I will be happy to move this to private email once I debunk some stuff...
-----Original Message-----
From: Paul Surgeon [mailto:[EMAIL PROTECTED]]
Sent: Monday, February 12, 2001 13:34
To: Noonan, Wesley
Cc: [EMAIL PROTECTED]
Subject: Re: Squid and Samba
----- Original Message -----
From: "Noonan, Wesley" <[EMAIL PROTECTED]>
To: "'Paul Surgeon'" <[EMAIL PROTECTED]>; "Noonan, Wesley"
<[EMAIL PROTECTED]>
Cc: <[EMAIL PROTECTED]>
Sent: Monday, February 12, 2001 8:56 PM
Subject: RE: Squid and Samba
> As for the culprits of blue screens, since 90% of them
> are drivers, and Microsoft doesn't make drivers, who is really responsible
> here?
Hmmm ... I was under the impression that what comes out of the MS box was
written by MS and is CERTIFIED by MS?!
Then you need to take a class, 'cause you couldn't be more wrong.
Who actually writes the drivers that come out of the box and why do
Microsoft certify them if they are buggy?
Depends on who the vendor is. Also, Microsoft doesn't tend to certify them
if they are buggy (besides, Veritest does the certification, not Microsoft).
The bugs tend to get discovered in the interoperability between differing
drivers due in large part to the sheer amount of hardware Microsoft
supports. See, the tests only specify that the driver needs to work with the
OS, not with any other drivers. It is easy to not have driver problems when
you support 10-15% of the hardware that Microsoft does. About 9 months ago I
did a comparison of RAID controllers (SCSI and Fibre Channel) supported by
NT, W2K and RedHat Linux. RedHat supported 18. NT supported 287. W2K
supported 97 by logo, plus a sizable chunk of the NT controllers by
compliant. There's no comparison.
I can take a brand spanking new version of Windows out of the box, install
it and boot it and it will crash within 2 weeks even if I don't use it at
all.
This is 100% pure unfounded, subjective garbage. If you can take a new
version of W2K out of the box, install it boot it and it will crash within 2
weeks without even being used... you have power problems my friend. Get a
UPS.
I have added no extra non-MS certified software and haven't run anything on
it during that time.
I can do the same with a HP-Unix, Linux, Solaris box and after idling for 2
weeks they are all ready to get some work done.
This isn't even worthy of response.
> Dare I say, it might not be the software that is the problem.
No, not if I notice that same thing on about the 10 different systems I have
used.
Different processors, different make of memory, different motherboards,
differnet hard drives, etc, etc, etc.
Same installer... besides, I (and many other folks I know) don't have these
problems any more/less with W2K vs. Unix (discounting Solaris running on
Sparc's... it's easy to run well when you make the hardware and the software
:))
> Not because it is really bad software but because it is flawed in both
> design and stability.
>
> Rarely have I found the design and stability to be the result of the
> software. For more often, it is the result of the admin.
Ummm ... yes you are right - I am a crap admin.
See, now you are personalizing this and letting your emotions over ride
logic. Not a good place to be when managing systems that are based purely on
logic (is it a 1, or is it a 0, or is it both for Quantum computers :)).
I installed the OS and it crashes so I got a Microsoft Certified Whatever to
do the install for me and Windows still crashes.
Guess MS don't know how to get their techies to do the magic stuff either.
So you got a bad tech. There's lots of them out there, and the good news is
they don't stick to Microsoft only.
The best they can do is delay the crashes by installing more RAM at my
expense which is a solution I do not accept when I am paying for a solution
that works!
Hogwash. Caveat Emptor. If you buy off on such low quality support, then you
are likely going to suffer from it.
I once learnt something (one of the few things I have learnt) in Ansi C and
C++ programming and it's this :
If you allocate memory, point a pointer to it and forget to de-allocate the
memory before destroying the pointer you get something called a "Memory
Leak".
Windows (out of the box) obviously suffers from this because after 1 week on
16 MB of RAM it is a slow as a dog and is thrashing the disk to pieces
because it has lost nearly all of the 16MB.
What, Windows 9x? Been a long time since a server product from Microsoft
would run on 16MB RAM. It definitely brings to light what some of your
problems might be, IMHO.
> This should really go to private mail...
I though this was a public mailing list?
You are right ... I should have discussed this in an MS mailing list.
Actually, I meant email me since it is no longer a firewall discussion.
Regardless, this will be my last post to the mailing list on this subject. I
think I have made my point, and any further belaboring of the point can be
done in private email.
-
[To unsubscribe, send mail to [EMAIL PROTECTED] with
"unsubscribe firewalls" in the body of the message.]
