Skip site navigation (1)Skip section navigation (2)
Date:      14 Nov 1999 15:33:11 +0200
From:      Giorgos Keramidas <>
To:        "David Schwartz" <>
Cc:        "Giorgos Keramidas" <>, <>
Subject:   Re: Judge: "Gates Was Main Culprit"
Message-ID:  <>
In-Reply-To: "David Schwartz"'s message of "Sat, 13 Nov 1999 14:32:53 -0800"
References:  <000001bf2e27$06a48230$>

Next in thread | Previous in thread | Raw E-Mail | Index | Archive | Help
"David Schwartz" <> writes:

> > Gawd's sake what do you mean by using the word `superior' here, because
> > I was assuming technical superiority is the goal.
> You were assuming technical superiority is the goal of what?

Of a company trying to produce competent software that will let it share 
some part of the market with others.

> Right. Perceived value is as important as actual value. This is
> especially true in the computer market, believe it or not.

As much as it grieves me, I do believe it ever since I started working
as a programmer in the computer market :|

> Yes, the limit should be what the consumer does not want. Not even
> Microsoft can make a consumer buy something they don't want. And not
> even the government should stop Microsoft from making and selling what
> consumers want.

This is certainly true.  The limit is always what the consumer wants.
However, raising the limit with marketing should not be that hard.  The
whole purpose of marketing is to actually /tell/ the consumer what he
wants, which creates a logical loop in that previous argument.

Oh and one more thing, unless I am totally wrong here, any company can
make some consumers buys something that they don't _need_ by making them 
think that they _want_ to buy it.  This is the idea behind commercials
and other ways of advertising.

> > As for the `user experience' thing, ask all those who have seen IE
> > stopping to download simultaneously from more than 4 sockets at a time,
> > and those who use Netscape and see all their active downloads proceeding
> > as usual.  And `user experience' is an objective criterion to define
> > superiority or `goodness'.  Both these terms usually depend on what one
> > calls `good' and what is regarded by each and every one of us as `bad'.
> 	Of course. I have my complaints about IE as well. Personally, I wish
> Microsoft had put 1/5 as much energy into designing ActiveX as Sun put into
> Java. I believe Java is technically superior to ActiveX and wish that Sun
> had made it more open and hadn't overpromised.
> 	But the fact still remains, the vast majority of people who compare IE to
> Netscape prefer IE. I have both on my desktop and use IE about 95% of the
> time. Yes, it crashes from time to time. So does Netscape.
> 	So why aren't the Netscape folks working on the next thing? Whatever it is
> that will make the browser obsolete. That's where the market for technical
> superiority is.
> > > You may consider some of Microsoft's innovations cheating, but that's
> > > not your decision to make.
> >
> > Using in half of the operating system the same libraries as IE uses is
> > certainly a nice design decision on behalf of Microsoft.  Having a
> > separate layer of abstraction, so that IE is not so closely bound to
> > Windows, would probably make porting IE to other platforms easier but
> > then again, it's already ported to a few platforms.  Not as many as I
> > would certainly wish to see it ported to, and not close to all the Unix
> > boxes supported by Netscape though.
> And the nice thing for developers targetting Windows 98 and later
> operating systems is that they can rely on those services being there
> and use them.  This is a benefit for the same reason that integrating
> memory management into the operating system is a benefit.

Memory management works transparently, without the programmer ever being 
in the need to tweak it, in order to do simple tasks.  However, all
other interfaces present in the system have to be documented properly,
so that one does not find that IE uses the OpenWindowFancy() call of a
system .dll, which seems pretty much undocumented in your system
manuals.  This will not let those `other' developers write their own

As for memory management...

> > > Go ahead, take memory management out of the OS. Take disk compression
> > > out.  How far do you want to set us back?
> >
> > Eh?  Did I ever even mention that any of these operating system
> > characteristics are *evil* and should be dropped by anyone?
> What about your argument against browser integration doesn't apply to
> disk compression integration? Or memory management integration? Or GUI
> integration? (Actually, GUI integration is something that I can't
> stand. But I'm not going to tell Microsoft how to design their
> products.)

Memory management is a characteristic of an operating system at the
"system" level.  Filesystems too.  Providing transparent encryption of
filesystem operations can be considered as a `feature' of a filesystem,
which is already in the "system" level.

However, one something is in the application level, i.e. programs like
the browser, the word processor, the mail reader, hell even the mail
transfer agent, it is no longer part of the OS.

Finally, if you think that I ever even hinted towards removing support
for system-level features, in order to do I don't what... well, try
reading once more my postings.

As for integration, Emacs has integrated support for the GNU debugger.
However, it's not mandatory to install Emacs when you install GDB.  The
fact the a program B can support a program A, means that if you want to
use B you'd probably find A useful.  However, this does not necessarily
become instantly true towards the other direction.

For instance, you can use X11 on Linux and/or FreeBSD.  But that does
not make it mandatory to install X11 on all Linux installations.  It
just adds to your `abilities'.  I would certainly like to see that level 
of fine-grained choise in Windows too.  Although it's not the only
design issue that makes me shy away from Windows, having a descent way
of working at them without the `integrated' GUI would really make me a
lot happier about them.

> If you don't like closed-source programs, don't buy them. But to tell
> everyone else that you no better than them and that closed-source
> should be eliminated is absurdly elitist.

I'm not buying them.  As for the elitistic part, telling others that
something is superior by something else, so you got to use it, because I 
and the rest of the planet use it 95% of the time... now that's e1it3 :)

> > > > Compare the price of running Netscape on *BSD, with that of running IE
> > > > on Windows.
> > [snip]
> > > Neither Netscape nor FreeBSD are commercial products. I'm talking
> > > about comparing Windows to other commercial operating systems.
> >
> > Pretty nice and rational now that you clarified it, but I thought that
> > we were not talking about commercial, and only commercial, software
> > products until now.
> When you're considering whether the cost of Windows is reasonable or
> not, you have to compare it to other commercial operating
> systems.  Microsoft had no means of enlisting thousands of people to
> work on Windows for them at no charge.

When as an engineer I am called to offer a solution, why is it important 
to limit myself to commercial only solutions?  I don't seem to follow
your reasoning here.

About the enlistment now, opening the source is one way.  So Microsoft
actually *has* the means to enlist all those developing on Windows now,
they are just too blind to see it.

> > However, many a time Microsoft decided to change the SMB protocol, for
> > no apparent reason.  The changes were in the key used to authenticate
> > oneself, but they did not seem to provide for better security (not
> > larger keys, just different and it has always been with proprietary
> > protocols, not documented anywhere).  They were there obviously in order
> > to make Samba unusable.  This is not what I usually call `better' when
> > I'm talking about a system in general.
> How did Microsoft's changes hurt samba?  It was still doing exactly
> what it was designed to do.  It simply no longer interoperated with
> Microsoft's operating systems.  This hurts Microsoft's operating
> systems as much or more than it hurts samba.

Yes and no.  Samba still worked.  Microsoft's neighborhood still worked,
they just did not interoperate well with each other.  By making just the
number of changes to be incompatible, you can bet on many people buying
the crap that goes like "our systems always interoperate well with each
other, theirs are just a heap of bull -- so use only our systems, be
happy and content that we provide you with all this interoperability."
Microsoft has never used my exact words, but the basic meaning most of
the time is right there.

> Microsoft did not ask for samba.  Microsoft does not have to suffer
> samba if they don't want to.  If they want their operating systems not
> to interoperate with other operating systems, that is their right.

By using the word "suffer" you don't mean of course that Samba is
getting _so_ good that Microsoft is in agony and all that...

Anyway, Digital, Sun, SCO, or anybody else did not ask for Linux or
FreeBSD either.  That does not mean that they should not follow any of
those dreaded POSIX standards which linux and freebsd strive to

> I develop products all that time that do not interoperate with
> products developed by other companies.  This is a deliberate design
> decision.  Or do you not believe that companies have the right to
> develop and use proprietary protocols and technologies?  Is this going
> to degenerate into 'information wants to be free'?

Developing products that do not interoperate, as you said above, is a
hurt to both the users and you, the developer.  I can only add to this
that you have every right to make a decision of your own, and I will
support your right to choose to do so, even if I disagree with your

About information, if it *does* want to be free, it will be.  Even if I
myself, you, or anybody else for that matter, tries to convince us all
of the opposite.

> > Lower prices are not a characteristic of Microsoft, unless you've never
> > actually payed for your copy of Windows and Office; but that's probably
> > something that does not conform with what Microsoft wants you to do.
> That's just not true.  Are you saying this for any particular reason?
> Do you really believe it?  Would any facts help, or is your mind
> already closed.

My mind is never closed.  I am always open to suggestions.  But I have
yet to see a Windows release priced less than FreeBSD or any other
Free*nix clone.  But, I know, I know... Free*nix is not commercial.

> Here's a quotation from Stan Liebowitz:

Nice stuff, can I find more of that online somewhere?

> Sure.  But I've had those same problems with Linux.  (Never with
> FreeBSD, but I think that's just been luck so far)

The problems, i.e. tech support on expensive and not so effective phone
lines, just don't exist with FreeBSD & Linux.  I can not recall the
year, I think it was the last one, when Linux received some award for
it's support (mind you, no real company does the support, just mailing
lists, irc channels, web pages, etc.)


Giorgos Keramidas, <>
"What we have to learn to do, we learn by doing." [Aristotle]

To Unsubscribe: send mail to
with "unsubscribe freebsd-chat" in the body of the message

Want to link to this message? Use this URL: <>