Re: An article from Microsoft

[Steven Pritchard <steve@silug.org>]

I want to preface my comments by saying that I have not read
Microsoft's FUD, nor do I intend to.  From what I've heard, it's more
of the same stupidity that Microsoft is famous for.  (Also, it sounds
like they've started to implement some of what they said they
normally would do in the Halloween documents.)

Also, Jay, please don't be offended by anything I say here.  I don't
intend for this to be any sort of a personal attack.  (My first
response might sound like it.  If so, I apologize.)

Jay Link said:
> I do agree with them about the desktop; Win9x is much better there (of
> course, my ideal layout would be *MACINTOSH* clients and a Linux server, 
> if it weren't for the lack of software issue).

Anybody who thinks that Windows, Macs, or any other single-user system
make good desktops obviously has never had to support a large
enterprise organization.  In my job (supporting Unix workstations),
I'm able to admin a large number of workstations all over greater
Peoria, plus some in remote facilities in northern Illinois (and I've
even helped with a system in Germany), all without leaving my desk.  I
*very* rarely have to travel to other facilities, even to do
application support.  This is due to the multi-user nature of Unix,
and the network-transparency of X.

All of the PC support people I know (here) never even see their desks.
They're too busy running around to remote facilities.

I'll concede that GNOME and KDE may not be the best answer to GUIs,
but they're *extremely* nice, and they're only getting better.

> I liked how the Linux swap file can only be 128 MB; they've conveniently
> forgotten that you can chain up *16* of the damned things for a COMBINED
> total of 2 gigs.

The 128MB limit for swap devices is long gone.

osiris:~$ cat /proc/swaps
Filename			Type		Size	Used	Priority
/dev/hda4                       partition	530136	43004	-1

> Linux was not designed from the ground-up to support symmetrical
> multiprocessing (SMP): am I to believe that NT was? Even so, Linux
> supports SMP prefectly well now, so it's a moot point.

NT wasn't designed from the ground-up to support TCP/IP.  I guess that
makes it unsuitable for Internet servers?  (Heh...  If that was the
only thing that made NT unsuitable, Microsoft would be doing pretty
good.)

Anyway, as you point out, that's a ridiculous argument.

> graphical user interfaces (GUI): This is their only legitimate point.

Again, I haven't read it, but somehow I doubt that they had a
legitimate point.

> fine-grained security model: HA HA HA, HO HO HO, HEE HEE HEE

Notice how many of the NT exploits are specifically because of this
fine-grained security model.  (Keep It Simple, Stupid.)

> For File and Print services: Can't respond here; no experience.

The only problem Linux has (really Samba) with being a file/print
server *for Windows* (other systems don't have this problem) is that
Microsoft likes to change things in service packs and new versions
that break competing software.  It's all undocumented, so you
constantly have to play catch-up.

> For e-commerce workloads: can't comment

Just thrown in for buzzword-compliance, I'm sure.

> For transaction-orientated Line of Business applications: What the hell
> does that mean? Is this an example of "Meaningless Information to Promote
> Sales"?

I'm sure.

> Linux only supports 2 gigabytes (GB) of RAM on the x86 architecture: They
> even admit that this is on the verge of being a non-issue. Plus, 64 bit
> systems will change all this.

Linux has supported more than 2 GB of RAM on 64-bit platforms for a
long time (3 or more years).  Linux also runs fully 64-bit on 64-bit
processors *right now*.  NT is still stuck at 32-bit, and will be for
probably at least a couple of more years.

BTW, the development kernels have support for up to 4 GB on x86.
Supposedly the newer Intel processors support 36-bit addressing, and
there are people working on adding that support to the kernel.  (In
other words, it is definitely a temporary situation.)

> Linux Needs Real World Proof Points Rather than Anecdotal Stories:
> Read: NT must hide behind distorted statistics, because their anecdotal
> stories are all BAD!

Heh.  Well put.

There are plenty of real-world success stories about Linux.  So far
I've not heard one success story about NT that wasn't being promoted
by Microsoft.

Just to add another one to the "anecdotal stories", I've seen Linux
boxes with >1 year uptimes (note

    http://www.silug.org/lists/silug-discuss/msg01052.html

as an example.)  This is not unusual in the Unix world either.  I've
got two HP-UX servers that I support that currently look like this:

  2:02pm  up 423 days, 21 hrs,  13 users,  load average: 0.28, 0.14, 0.09

and

  2:03pm  up 367 days, 19 hrs,  2 users,  load average: 0.17, 0.11, 0.26

I've never seen an NT box come close to this.

> Therefore, commercial support services for Linux will be fee-based and
> will likely be priced at a premium: am I to believe that Win support is
> free? HO HO HO HEE HEE HEE HA HA HA!

And do people supporting Windows have access to the source code and
developers?

> See how these leading companies and organizations have deployed Windows NT
> Server 4.0: ignorant mother f*ckers...  Should I laugh or cry?

Depends on whether or not you work there, I suppose.

> With the release of the Windows 2000 operating system: no, no, let's not
> bait & switch here. If NT is so good, why are they dropping the name?

Standard Microsoft tactic.  Talk down a new technology, then promise
that a new piece of software or new version will do that, but even
better.  There are a million examples of this.  See their response to
NDS.  (Why would you want that when you can wait for NT 5 and get
Active Directory?!)

*sigh*

Steve
-- 
steve@silug.org           | Linux Users of Central Illinois
(217)698-1694             | Meetings the 4th Tuesday of every month
Steven Pritchard          | http://www.luci.org/ for more info

--
To unsubscribe, send email to majordomo@luci.org with
"unsubscribe luci-discuss" in the body.