Linux-Advocacy Digest #116, Volume #27 Fri, 16 Jun 00 09:13:06 EDT
Contents:
Re: Would a M$ Voluntary Split Save It? ("Daniel Johnson")
Re: The Tholenbot (was: Microsoft invites Canada south) (Jacques Guy)
Re: Would a M$ Voluntary Split Save It? ("Daniel Johnson")
Re: Would a M$ Voluntary Split Save It? ("Daniel Johnson")
Re: Would a M$ Voluntary Split Save It? ("Daniel Johnson")
Re: Coherency (2:1)
Re: Claims of Windows supporting old applications are reflecting reality or fantasy?
(Sam)
Re: Good books on writing a kernel. (Matthias Warkus)
Re: Alpha vs Intel (Ben Chausse)
Re: Claims of Windows supporting old applications are reflecting reality or fantasy?
(2:1)
----------------------------------------------------------------------------
From: "Daniel Johnson" <[EMAIL PROTECTED]>
Crossposted-To:
comp.os.os2.advocacy,comp.os.ms-windows.nt.advocacy,comp.sys.mac.advocacy
Subject: Re: Would a M$ Voluntary Split Save It?
Date: Fri, 16 Jun 2000 12:15:36 GMT
"Bob Hauck" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> On Thu, 15 Jun 2000 23:15:27 GMT, Daniel Johnson
> <[EMAIL PROTECTED]> wrote:
[snip]
> >> I didn't "make" them do anything. Mostly, they would read some book
> >> and find out that they could do things on the server.
> >
> >You've got unusual users there.
>
> Not in 1994-1996. The Internet was much more geek-oriented then.
That is just to say that back then, ordinary users didn't use the 'net.
Probably because it was too hard. :D
[snip]
> >That's your opinion; MS may well feel it made it easier to achieve
> >the functionality they wanted.
>
> Yes, it is my opinon. And since other products have similar features
> without requiring special hacking of Apache or a particular web server,
> I think it is a reasonable opinion.
Pardon me if I express doubt, but I've too often heard that
thus-and-such a product has "similar features" which,
when examined, don't really measure up.
[snip]
> >This isn't true. I realize that it's popular in some circles to accept
> >all that the DoJ says as the gospel truth, but it ain't so.
>
> What isn't true? That they try to lock out other products?
Yes. They go through a lot of trouble to provide ways for other
products to integrate with their own.
> Well, if
> not, then there's been an amazing series of lucky coincidences. It
> sure is true that so-called "product tying" was an issue in the trial.
For awhile, it was *the* issue.
> >MS is very gung ho about interoperability.
>
> Yup, they are "gung-ho" about interoperability *between their own
> products*, just not with other vendor's products or standards-based
> products.
That's not so. MS goes *much* farther than any Unix does to
interoperate with other vendors products.
> They actively try to sabotage interoperability with other
> operating systems by claiming full compatibility and then doing the
> absolute minimum required by the letter of the spec, never mind that
> every single other product in the space has certain common extensions.
IMHO, this is a problem with the specs- a problem that is almost inevitable
given the standards process used to produce them.
> They then go and make an incompatible version that does do what people
> expect and say "see, our stuff works better than that other crap,
> we have better features than the standard, switch to MS".
Yup. That's their strategy for beating open source in a nutshell.
[snip]
> >No doubt. But it is perfectly reasonable for them to try to peddle
> >their internet servers as "easier to use"; they should not be obliged
> >to provide similar improvements for their competitors.
>
> You just don't want to hear what I'm saying here.
I don't want to *agree* with what you are saying.
> Their server was _not_ easier to use for an ISP,
I didn't say "for an ISP". The real problem is with
end user support.
Remeber my Standard Explaination (tm): MS succeeds
because it sees problem that must be solved, rather
than trying to find problems to solve with the technology
it already has.
The problem to be solved here is giving end users
their own web pages, which they can maintain themselves.
> because it did not support the ISP's
> existing infrastructure. They were trying to *tie* the FP client
> (which they could bundle with Office and lowball the price on to help
> acceptance) to the IIS server in order to force people to buy the
> server whether they wanted it or not.
You say "tie" as if it were a bad thing. I certainly don't think it is,
not in this case.
They wanted to sell IIS by giving it better features than the
next product. Saying they "forced" you to buy IIS because they
didn't implement the same features for competitors products
is just not reasonable.
> Nobody else could play that
> game because nobody else had a lock on the desktop market.
Nonsense. Distributing your client for free is childs play.
You can make it available for free downloads.
You can give it to ISPs, and *they* can give it to their clients.
You can give it to desktop computer makers and they can
bundle it with their computers.
You can give it to other software makers and they can bundle
it with the applications.
You can mail CD-ROMs out to all and sundry.
You can think of something that hasn't been done a million
times before, by why bother? The above techniques *work*.
> >I really do object to this whole "they didn't give *us* an easy to use
> >front end, too!" line here. MS is not obliged to write software for
> >Apache just because you would like to have such software.
>
> A normal company, one without a river of cash flowing in from other
> sources,
That's a "normal" company?
> could not afford to play the game the way MS did.
Sure they could. It's the razors/razor blades thing. Give the client
away; sell the server.
> They would
> have had to make the server side compatible with the ISP's existing
> servers,
Bad idea. You can't sell your server if you do this.
> and be reasonably easy to support, or nobody would have
> installed it.
It must be supportable, yes, or it won't be used. I think MS managed
to acheive that, more or less.
> The ISP's could have just said "sorry, we don't support
> that".
They could do that. They'd be at a disadvantage next to other
ISPs that offer the cool new feature to their customers.
This, of course, applied regardless of who makes FrontPage;
you don't need a monopoly for it.
> But MS could get millions of copies of their client out there
> through bundling deals and lowball pricing, making this an untenable
> position for most ISP's to take.
It's not exactly "untenable", it's just a disadvantage not to have
FrontPage or something like it, if your competitors have it.
> What are they selling? A web-publishing tool? A web server? Yes,
> both,
No, just a web server. The web-publishing tool is free.
> and they want to tie them together so you can't fully use one
> without the other. And they have the means to give a large percentage
> of their customers a copy of the client included with other products.
> Does this sound like a fair and honest way to enter the web server
> market to you? Or does it sound like leverage?
It's fair and honest and *commonplace*. What MS has done here
is completely ordinary; you are the one demanding radical
strange policies of MS here.
> >MS had focused on integrating with Netware. That's why it didn't
> >ship with a kerberos client- until now.
>
> Yes, they were latecomers to the game. Their product was not ready,
> which is why ISP's did not want it. How that happened is irrelevant.
Surely you don't mean that. Perhaps you mean 'it is irrelevant unless
I can use it to potray MS as evil'?
I realize that NetWare isn't Unix, but in reality it was the big
competitor to NT for years.
> >Well, you *can* use MS's tools with other peoples authentication and
> >databases and so on. MS tries to make theirs work *better*, which
> >hardly seems so awful.
>
> How do you make NT authenticate to an NIS database? You buy a third
> party product, which is likely to break with the next service pack. It
> isn't clear to me how this is "better".
What makes you think it is likely to break with the next service pack?
Most things don't.
Anyway, it's better because you aren't limited to Unix. Anyone can
write these client modules and put NT computers on their
network.
With "open" standards, you are just committing to Unix. This is no
better in princinple than committing to OS/2 LAN Manager would
have been.
> Maybe it is "better" if you use all NT and get rid of NIS, but that's
> just not a realistic option for everyone.
Well, if you do that you get to have the features of an NT server;
this may or may not be worthwhile. For a long time, many people
used NetWare servers for this because they felt that NT 4's
directory services were not up to snuff.
Windows 2000 is supposed to fix that, of course. We'll see.
> >One way MS does this is by providing very rich APIs that let their
> >product work together closely. But these APIs are documented
> >and can be used in the same way by other products as well.
>
> Yeah, right, they care deeply about interoperability, and they document
> things so well that it is trivial to implement a compatible version.
I don't say "trivial". It's possible.
> That's why there are so many non-MS clients for Exchange. That's why
> there are so many non-MS PDC and BDC servers.
Now I don't know much about Exchange. But I *do* know there's a
documented API with which you can provide your own domain
controllers. If no-one has done it, maybe that's because no-one
feels it is worthwhile to go up against MS and NetWare both.
> Once again, MS redefines
> the language. MS's idea of interoperability is for their products to
> interoperate with their other products.
And other peoples products.
> On some days, it is ok for
> other products to interoperate too, if they run on an MS platform.
Also sometimes if they don't. The security API, for instance,
does not require that the domain controller or directory server
or whatnot be running an MS OS.
> That's why they think API compatibility is all that's needed and
> Win-supporters like yourself are derisive of the need for "wire
> protocols" to be documented. But that is the very thing that proves
> that they (and you) don't really care about true interoperability.
"True interoperability" to you seems only to mean "works with
Unix.". Using Unix wire protocols will not buy you compatibility
with MacOS or NetWare or any other OS but Unix.
That's interoperability?
> No other platform supports the Windows API, so documented API's are
> worthless for vendors who wish to interoperate on other platforms.
Windows it itself irrelevant to vendors who wish to interoperate only
between other platforms. They become relevant only when you bring
Windows into the mix.
> If
> a third party wants to make a client for Exchange that runs on Unix or
> VMS or VM, they have to know how to talk to Exchange over the wire.
> The API's used are irrelevant.
I don't know Exchange well enough to say. I know this is not true
of other things MS has done.
> If MS wants to document the API only, then if they really truly care
> about interoperating and supporting their customers who might not be
> ready to go all-MS, they should provide a reference implementation of
> that API with source. Otherwise they are just blowing smoke.
MS does not generally provide reference implementations of these
things with source; Windows developers don't seem to need it.
Perhaps we're smarter than Unix developers. :D
> >> Leveraging a monopoly in this way is illegal.
> >
> >Very dubious.
>
> No, that's what the law says. It is not a "strange interpretation"
> either, as it has been applied before. AT&T was prevented from getting
> into the computer market because they had a communications monopoly.
That's not even the same interpretation you had before. *Now* you
are saying that if a company is successful in one market, they
are not allowed to expand into any other market, *regardless* of
how they do it. 'Leveraging' is not the point; you just can't expand.
That is a *horrible* stupid policy; it's destructive, immoral, and wrong
it just about every possible way.
I'm not surprised to DoJ favors it. :D
I realize that the DoJ has been into creative intepretation of the
Sherman act for years and years, and one might argue that
it is, through familiarity, no longer strange. But I think that
quibble leaves my original point unaltered.
You might argue that they can be expected to get away with it,
I guess- as they have before. But that hardly means the
law- still less morality- is on their side.
> >Yes, that's what makes this so scary. The DoJ really seems to think
> >they should be micromanaging software design in general.
>
> Then why the breakup solution?
They want to take the web browser feature out of windows, and
the breakup does that.
> If they wanted to micromanage, they
> could have asked for a supervisory remedy.
Wouldn't have a chance, given the appeals courts previous
ruling.
> I don't think the DoJ has
> any interest in directing software design.
Nonsense. They keep insisting on making product design
decisions for Microsoft. What is that if not directing software
design?
> I think they have an
> interest in preventing an excessive concentration of market power in an
> important industry.
No, they don't. I realise they are supposed to work for 'the interests of
the
people', but it takes a remarkably naive person to think that's what
they really do. Pious political posturing is cheap.
They have an interest in perpetuating and expanding their beauracracy;
concentrations of market power are at most opportunities for this.
------------------------------
Date: Fri, 16 Jun 2000 12:26:50 -0700
From: Jacques Guy <[EMAIL PROTECTED]>
Reply-To: [EMAIL PROTECTED]
Crossposted-To:
comp.sys.mac.advocacy,comp.os.os2.advocacy,comp.os.ms-windows.nt.advocacy
Subject: Re: The Tholenbot (was: Microsoft invites Canada south)
Tholen the Bot wrote:
>
> In article <[EMAIL PROTECTED]>, [EMAIL PROTECTED]
> wrote:
[never mind what I wrote, listen to the bot:]
> When did you first know that appeared at the same time as this from
> tholenbot?
[snip me]
> What makes you believe it is a bot?
[snip me]
> Does it bother you that the armenians genocided the Turks?
[snip me]
> Maybe your plans have something to do with this.
[snip me]
> I don't understand.
[snip me]
> You mentioned that the armenians genocided the turks?
[snip me]
> Is it because of your life that you are going through all of this?
[snip me]
> Eliza? Hah! I wish you would continue.
If it were a real, even only vaguely human, being, such as they may
harbour not at Ann but at Cornell, he/she would have picked
the intentional blooper which you may find in one of the [snip me]'s.
Welcome back, Hasan B. Mutlu, glad to see you are risen from the
grave! Are you still with AT&T?
------------------------------
From: "Daniel Johnson" <[EMAIL PROTECTED]>
Crossposted-To:
comp.os.os2.advocacy,comp.os.ms-windows.nt.advocacy,comp.sys.mac.advocacy
Subject: Re: Would a M$ Voluntary Split Save It?
Date: Fri, 16 Jun 2000 12:25:32 GMT
"Grant Fischer" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> On Thu, 15 Jun 2000 23:15:29 GMT, Daniel Johnson
> <[EMAIL PROTECTED]> wrote:
[snip]
> >MS thinks layers too, but they have different layers. MS's security
> >support providers were not designed to work with Kerberos. They
> >expect to get more information about the user in question that
> >just his identity.
>
> We're talking about the people who just redesigned their whole
> security structure, and chose to design Kerberos into it.
No, they didn't.
The didn't need to because their original design was flexible
enough to permit them to put Kerberos in with only a small
addition to Kerberos.
That's why APIs and plug-ins are better than wire protocols- they
lets you cope with change far more easily.
> You're saying that they had to extend it because it was limited;
> I pointed out that there were other, more complete frameworks.
> Now you're saying it didn't fit right; that's a funny thing
> to say about their design choices.
I don't think DCE would have fit *better* in this area.
> >This could be kludged on as a separate protocol, but this would
> >do nothing for interoperability, it would be slower, and it would
> >open up new opportunities for spoofing the security system.
>
> What the heck are you talking about? Kerberos takes care of
> the identity of both the users and the services they talk to.
> That's what it does.
Sure, provided you understand identity as a user name,
as Unix does. NT needs a little more.
> Interoperability would take care of itself if MS made it an
> open protocol -- publish the complete data specs rather than
> API's.
MS is not obliged to implement things the way Unix does.
> I agree that it will be slower to implement authorization in
> a second service; however it would remain to be seen whether
> it would be noticeably slower. If so, there might be other ways
> to mitigate that.
True. But there do not seem to be any advantages to doing it
that way; you don't gain any interoperability. NT clients need
a server with that extra protocol, that's all.
> >Kerberos was designed to fit into Unix's structure. It didn't quite
> >fit NT.
>
> It actually has no concept of what OS it is running on; it uses
> simple principal names to name users and services. You're thinking of the
> implementation of the kerberized UNIX login services; they by
> definition were designed to fit into UNIX's structure, and are
> quite irrelevent to how NT could use Kerberos.
"Simple principle names" are the problem. NT's principles are not so
simple.
NT security ids contain a list of user ids, groups, and other attributes
that can be compared against ACLs. Some are dynamically
generated, but others must be sent over the wire at *some* point.
------------------------------
From: "Daniel Johnson" <[EMAIL PROTECTED]>
Crossposted-To:
comp.os.os2.advocacy,comp.os.ms-windows.nt.advocacy,comp.sys.mac.advocacy
Subject: Re: Would a M$ Voluntary Split Save It?
Date: Fri, 16 Jun 2000 12:25:33 GMT
"JEDIDIAH" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> On Thu, 15 Jun 2000 23:15:29 GMT, Daniel Johnson
<[EMAIL PROTECTED]> wrote:
> >"Grant Fischer" <[EMAIL PROTECTED]> wrote in message
> >news:[EMAIL PROTECTED]...
[snip]
> >Kerberos was designed to fit into Unix's structure. It didn't quite
> >fit NT.
>
> Then Microsoft shouldn't be telling us pleasant lies.
You prefer unpleasant lies? :P
------------------------------
From: "Daniel Johnson" <[EMAIL PROTECTED]>
Crossposted-To:
comp.os.os2.advocacy,comp.os.ms-windows.nt.advocacy,comp.sys.mac.advocacy
Subject: Re: Would a M$ Voluntary Split Save It?
Date: Fri, 16 Jun 2000 12:25:33 GMT
"Grant Fischer" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> On Thu, 15 Jun 2000 23:15:16 GMT, Daniel Johnson
> <[EMAIL PROTECTED]> wrote:
[snip]
> >> A lot of UNIX users use Windows desktops. Why? MS Office.
> >
> >If they can get MS Office, can they not also get a good
> >telnet client?
>
> Of course. But they could also get a web browser.
> Their HTML help system doesn't justify the expense
> put into IE.
I think IE has more utility that just being the help system,
really. :D
> >MS telnet is certainly half-assed but I don't see MS support
> >for "open standards" to be obligatory unless those
> >standards are useful, and telnet is marginal at best for
> >an NT system.
>
> Hey, we're users too. Telnet is marginal for talking to
> an NT system, but very useful for talking to a lot of other
> devices on a corporate WAN.
That's nice.
Telnet is still marginal at best for an NT system.
Demanding that MS is *obliged* to provide support
for your stuff *just* because you want it, darn it, is
pretty weak. The demand for a kick-butt telnet isn't
there.
------------------------------
From: 2:1 <[EMAIL PROTECTED]>
Subject: Re: Coherency
Date: Fri, 16 Jun 2000 13:25:20 +0100
> Having one icon format is absurd? That's probably why KDE is so bloted.
> It has to be abal to read every immage format out thear (except the
> pollitically incorrect once like GIF).
That need not cause bloat. Linux already has the capability to let
almost any app read almost any image file. The pnm utils.
-Ed
--
The day of judgement cometh. Join us O sinful one...
http://fuji.stcatz.ox.ac.uk/cult/index.html
------------------------------
From: Sam <[EMAIL PROTECTED]>
Crossposted-To: comp.os.ms-windows.advocacy,comp.os.ms-windows.nt.advocacy
Subject: Re: Claims of Windows supporting old applications are reflecting reality or
fantasy?
Date: Fri, 16 Jun 2000 22:30:18 +1000
On Fri, 16 Jun 2000 12:22:35 +0100, 2:1 <[EMAIL PROTECTED]> wrote:
>I also meant to say (but forgot) that the lifespan of other consumer
>products is longer thasn 8 years. TVs, fridges etc can have lifespand
>over 20 years. Technology may have improged, but they have become no
>worse. Software is not different
You need an Oxford dictionary
Sam
------------------------------
From: [EMAIL PROTECTED] (Matthias Warkus)
Crossposted-To: comp.unix.programmer,comp.unix.solaris,comp.unix.aix
Subject: Re: Good books on writing a kernel.
Date: Fri, 16 Jun 2000 00:19:37 +0200
Reply-To: [EMAIL PROTECTED]
It was the Thu, 15 Jun 2000 16:48:52 -0500...
...and Manish Ahuja <[EMAIL PROTECTED]> wrote:
> What I find most difficult is that it is very difficult to get a BIG
> picture when
> you look at any OS code. The complexity of the whole thing eludes you.
> There
> are lotsa good books out there, but reading helps you to clear stuff up
> but only when
> you write and implement it. Any good books that can help me write an
> absolute rudimentary kernel, or maybe parts of it like FS, VMM or proc
> to begin with.
The Logical Design of Operating Systems, Second Edition
by Lubomir Bic and Alan C. Shaw
(C) 1988 by Prentice-Hall
mawa
--
Manchmal fühle ich mich im Usenet wie eine Kreuzung aus Don Giovanni
und dem Zauberlehrling; während es um mich herum unkontrollierbar
wuselt, weil ich etwas losgetreten habe, möchte ich rufen "Genug!
Genug!", und darauf warten, daß der Teufel mich holt. -- mawa
------------------------------
From: Ben Chausse <[EMAIL PROTECTED]>
Crossposted-To:
comp.unix.admin,comp.os.linux.misc,comp.infosystems.www.servers.unix,linux.redhat,comp.os.linux,comp.os.linux.networking,comp.os.linux.help
Subject: Re: Alpha vs Intel
Date: Fri, 16 Jun 2000 12:49:15 GMT
Well, we will have RealServer 7.0, plus about 3000 MP3's to serve to 600
employee, 3000 static web pages, about 500 php and perl web applications, a
huge database...
Mark Rafn wrote:
> Ben Chausse <[EMAIL PROTECTED]> wrote:
> >I build a Intranet WebServer on Linux with Apache 1.3.12, mod_perl 1.49
> >and PHP4 and I would like to know what will the best between a server
> >with 2x667 MHZ Alpha Processer and a 4x700 MHZ Xeon Processer ??
>
> Lordie! What are you doing on your intranet that you need such power?
> Unless you've got a pretty specific task that's very processor-intensive,
> you'll bottleneck on I/O long before CPU on either of these platforms.
>
> In general, for webserving, you'd rather have multiple
> load-balanced/redundant machines over one monster machine.
> --
> Mark Rafn [EMAIL PROTECTED] <http://www.dagon.net/>
------------------------------
From: 2:1 <[EMAIL PROTECTED]>
Crossposted-To: comp.os.ms-windows.advocacy,comp.os.ms-windows.nt.advocacy
Subject: Re: Claims of Windows supporting old applications are reflecting reality or
fantasy?
Date: Fri, 16 Jun 2000 14:05:57 +0100
Sam wrote:
>
> On Fri, 16 Jun 2000 12:22:35 +0100, 2:1 <[EMAIL PROTECTED]> wrote:
>
> >I also meant to say (but forgot) that the lifespan of other consumer
> >products is longer thasn 8 years. TVs, fridges etc can have lifespand
> >over 20 years. Technology may have improged, but they have become no
> >worse. Software is not different
>
> You need an Oxford dictionary
>
> Sam
Oh, dear. I should start proof reading emails. Dunno anout a dictionary,
but typing lessons might help;-)
-Ed
--
The day of judgement cometh. Join us O sinful one...
http://fuji.stcatz.ox.ac.uk/cult/index.html
remove foo from the end and reverse my email address to make any use of
it.
------------------------------
** 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
******************************
