subject: Re: NT vs *nix performance
Date: Tue, 27 Feb 2001 04:54:34 GMT
Jon Johanson wrote:
> Special cache? You mean some software anyone can buy and run outta the box?
> You didn't notice the AIX box has 12 processors versus 8 for the w2k box
> (and it only beat it by 14% with 50% more processors).
You have to ask yourself why microsoft doesn't just
submit benchmarks for the 32-way system on which
win 2k can ostensibly run. Think about it.
> I did see the custom benchmark buster Tux system beating it by a sizzling 3
> whole percentage points! Woo hoo!!
You've got the chronology all wrong.
Linux blasted into first place in specweb last summer.
In the process is beat all existing windows results by
a wide margin, something like 2:1.
We haven't yet seen the Linux response to the latest
microsoft results, but they will come.
> > I have a very low end server (Compaq Pentium Pro 200)
> > here at the office - I did some quick benchmark tests with
> > apache, mind you - not something really fast like zeus or
> > tux, just plain old apache on Red Hat 7.0,and I get 4200
> > requests per second from that tired old Linux box, over
> > a single100 mbit ethernet interface.
>
> Amazing... but I think you are lying. Why? Well, how do you figure that you,
> joe blow average user,
Well, first of all, I'm hardly "joe blow average user", but...
> using a piece of crap pentium pro scraper can get
> 4200 but it takes IBM a quad Xeon 700 to get 4200? You really expect me to
> believe that you generated 4200 rps on a PP200? ahahahaha
You completely misunderstand the whole concept here.
The thing you have to try and get your mind wrapped around
is something called "units of measurement".
4200 requests per second is not the same as 4200 kb/sec,
nor is it the same as "4200 conforming connections".
Let me put it this way - 17 dogs has nothing to do with
17 days, or 17 dollars - does any of this make sense?
> Where is your gigabit ethernet?
No gigabit, just the 10/100 tlan ethernet card that comes
with compaq servers.
> Did you have 4 nics
Just 1 nic, like I said.
> and 6 10k scsi drives in
> it?
Nope, just the plain old compaq raid controller.
> or maybe your just such a software guru that you figured out how to do
> what all the engineers at IBM couldn't.
Oh Dear, whatever are you talking about?
The test was very simple and quick, as I said all along.
>From another Unix server on the same lan, I typed:
ab http://case/manual/index.html -k -n 8193 -c 128
and the results said, among other things, that the
"Requests per second" figure was 4200.
This is not rocket science.
jjs
------------------------------
From: J Sloan <[EMAIL PROTECTED]>
Subject: Re: SSH vulnerabilities - still waiting [ was Interesting article ]
Date: Tue, 27 Feb 2001 05:01:01 GMT
Shane Phelps wrote:
> The really funny thing about this is that there *is* at least one Windows
> ssh server version. There is a port of the ssh 1,2,x sshd which runs with
> Cygwin (it only needs cygwin.dll, not the entire package). Data Fellows
> *may* have an ssh server for Windows as well (it is US-only if there is one)
That's interesting, somebody finally managed to do it eh?
(well, actually it doesn't really interest me, but what is
intersting is that no windows admin I've ever talked to
knew there was such a thing).
> and yes, the sshd is a very good way to remotely administer large numbers
> of Windows NT boxes (though a little limited),
It's definitely more limited on the pc platform, since the
major advantage of X over ssh is lost.
> and also much better
> than FTP for automated file transfers.
>
> Ssh wasn't designed to be specific to Unix
Of course not, but Unix has always had all the pieces
needed to build ssh and to have it be useful. I suppose
the windows port was bound to happen eventually, but
I'm not sure I see the point....
jjs
------------------------------
From: J Sloan <[EMAIL PROTECTED]>
Crossposted-To:
alt.destroy.microsoft,alt.linux.sux,comp.os.ms-windows.advocacy,comp.os.ms-windows.nt.advocacy
Subject: Re: NT vs *nix performance
Date: Tue, 27 Feb 2001 05:08:32 GMT
Boris Dynin wrote:
> We discussed it on this group already. Linux results are for some strange
> kernel mode web server called Tux which nobody seems to use.
Actually there's no thing strange about it except the
price/performance ratio - and it's a brand spanking
new product, so there's naturally not been an instant
wholesale conversion. Give it a few years.
> This means (to
> me) that Linux results are quite useless.
> As for AIX results, they were achieved on more powerful end expensive
> hardware than IIS test above: 12 CPUs for AIX vs. 8 for W2k, 64GB RAM for
> AIX vs. 32 GB for W2k.
So why doesn't microsoft post some benchmarks with
a 32-way box? think about it.
> I wouldn't be surprised if AIX test configuration did cost several times
> more than W2k.
However the Linux configuration was much cheaper,
and the software cost was zero.
BTW how much do you think I'd pay to obtain
- win2k datacenter server?
- iis 5?
- swc?
> The question is what you were measuring.
I thought that was clearly stated - it was requests per second,
and as I said, it was a quick and dirty test, but the results are
completely repeatable.
> What makes you think that your
> "quick benchmark tests" make any sense at all?
Because the purpose of benchmark programs is to quantify performance.
jjs
------------------------------
From: Trevor Zion Bauknight <[EMAIL PROTECTED]>
Crossposted-To:
comp.sys.mac.advocacy,comp.sys.next.advocacy,comp.os.ms-windows.advocacy
Subject: Re: Information wants to be free, Revisited
Date: Tue, 27 Feb 2001 04:55:01 GMT
Don't anthropomorphize information. It hates that.
Trev
--
"I think Trevor is an idot. Just the kind of robot President CLITton likes.
Supid people!" - Husker Kev
------------------------------
From: "Adam Warner" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.ms-windows.nt.advocacy
Subject: Re: Is this odd security behaviour by MS?
Date: Tue, 27 Feb 2001 18:20:17 +1200
Hi Erik,
> > Now I would be very surprised if event viewer contained hardware
specific
> > code. Does Microsoft have a policy of not providing security patches for
> > Windows 2000 Datacenter Server, instead leaving it up to OEMs?
>
> MS provides the patches to the OEM, but since the OEM provides hardware
> specific patches in a service pack (No, the even log viewer patch isn't
> hardware specific, but other patches in the service pack are) they are the
> sole source for providing patches for those customers.
I've been lucky enough to discover by trying different Datacenter vendors
that Compaq provides some of their Windows 2000 Datacenter downloads
publicly from their web site:
http://www.compaq.com/support/files/datacenter/us/locate/47_1205.html
The most recent Compaq ProLiant 8500 Data Center Support Paq was issued 16
January 2001.
This appears to be a Compaq-specific security advisory:
http://www.compaq.com/solutions/datacenter/service-advisory.html
"Data Center Service Advisory -
CIM web-enabled agent issue
In reference to Compaq Security Advisory, "Compaq web-enabled
management software security vulnerability. Reference SSRT0705."
"CSA SSRT0705 does affect the Compaq Data Center Solution. We are aware
of this impact and are currently conducting a qualification test on the
Compaq patch to verify its stability on the Data Center platform. Compaq
Data Center customers should NOT APPLY the Compaq SoftPaq to their Data
Center systems. We recommend that Data Center customers immediately disable
their web-enabled agents.
"After the Data Center Solutions Team has completed the qualification
tests on the patch, we will post further information and instructions on
http://www.compaq.com/datacenter.
"Please contact your TAM for further information and assistance."
The original advisory is here:
http://www.compaq.com/products/servers/management/agentsecurity.html
The original advisory was posted 9 January 2001.
It's now 27 February 2001. This does indicate (but only indicate) that it
might take longer to obtain security-related patches specifically for
Windows 2000 Datacenter from a specific OEM.
Regards,
Adam
------------------------------
From: [EMAIL PROTECTED] (Donovan Rebbechi)
Crossposted-To: alt.linux.sux,alt.destroy.microsoft
Subject: Re: Mircosoft Tax
Date: 27 Feb 2001 05:26:35 GMT
On Tue, 27 Feb 2001 02:43:45 GMT, Bob Hauck wrote:
>On 26 Feb 2001 03:30:43 GMT, Donovan Rebbechi <[EMAIL PROTECTED]> wrote:
>>You're assuming that there are no other costs. Tell MS's operating system
>>group that they don't have to go to work anymore, because the "initial
>>development" has already been done.
>
>They still have to work, but if MS sells twice as many copies they don't
>have to work twice as hard.
Yes, but if they're doing well, then they can afford to spend more on
development. If they can forecast an increase in sales, then they may
well put more time into development because of it.
--
Donovan Rebbechi * http://pegasus.rutgers.edu/~elflord/ *
elflord at panix dot com
------------------------------
From: [EMAIL PROTECTED] (Donovan Rebbechi)
Crossposted-To: alt.linux.sux,alt.destroy.microsoft
Subject: Re: Mircosoft Tax
Date: 27 Feb 2001 05:27:50 GMT
On Tue, 27 Feb 2001 02:43:43 GMT, Bob Hauck wrote:
>On 26 Feb 2001 00:41:54 GMT, Donovan Rebbechi <[EMAIL PROTECTED]> wrote:
>
>>Why should the operating system go down in price ? Has it
>>become cheaper to design and write operating systems ?
>
>The unit volume is higher than it was. Software development is almost
>all fixed costs. So the unit price should be lower unless development
>costs have increased at least as fast as the market has grown.
>
>This also explains why CAD software costs thousands while Office is
>hundreds. There's about the same level of effort in development, but
>CAD is a much smaller market.
This explains why OEM Windows is about $50- despite being an enormous
project (-;
>Because unit volume is much higher now than it was in 1995 and software
>development costs are independent of volume.
How much higher ?
--
Donovan Rebbechi * http://pegasus.rutgers.edu/~elflord/ *
elflord at panix dot com
------------------------------
From: Craig Kelley <[EMAIL PROTECTED]>
Subject: Re: [OT] .sig
Date: 26 Feb 2001 22:40:37 -0700
Brent R <[EMAIL PROTECTED]> writes:
> mlw wrote:
> >
> > Ben Pfaff wrote:
> > >
> > > Aaron Kulkis <[EMAIL PROTECTED]> writes:
> > >
> > > > What part of United States Army do you not understand?
> > > >
> > > > The US Army defends the US Constitution, which secures my right,
> > > > by virtue of being in the US, to say anything I damn well please
> > > > on USENET or any other place.
> > >
> > > [Snip other nonsense]
> > >
> > > Wonderful. Only a week or so after I empty my killfile, another
> > > fuckhead Ugly American prick shows up.
> >
> > Please do not think Aaron is representative of all Americans. All countries and
> > all cultures have their share of people like Aaron.
>
> I'm an American too, and sadly I'd have to disagree. America breeds more
> immature, online, shit-talkers than any other country online.
They also have the most on-line.
Coincidence? Probably not.
--
The wheel is turning but the hamster is dead.
Craig Kelley -- [EMAIL PROTECTED]
http://www.isu.edu/~kellcrai finger [EMAIL PROTECTED] for PGP block
------------------------------
From: Craig Kelley <[EMAIL PROTECTED]>
Crossposted-To:
alt.destroy.microsoft,alt.linux.sux,comp.os.ms-windows.advocacy,comp.os.ms-windows.nt.advocacy
Subject: Re: NT vs *nix performance
Date: 26 Feb 2001 22:48:43 -0700
"Jan Johanson" <[EMAIL PROTECTED]> writes:
> I agree, anyone thinking that an OS has no effect on the performance of a
> database running on it is an idiot.
It actually has very little effect. DBM systems pretty much
re-implement all OS features within themselves (vfs, caching, raw
organization). The OS just needs to get out of the way. Some OSes
may not be so good at that (like MacOS 9 or lower, for instance) --
but by and large the OS has very little say.
--
The wheel is turning but the hamster is dead.
Craig Kelley -- [EMAIL PROTECTED]
http://www.isu.edu/~kellcrai finger [EMAIL PROTECTED] for PGP block
------------------------------
From: Craig Kelley <[EMAIL PROTECTED]>
Subject: Re: The Windows guy.
Date: 26 Feb 2001 22:52:26 -0700
"Erik Funkenbusch" <[EMAIL PROTECTED]> writes:
> "Steve Mading" <[EMAIL PROTECTED]> wrote in message
> news:976bmc$drc$[EMAIL PROTECTED]...
> > Erik Funkenbusch <[EMAIL PROTECTED]> wrote:
> >
> >
> > : The definition says it's a queue of bytes between two processes. A file
> is
> > : most definately a queue of bytes. And it bridges two programs via their
> > : stdin and stdout.
> >
> > Programs != Processes.
> >
> > The DOS style is actually a temporary storage between one process and
> > ITSELF, because there is only one process in DOS. At different times it
> > is populated by different program images, but it is only one process.
>
> I see you didn't comment on my arguments about other OS's that also don't
> have processes, but do have multitasking such as AmigaOS and MacOS <= 9.x
Maybe I'm misunderstanding here, but Amiga and Mac OS both have
processes.
--
The wheel is turning but the hamster is dead.
Craig Kelley -- [EMAIL PROTECTED]
http://www.isu.edu/~kellcrai finger [EMAIL PROTECTED] for PGP block
------------------------------
From: Tim Hanson <[EMAIL PROTECTED]>
Crossposted-To: alt.linux.sux,alt.destroy.microsoft
Subject: Re: Mircosoft Tax
Date: Tue, 27 Feb 2001 06:06:13 GMT
Donovan Rebbechi wrote:
>
> On Mon, 26 Feb 2001 00:47:28 -0700, Dave wrote:
> >On Sun, 25 Feb 2001 19:40:43 -0600, "Erik Funkenbusch"
> ><[EMAIL PROTECTED]> wrote:
> >
>
> >Amazing. MS has achieved the impossible - they've exceeded the US
> >Government in inefficiency! I mean, look at how much more Linus
> >Torvalds and his merry band have accomplished for practically nothing,
> >and here it takes MS "100's of millions or even multiple billions" to
> >create merely incremental improvements to thier sorry excuse for an
> >OS.
>
> No, Linus and friends certainly haven't accomplished it for "practically
> nothing". They are generous enough to volunteer their valuable time
> towards Linux development. The difference is not really that Linux is
> cheap -- the difference is that Linux is a gift. And a very valuable one.
That's not quite true and hasn't been for a while. Linus is paid well
by Transmeta for work which dovetails into his "hobby." In addition he
and many of the kernel as well as application developers are paid to
work on free software. He and live the dream of all of us Dilberts, to
do what we really want to do and get paid really well for it. Linus is
also on the boards of some Linux companies which pay him director's
fees.
--
There is no realizable power that man cannot, in time, fashion the
tools to attain, nor any power so secure that the naked ape will not
abuse it. So it is written in the genetic cards -- only physics and
war hold him in check. And also the wife who wants him home by five,
of course.
-- Encyclopedia Apocryphia, 1990 ed.
------------------------------
From: "Adam Warner" <[EMAIL PROTECTED]>
Subject: Redhat's CEO wants the Open Source Comm. to address the US Senate
Date: Tue, 27 Feb 2001 19:08:41 +1200
Hi all,
Wired.com has published a letter by Redhat's CEO:
http://www.wired.com/news/business/0,1367,42008,00.html
And a story based around a telephone interview:
http://www.wired.com/news/business/0,1367,42011-2,00.html
Matthew Szulik says:
"Red Hat, as a representative of the open-source community, would love to
have an opportunity to provide a counter-argument to (Microsoft's) claims to
the U.S. Senate. We'd love to bring the brightest minds in the open-source
community -- both within and outside of Red Hat -- to the U.S. Senate."
Cool :-)
Regards,
Adam
------------------------------
** 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 by posting to comp.os.linux.advocacy.
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
******************************
