From: "Frank C. Garber" <garbe...@coolsite.net>
Subject: Microsoft monopoly threatens consumer buying choices
Date: 1997/08/12
Message-ID: <33F096DF.4E8ABB8B@coolsite.net>#1/1
X-Deja-AN: 263746033
X-Priority: 3 (Normal)
Organization: Prosoft Data Systems
Reply-To: garbe...@coolsite.net
Newsgroups: comp.lang.c,microsoft.public.vc.events,comp.lang.java.advocacy,
comp.lang.c++,comp.lang.java,symantec.support.devtools.pc.visual-cafe.java
Boycott Microsoft
Full story: http://www.vcnet.com/bms/default.html
Excerpt:
Microsoft's strategy includes giving away its competing product,
providing exclusive free access to information, offering content which
can
only be viewed with their browser, and supporting this product
give-away program with a massive national advertising campaign.
Microsoft understands that none of its competitors can afford to give
away products that cost millions of dollars to manufacture. In any other
industry, selling products at a loss for the purpose of driving another
company out of business would be instantly disallowed as "dumping."
Why should this industry be treated differently?
Lest you mistake their purpose, Microsoft is amazingly candid about the
consumer-defeating designs of its Internet Explorer scheme. Said Steve
Ballmer, Microsoft's vice president of sales, "We're giving away a
pretty
good browser as part of the operating system. How long can they
survive selling it?" [Forbes, 27 Jan 1997]
From: "Paul Gustavson" <pgus...@illuminet.net>
Subject: Re: Microsoft monopoly threatens consumer buying choices
Date: 1997/08/14
Message-ID: <5sukls$lpm$1@news.netrail.net>#1/1
X-Deja-AN: 264166921
X-MimeOle: Produced By Microsoft MimeOLE Engine V4.71.1008.3
Organization: Netrail. Inc.
Newsgroups: comp.lang.c,microsoft.public.vc.events,comp.lang.java.advocacy,
comp.lang.c++,comp.lang.java,symantec.support.devtools.pc.visual-cafe.java
I couldn't resist adding my comments to this thread.
This may be obvious, but my personal feeling is that competition is
good for the consumer. As I scan back over the last few years
I see that somebody else (other than Microsoft) came up with a
good idea and MS choose to compete against them. As a result,
either the competitor improved on the product, or Microsoft developed
an application that proved equivalent or better.
Here are some examples...
Borland C++, once the top dog in development environments, it was
subplanted by Microsoft when MS VC++ 4.0 came out, BUT Borland
seems to have rebounded with the Borland C++Builder line.
Netscape!. We wouldn't have IE if it wasn't for Netscape. And
wether you're a MS fan or not, IE 3.0 & 4.0 is a pretty good browser
(especially for the money). Personally I hope Netscape survives
(like Norton has in their competition with Microsoft System Utilities).
Windows (whether you love it or hate it) wouldn't be so popular if
it wasn't for the work done by Apple and some of the X-Window
environments for UNIX (I think of SGI's environment). The problem
with 95 & 98 is that it's built on top of DOS and contains legacy win16
architecture to support your old DOS and Win 3.1 apps. NT is the
better choice for pure Win32. A consumer version of NT is not
far off (year 2K).
GL -> OpenGL. We have Silicon Graphics to thank for this. OpenGL
is available for Win32 developers, but MS also has DirectDraw &
Direct3D. If you have ever used the Windows GDI, you know that
GDI graphics programmin is the pits, both opengl and directX are a
welcome addition.
WordPerfect & Lotus 123. There would be no MS Office if it wasn't
for these two past winners. And they maybe winners again, if
MS doesn't do something about their memory hogging bloatware.
I could go on.
The bottom line, I think we (consumers) are better off now, then we
would be if it Microsoft was the only one pioneering these applications
and technologies. Microsoft maybe the marker share leader, but it's the
consumer & his pocket book who chooses who's on top.
I personally believe that if IBM gave away OS/2 Warp a few years
back (a true Win32 OS before there was ever an Win95 beta), we would
be discussing which OS is better OS/2 Warp or Windows. And I think
OS/2 Warp would be slightly ahead of Microsoft. And consider this, if
both operating systems were Win32, then as developer's it wouldn't
matter which OS we developed on, the application (in an ideal world)
would be able to run on either OS/2 Warp or Win.
Okay so back to the topic, "Microsoft monopoly threatens consumer buying
choices."
Suppose MS went through a breakup like Ma Bell did back in the early 80s
(or was it late 70s). I think that we would see even better products
coming out of redmon. Why? The gov't would have equalized the competition,
and as we all know MS (wether one big company or several smaller ones)
loves to compete. To me that is what makes MS successful.
Paul
From: sc...@lighthouse.softbase.com (Scott McMahan)
Subject: Re: Microsoft monopoly threatens consumer buying choices
Date: 1997/08/14
Message-ID: <5svm9f$igb$1@scully.new-era.net>#1/1
X-Deja-AN: 264700541
References: <5sukls$lpm$1@news.netrail.net>
Followup-To: comp.lang.c,microsoft.public.vc.events,comp.lang.java.advocacy,
comp.lang.c++,comp.lang.java,symantec.support.devtools.pc.visual-cafe.java
Organization: New Era Technologies, Inc.
Newsgroups: comp.lang.c,microsoft.public.vc.events,comp.lang.java.advocacy,
comp.lang.c++,comp.lang.java,symantec.support.devtools.pc.visual-cafe.java
Paul Gustavson (pgus...@illuminet.net) wrote:
: The bottom line, I think we (consumers) are better off now, then we
: would be if it Microsoft was the only one pioneering these applications
: and technologies. Microsoft maybe the marker share leader, but it's the
: consumer & his pocket book who chooses who's on top.
Name one market other than operating systems where Microsoft has won
other than by default. There is almost *NO* competition at all for
Microsoft. Most software companies in the '90s could not get out of
their own way, let alone compete. Manzi made his millions running Lotus
into the ground and it'll never recover; Wordperfect was sold and
resold and with some of the bad, buggy releases people lost confidence
in them; Borland, well, was Borland; and so on. Who the heck is
supposed to be offering an alternative to Microsoft? It wasn't until
NT's 3.51 and 4.0 releases that traditionally high end UNIX companies
like Oracle and IBM took the Wintel platform seriously, and they
haven't really had time to do anything.
: I personally believe that if IBM gave away OS/2 Warp a few years
: back (a true Win32 OS before there was ever an Win95 beta), we would
: be discussing which OS is better OS/2 Warp or Windows. And I think
: OS/2 Warp would be slightly ahead of Microsoft.
And I think you couldn't pay people to use an IBM operating system.
OS/2 appeals to a certain narrow type of technical people, but
to most people it is unfriendly, bizarre, and unfathomable.
: Okay so back to the topic, "Microsoft monopoly threatens consumer buying
: choices."
The topic's premise is flawed: people *HAD* choices. During the 90s,
there were a *LOT* of choices. At one time, Microsoft, Lotus,
WordPerfect, and Borland had roughly equivalent products. The
companies offering the other choices dropped the ball in various
different ways while Microsoft continually improved its offerings.
How come you always hear about Microsoft's "monopoly" being bad, but
you never hear anyone complain IBM is the most incompetent PC company
in history, Lotus was the worst managed company in the 90s, Borland was
an aimless company with no plan (they had a database, and BOUGHT
ANOTHER DATABASE!), etc? Microsoft hasn't *HAD* to do anything
bad, all they've had to do is watch their so-called "competition"
go down the tubes.
Scott
From: ktur...@pug1.sprocketshop.com (Kenneth P. Turvey)
Subject: Re: Microsoft monopoly threatens consumer buying choices
Date: 1997/08/14
Message-ID: <slrn5v75gb.tb.kturvey@pug1.sprocketshop.com>
X-Deja-AN: 264290424
References: <5sukls$lpm$1@news.netrail.net> <5svm9f$igb$1@scully.new-era.net>
Organization: No.
Newsgroups: comp.lang.c,microsoft.public.vc.events,comp.lang.java.advocacy,
comp.lang.c++,comp.lang.java,symantec.support.devtools.pc.visual-cafe.java
[I started to change the follow up to, but what should it be? I guess
this effects us all. Any suggestions?]
On 14 Aug 1997 19:28:47 GMT,
Scott McMahan <sc...@lighthouse.softbase.com> wrote:
>Name one market other than operating systems where Microsoft has won
>other than by default. There is almost *NO* competition at all for
>Microsoft. Most software companies in the '90s could not get out of
>their own way, let alone compete. Manzi made his millions running Lotus
>into the ground and it'll never recover; Wordperfect was sold and
>resold and with some of the bad, buggy releases people lost confidence
>in them; Borland, well, was Borland; and so on. Who the heck is
>supposed to be offering an alternative to Microsoft? It wasn't until
>NT's 3.51 and 4.0 releases that traditionally high end UNIX companies
>like Oracle and IBM took the Wintel platform seriously, and they
>haven't really had time to do anything.
I disagree. The reason there isn't any competition anymore isn't
because Microsoft has done such a great job. It is because Microsoft
owned the operating system and they could leverage that market to
control the other markets you speak of. Why didn't Word Perfect win
the battle of the word processors? They had a superior product. They
had better support. Why does the windows Wordpad support the Word 6
document format, and look suspiciously like Word 6? Because it shares
code with Word 6, I am sure of it. Anyone in a software company that
is competing head to head with Microsoft is praying they get bought
instead of crushed. There isn't an alternative.
Perhaps Word Perfect isn't the best example, but there are many
others. Microsoft drives its competition out of business by
convincing its customers that purchasing applications from the same
company that provides the operating system is advantageous. They
force it to be advantageous by collaberation between the operating
systems programmers (who have a monopoly) and the application
programmers (that are working on getting a monopoly of their own).
Microsoft adds (and probably doesn't add) features to its operating
system with these goals in mind. Why can't you keep Windows 95 from
putting that Microsoft Network icon on the desktop in the
installation? If you tell the setup program you don't want to install
it, it puts a shortcut to the install on your desktop (just in case
you change your mind). This was an effort to put compuserve and AOL
out of business, and take over the content provider market.
Microsoft has a monopoly in the operating systems market for the
desktop PC. They have regularly used this monopoly to push others out
of competitive markets. They have purchased the best of their
competition and destroyed the rest.
I hope the FTC does something about this soon. We are already
starting to feel the effects of the Microsoft monopoly. Software
quality is going down, software prices are increasing (marginally in
the initial price, but look at upgrade prices), support is at an all
time low.
It is effecting the entire market.
[Stuff Deleted]
>
>The topic's premise is flawed: people *HAD* choices. During the 90s,
>there were a *LOT* of choices. At one time, Microsoft, Lotus,
>WordPerfect, and Borland had roughly equivalent products. The
Not so, they had roughly equivalent applications that ran on a
Microsoft operating system. Wasn't it Borland that sued about
Microsoft hiding information about its API from their competition,
until it had an edge in the compiler market too?
>companies offering the other choices dropped the ball in various
>different ways while Microsoft continually improved its offerings.
Microsoft dropped the ball on many of these offerings as well, but
lucky for them, they started out way ahead. Microsoft will win the
'battle of the browsers', but it will have nothing to do with hard
work or having the best product. It will be because Microsoft
controls the desktop operating system. They can leverage one monopoly
to gain another.
>
>How come you always hear about Microsoft's "monopoly" being bad, but
>you never hear anyone complain IBM is the most incompetent PC company
>in history, Lotus was the worst managed company in the 90s, Borland was
>an aimless company with no plan (they had a database, and BOUGHT
>ANOTHER DATABASE!), etc? Microsoft hasn't *HAD* to do anything
>bad, all they've had to do is watch their so-called "competition"
>go down the tubes.
Come on.. everyone else was just incompetent? Microsoft just fell
into having a monopoly in the entire desktop software market?
Servers, Applications, Operating systems, Development systems? They
took it all over because every other software company in the world was
incompetent? Don't be ridiculous. The had an advantage and they used
it. I might add, they used it illegally. Leveraging one monopoly to
gain another (in the manner Microsoft has) is a violation of antitrust
laws in the US.
It is simply a matter of time and Microsoft will be split up or some
other reasonable resolution will come to pass. I hope it is soon.
Who knows, maybe some currently unknown threat will end the Microsoft
monopoly without government intervention. Maybe we shouldn't be
laughing at Web TV and network computers? Nah.
--
Ken Turvey <ktur...@pug1.SprocketShop.com>
(Running Linux)
P.S. As a side note, I should say that Microsoft does produce some of
the best software available. It is unfortunate, however, that they
are the only reasonable choice most people have.
From: "John Poole" <jfpo...@bulk.email.spoof.uwaterloo.ca>
Subject: Re: Microsoft monopoly threatens consumer buying choices
Date: 1997/08/15
Message-ID: <01bca982$07c5d6e0$8d396478@johnpo_x2>#1/1
X-Deja-AN: 264901289
References: <5sukls$lpm$1@news.netrail.net> <5svm9f$igb$1@scully.new-era.net>
<slrn5v75gb.tb.kturvey@pug1.sprocketshop.com>
Organization: Corel Corporation
Newsgroups: comp.lang.c,microsoft.public.vc.events,comp.lang.java.advocacy,
comp.lang.c++,comp.lang.java,symantec.support.devtools.pc.visual-cafe.java
Kenneth P. Turvey <ktur...@pug1.sprocketshop.com> wrote in article
<slrn5v75gb.tb.ktur...@pug1.sprocketshop.com>...
>
> I disagree. The reason there isn't any competition anymore isn't
> because Microsoft has done such a great job. It is because Microsoft
> owned the operating system and they could leverage that market to
> control the other markets you speak of. Why didn't Word Perfect win
> the battle of the word processors? They had a superior product. They
> had better support. Why does the windows Wordpad support the Word 6
> document format, and look suspiciously like Word 6? Because it shares
> code with Word 6, I am sure of it. Anyone in a software company that
> is competing head to head with Microsoft is praying they get bought
> instead of crushed. There isn't an alternative.
WordPerfect *wasn't* a superior product when it hit Windows (the DOS
version is another matter entirely). They did a half-assed job writing
(porting?) the Windows version of WordPerfect, and I think that's when they
started to lose users. Otoh, Word for Windows 2.0 was quite a nice product.
For once it was a superior product beating out an inferior one :)
[rest deleted]
--
jfpoole - {}
http://www.undergrad.math.uwaterloo.ca/~jfpoole
"When they start treating us for cynicism,
we'll have no reason to drink." -- suck.com
From: Alicia Carla Longstreet <ca...@ici.net>
Subject: Re: Microsoft monopoly threatens consumer buying choices
Date: 1997/08/30
Message-ID: <3408B3A9.4E55@ici.net>#1/1
X-Deja-AN: 269657190
References: <5sukls$lpm$1@news.netrail.net> <5svm9f$igb$1@scully.new-era.net>
<slrn5v75gb.tb.kturvey@pug1.sprocketshop.com> <01bca982$07c5d6e0$8d396478@johnpo_x2>
Reply-To: carlaNO_S...@ici.net
Organization: The Computer Solution
Newsgroups: comp.lang.c,microsoft.public.vc.events,comp.lang.java.advocacy,
comp.lang.c++,comp.lang.java,symantec.support.devtools.pc.visual-cafe.java
John Poole wrote:
>
> Kenneth P. Turvey <ktur...@pug1.sprocketshop.com> wrote in article
> <slrn5v75gb.tb.ktur...@pug1.sprocketshop.com>...
> >
> > I disagree. The reason there isn't any competition anymore isn't
> > because Microsoft has done such a great job. It is because Microsoft
> > owned the operating system and they could leverage that market to
> > control the other markets you speak of. Why didn't Word Perfect win
> > the battle of the word processors?
> > They had a superior product.
That is a matter of opinion. Actually, until version 5.x came along
WordPerfect could not really handle porportionally spaced fonts, output
was awful on version 4.3. WordPerfect also had several foolish
idiosyncracies. Like pressing page down jumped to the next page. It may
seem sensible but it was counter-intuitive since most every other
product on the market used page down to jump to the next screen page.
On top of that, no one could ever remember all of the command sequences,
you always needed *some* kind of help. Word had the memu system with
shortcuts, I could always use word without *any* aids to remember
command sequences.
> > They had better support. Why does the windows Wordpad support the Word 6
> > document format, and look suspiciously like Word 6? Because it shares
> > code with Word 6, I am sure of it.
WordPad uses the Microsoft Foundation Classes, which happen to be based
on code in Word, Excel, etc. This is why all of the Microsoft
applications have the same look and feel.
WordPad supports Word 6 file formats for the same reason the latest
version of WordPerfect supports them, because Microsoft makes the
formats available for a fee, just like WordPerfect makes their file
formats available for a fee. So Pagemaker, etc and support their
products.
The Microsoft Applications programming staff does not (and this has been
verified by the Feds) have any access to OS code that is not provided to
any other company producing Windows applications. In any event, Borland
has consistantly produced products that made the same excellent use of
OS resources and capabilities as Microsoft, proving the the info was
available if the companies wanted to look for it. Problem was,
WordPerfect, Lotus, etc didn't want to work for it, they just wanted
Microsoft to give them the data that Microsoft applications programmers
had researched.
> > Anyone in a software company that
> > is competing head to head with Microsoft is praying they get bought
> > instead of crushed. There isn't an alternative.
I was wondering, does anybody know of any company with a superior
product than Microsoft, or even an equal product, that got 'crushed'.
> WordPerfect *wasn't* a superior product when it hit Windows (the DOS
> version is another matter entirely). They did a half-assed job writing
> (porting?) the Windows version of WordPerfect, and I think that's when they
> started to lose users. Otoh, Word for Windows 2.0 was quite a nice product.
> For once it was a superior product beating out an inferior one :)
WordPerfect ported to the Macintosh and did a lousy job, then they
finally ported from the Mac to Windows. Their problem was they didn't
want to lose sales for their DOS product.
[snip]
--
************************************************
* Alicia Carla Longstreet ca...@ici.net *
************************************************
* Remove NO_SPAM when replying to me. *
************************************************
My compassion for someone is not limited
to my estimate of their intelligence.
Dr Gillian, Star Trek IV, The Voyage Home
************************************************
One often contradicts an opinion when
what is uncongenial is really the
tone in which it was conveyed.
Friedrich Nietzsche
From: Larry Brasfield <larr...@SsPqAiM.com>
Subject: MS's secret OS/App channel (was Re: Microsoft monopoly threatens ...)
Date: 1997/09/03
Message-ID: <340E008F.2781E494@SsPqAiM.com>#1/1
X-Deja-AN: 269793089
References: <5sukls$lpm$1@news.netrail.net> <5svm9f$igb$1@scully.new-era.net>
<slrn5v75gb.tb.kturvey@pug1.sprocketshop.com> <01bca982$07c5d6e0$8d396478@johnpo_x2>
<3408B3A9.4E55@ici.net> <340D8A10.2B41@world.std.com> <340DBF66.5C9B@se80.dseg.com>
Followup-To: comp.lang.java.advocacy
Organization: Siemens Medical Systems, Inc. - Ultrasound Group
Newsgroups: comp.lang.c,microsoft.public.vc.events,comp.lang.java.advocacy,
comp.lang.c++,comp.lang.java,symantec.support.devtools.pc.visual-cafe.java
Followups directed to comp.lang.java.advocacy, where
Microsoft is a subject more favored than Java.
Russ Lyttle wrote:
>
> David Chase wrote:
> >
> > Alicia Carla Longstreet wrote:
> >
> > > The Microsoft Applications programming staff does not (and this has been
> > > verified by the Feds) have any access to OS code that is not provided to
> > > any other company producing Windows applications.
> >
> > They eat in separate cafeterias, then? In my experience, interesting
> > information often travels through informal channels.
> >
> > --
> > David Chase, ch...@world.std.com
> Actually, if I remember the case correctly, the judge found Microsoft
> guilty of sharing insider info on code, and fined them big time. The
> (in)Justice Department then objected and had the judge removed. The new
> judge the reduced the fine to a few hundred k. Anyone have references to
> the case? I would like to look it up for an ethics class I am taking.
As a former Microsoft employee, I find this speculation
about OS/App communication pretty laughable. Russ's fantasy
has no factual basis that I know of, and I have been following
Microsoft's legal situation carefully (as a very interested
shareholder and options trader) for at least 5 years. I am
willing to bet that Russ cannot cite a case in any actual
U.S. court with findings resembling what he states above.
Later tonight, on my own time and equipment, I will respond
at greater length regarding the "secret channel" issue.
--
--Larry
work: (425)557-1670 larr...@SsPqAiM.com
home: (206)236-2121 larr...@eaSrtPhlAinMk.net
Aforementioned views are likely mine alone.
(Remove "SPAM" from address for an email reply.)
From: mellon+use...@pobox.com (Anatoly Vorobey)
Subject: Re: MS's secret OS/App channel (was Re: Microsoft monopoly threatens ...)
Date: 1997/09/05
Message-ID: <slrn61132q.30d.mellon+usenet@sasami.jurai.net>
X-Deja-AN: 270125397
References: <5sukls$lpm$1@news.netrail.net> <5svm9f$igb$1@scully.new-era.net>
<slrn5v75gb.tb.kturvey@pug1.sprocketshop.com> <01bca982$07c5d6e0$8d396478@johnpo_x2>
<3408B3A9.4E55@ici.net> <340D8A10.2B41@world.std.com> <340DBF66.5C9B@se80.dseg.com>
<340E008F.2781E494@SsPqAiM.com> <01bcb904$a4a6e4e0$8101019d@bit_blaster>
Organization: Erol's Internet Services
X-Received-On: 5 Sep 1997 22:49:07 GMT
Reply-To: mellon+use...@pobox.com
Newsgroups: comp.lang.java.advocacy
On 4 Sep 1997 07:48:16 GMT, Larry Brasfield <SpamGuard_larr...@earthlink.net>
wrote:
>
>I don't know about "the Feds" having verified such a thing, but I can
>state from personal experience that the O.S. sources are not generally
>available outside of the O.S. development/test groups.
This is, of course, patently false. Microsoft sells NT source
licenses for $1 million a year to companies and licenses NT
source for free to many universities. Maybe you meant "are not
generally available in Microsoft outside..." - that might
or might not be true. It's probably false though: the NT
DDK team and possibly SDK team almost certainly probably got access
to the sources. Now, let's just exclude these too...
>Further, there
>is no reason the other groups want or need to peruse those sources.
This isn't true either. NT DDK is impossible to create without
sources; Win32 SDK is problematic to create without sources.
There are many problems which have very slow or in other
ways inferior solutions using documented methods and very
fast solutions using undocumented ones. Two examples are
thunks in '95 and OS information APIs in NT.
>would be available if there was any reason to use them. The trouble
>is that "The Secret API's" is a myth without reason to be real.
This is also false. "Secret" APIs have been documented in the past.
One of the better sources are relevant Pietrek and Richter books
dealing with internals of Win31 and Win95. There are many
examples there of how applications that are not
part of the OS use these APIs.
Things are *much* better with Win32 in this respect than
they were with Win16. Win32 has a much better API for one
thing; and secondly, Win95 and NT internals are so much
different and are controlled by very different groups
so that sneaking the same secret API in would be
quite impractical technically.
>I am tempted to cite my personal experience with Application
>and Languages development at Microsoft to put the kabosh on
>the oft-repeated "secret channel" conjecture. I believe that
>if there was any real truth to it, I would have been able to
>get a glimpse of it. But nothing I could state along those
>lines would have the slightest chance of convincing those who
>love to latch onto and spout such conjecture.
It isn't a conjecture, Larry. It is a fact, which
very many programmers dealing with lower-level APIs
and messing with lower-level details of NT and 95 inevitably
discover. I am not saying this because I am a MS-hater. In
fact, although I'm by no means MS-lover, I recognize and
applaud the fact that there has been a tremendous improvement
in the area if hidden APIs from Win16 and DOS to Win32. However,
it's simply not true that they're gone. You can always
write me off as an "Ms-hater" of course...
I suspect I
>would be deemed ignorant, a liar, or both. So, instead, I
You're neither. But I suspect you might be ill-informed.
>will state why it makes no sense for Microsoft developers
>to attempt to exploit "The Secret API's" and I will pose a
>challenge for anybody to show any actual exploitation.
You said you were from Applications and Languages group, right?
Visual C++ versions 4.0, 4.1, 4.2 and 5.0 (maybe earlier ones
too, I can't remember) all include a program, pstat.exe, which
is installed only on NT, not on 95. This program uses
highly undocumented APIs from ntdll.dll (no API from this
DLL has ever been documented by Microsoft; it's actually the
most crucial DLL in NT) to print lots of information about
processes, threads, and general system info. Some of this
information is available by using documented APIs (special
registry keys); however, due to either deliberate mucking or
extremely faulty design (personally, I would immediately
fire the person who came up with it) these it is much much slower
than these undocumented APIs and provides much less information.
While we're at it, the whole mechanism by which Visual
C/C++ supports C and C++ exceptions depends on
undocumented OS interface of supporting exception
handlers chains. The interface is undocumented quite
deliberately; for example, the C runtime library
source lacks exactly the source assembly file which
is needed to support them (only .obj is provided). Noone
has made too much fuss over it, probably because
the mechanism is very simple and was easily reverse-engineered
by other compilers' authors.
About a year ago there was a story about special APIs
put into NT by Microsoft especially to greatly speedup
Web servers and similar applications. The MS web server
of course exploited the APIs, and Microsoft refused
to supply their details to Netscape, O'Reilly, and
other Win32 Web servers vendors. (MS eventually supplied
the details half a year later).
These are three random examples which immediately sprang
to my mind. There are more of these (especially Win95-related).
Some of them are documented in several books about 95 internals
available.
This is my answer to your challenge.
[argument about unwillingness of OS designers
to release undocumented APIs because it prevents further
OS changes snipped]
The argument is valid, but it misses one point. Some
of these APIs (for example, the convention that fs:[0]
holds a pointer to the first handler record in
exceptions handlers list on x86) are as "documented"
as others to the designers. In other words, they're
fixed and will not changed and designers consider
them valid and documented; but another authority in
MS prevents them from being released to the public.
Then you have an API which is as good as a documented
one (the OS team is committed to support it) and
is only available to MS teams.
>by using secret O.S. features, especially ones that
>have to be used so slyly as to be invisible to folks
>willing to examine DLL linkages.
Not necessarily. Many of these hidden APIs
are exported through (undocumented) DLLs such
as ntdll.dll, psapi.dll, (some of)Toolhelp-32
DLL on 95, and some more.
>As far as I'm concerned, "the secret API's" are no
>more than fodder for the same sort of conspiracy
>theorists that maintain aliens crashed at Roswell.
Your irony is misplaced. It seems that you are the
conspiracy theorist - you dismiss every mention
of hidden APIs as a product of "MS-hater" mindset.
Well, you're blind to the facts then.
>I read endless bald statements that Microsoft
>exploits this "secret channel", but I have yet
>to see anybody offer a shred of evidence that
>would suggest such exploitation to an unbiased
You had to avoid reading some of the
Windows internals books then. Would you also
claim that MS officials have never issued deliberate
lies about MS products (such as integration
of '95 and DOS etc.)? There are many examples of
such proven lies in these books as well.
>technical observer. Do the "secret channel"
>kooks think that Microsoft applications call
>directly into O.S. locations, bypassing the
>well-established DLL linkage conventions? Or
>is it done via mysterious software interrupts?
No and no. You're setting up a strawman
argument.
>Can anybody find a single connection between a
>Microsoft application and the O.S. that does
>not go thru a published API? Not yet. I guess
I have offered you three examples before. I could
probably offer another three easily, or
another thirteen after a few
days of working through books, articles and
monitoring all calls of some applications
(_especially_ in 95) but I do not get payed for this.
>and "Microsoft is evil enough to do it." Again, I
>chuckle at what passes for thinking by some folks.
Your patronizing is noted and not appreciated.
Apparently there is no place in your world
to developers who do not hate MS fanatically,
yet recognize its weaknesses and questionable
practics. If that is so, I suggest that you read
a good book on critical thinking.
>An interesting question for those who believe in the
>"secret channel" hypothesis is: If the secret API's
>are so valuable for application performance, and if
>Microsoft personnel are so immoral, why is there no
>active black market for that information? I would
You're evidently smart enough to answer that yourself.
Because only MS programmers can be sure the OS group won't
change the interfaces in the next OS release. Also,
they are rarely really critical to justify illegal black
market operations. When they _are_ critical, the company can
usually afford to pay a good programmer to
reverse-engineer them.
In fact, there are many companies whose products
need undocumented abilities of 95 or NT so desperately
that they're ready to risk having to redo it
when the next OS version comes out. A good portion
of my paycheck often comes from such companies, so
I would know.
Also (this isn't really relevant, but amusing) I hear
that NT sources are quite well distributed in
some circles of hackers' community.
--
Anatoly Vorobey,
mel...@pobox.com http://pobox.com/~mellon/
"Angels can fly because they take themselves lightly" - G.K.Chesterton
From: "Larry Brasfield" <larr...@earthlink.net>
Subject: Re: MS's secret OS/App channel (was Re: Microsoft monopoly threatens ...)
Date: 1997/09/13
Message-ID: <01bcc092$47648720$8101019d@bit_blaster>
X-Deja-AN: 272245478
References: <5sukls$lpm$1@news.netrail.net> <5svm9f$igb$1@scully.new-era.net>
<slrn5v75gb.tb.kturvey@pug1.sprocketshop.com> <01bca982$07c5d6e0$8d396478@johnpo_x2>
<3408B3A9.4E55@ici.net> <340D8A10.2B41@world.std.com> <340DBF66.5C9B@se80.dseg.com>
<340E008F.2781E494@SsPqAiM.com> <01bcb904$a4a6e4e0$8101019d@bit_blaster>
<slrn61132q.30d.mellon+usenet@sasami.jurai.net>
Organization: Phrenetic Endeavors, Limited
Reply-To: "Larry Brasfield" <l_brasfield@computer_org>
Newsgroups: comp.lang.java.advocacy
Anatoly Vorobey <mellon+use...@pobox.com> wrote in article
<slrn61132q.30d.mellon+use...@sasami.jurai.net>...
>
> On 4 Sep 1997 07:48:16 GMT, Larry Brasfield <SpamGuard_larr...@earthlink.net>
> wrote:
> >
> >I don't know about "the Feds" having verified such a thing, but I can
> >state from personal experience that the O.S. sources are not generally
> >available outside of the O.S. development/test groups.
>
> This is, of course, patently false. Microsoft sells NT source
> licenses for $1 million a year to companies and licenses NT
> source for free to many universities. Maybe you meant "are not
> generally available in Microsoft outside..." - that might
> or might not be true.
The latter is what I meant. If you look at the context which
immediately preceded your above quotation, you would see:
(written by the indubitable Ms. Longstreet)
> > > > The Microsoft Applications programming staff does not (and this has been
> > > > verified by the Feds) have any access to OS code that is not provided to
> > > > any other company producing Windows applications.
So, to be more precise and create a standalone statement:
I can state from personal experience that the O.S. sources are
not generally available to Microsoft developers outside of the
O.S. development/test groups.
In your mind that might or might not be true. That's fine.
I'm reporting my observations for those who have not put
me into the liar category. Do what you wish with it.
> It's probably false though: the NT
> DDK team and possibly SDK team almost certainly probably got access
> to the sources. Now, let's just exclude these too...
I suppose you could quibble about whether the phrase
"O.S. development/test groups" includes whoever it is
that produces the DDK's and SDK's. But since I was
addressing use of the sources by developers in other
groups, it hardly matters. So yes, let's exclude
them because it is not germane to the issue of the
"secret channel", not just to whittle away at what I
stated until it might be true.
(I like your "almost certainly probably" phrase. I'll
hoard that for when its time to confound someone. ;-)
> >Further, there
> >is no reason the other groups want or need to peruse those sources.
>
> This isn't true either. NT DDK is impossible to create without
> sources; Win32 SDK is problematic to create without sources.
The DDK and SDK are products of the same group that is
responsible for the O.S. I do not see the relevance of
this tangent to the "secret channel" issue. Are you
saying that because the SDK headers and API docs ship
with Visual C++ that the VC group is exploiting the
"secret channel"? As far as I know, that material has
always been available to other vendors of language
products under reasonable license terms. Would you
not agree that this removes that particular flow of
information from inclusion in the "secret channel"?
> There are many problems which have very slow or in other
> ways inferior solutions using documented methods and very
> fast solutions using undocumented ones. Two examples are
> thunks in '95 and OS information APIs in NT.
I thought thunks were a way to cross 16/32 boundaries.
They were well documented in MSDN long before Win95
was released. Is there an undocumented way to build
a faster thunk? You've got me baffled with this.
The OS information API's in NT are a more interesting
case. I'll address that where you touch upon it later.
> >would be available if there was any reason to use them. The trouble
> >is that "The Secret API's" is a myth without reason to be real.
>
> This is also false. "Secret" APIs have been documented in the past.
> One of the better sources are relevant Pietrek and Richter books
> dealing with internals of Win31 and Win95. There are many
> examples there of how applications that are not
> part of the OS use these APIs.
Richter? As in Jeffrey Richter? What titles? Are
you claiming Microsoft applications were dissected
there and shown to use unpublished API's? I have
read a few of Richter's books and managed to miss
seeing anything like that. I'm not so sure about
Pietrek's stuff, but I'll examine it if you would
be willing to give me a starting point. Until then,
I have to treat the claim as too vague to argue.
I am not claiming that there are no unpublished PI's.
What I mean by '"The Secret API's" is a myth' is that
there is no set of API's designated within Microsoft
as being available for safe use by applications and
language tools that are unpublished. On occasion, via
the same route that other vendors sometime create an
application that relies upon undocumented (or under-
documented) PI's, a Microsoft application may end
up using one. That is not the "myth" I refute. The
myth I keep seeing is the proposition that the O.S.
group has created API's with similar commitment and
documentation to published ones, but they are kept
secret and only published internally for the benefit
of Microsoft non-O.S. products, to give them unfair
competitive advantage. (I readily grant that such a
practice would be unfair and would support a breakup
of Microsoft if that was what it took to preclude
such unfair tactics.)
> Things are *much* better with Win32 in this respect than
> they were with Win16. Win32 has a much better API for one
> thing; and secondly, Win95 and NT internals are so much
> different and are controlled by very different groups
> so that sneaking the same secret API in would be
> quite impractical technically.
I would grant that certain situations which arose with
16-bit Windows are better ammunition for an argument.
The way that API evolved (accreted), there was a lot
more room for that sort of thing to happen. Whether
the actual incidents were part of a master plan or a
by-product of the usual chaos at Microsoft is a good
question, one I have no solid basis for addressing.
> >I am tempted to cite my personal experience with Application
> >and Languages development at Microsoft to put the kabosh on
> >the oft-repeated "secret channel" conjecture. I believe that
> >if there was any real truth to it, I would have been able to
> >get a glimpse of it. But nothing I could state along those
> >lines would have the slightest chance of convincing those who
> >love to latch onto and spout such conjecture.
>
> It isn't a conjecture, Larry. It is a fact, which
> very many programmers dealing with lower-level APIs
> and messing with lower-level details of NT and 95 inevitably
> discover.
In my lexicon, a conjecture by one person could be the
same word sequence as a statement of fact by another.
It all depends on how the "information" arose. And we
should be careful to distinguish criticisms of the API
documentation and the difficulty of dealing with under-
specified behavior from the "Secret API" concept. They
may border each other, or even overlap slightly, but it
clouds the issues to mix them together willy-nilly.
Just what "is a fact" is still pretty murky here.
> I am not saying this because I am a MS-hater. In
> fact, although I'm by no means MS-lover, I recognize and
> applaud the fact that there has been a tremendous improvement
> in the area if hidden APIs from Win16 and DOS to Win32. However,
> it's simply not true that they're gone. You can always
> write me off as an "Ms-hater" of course...
I don't worry about MS-hating or MS-loving among those
willing to deal in evidence and logic, and I try not to
deal with those who are not so willing. So there is not
much reason for me to write you off as anything, yet.
> I suspect I
> >would be deemed ignorant, a liar, or both. So, instead, I
>
> You're neither. But I suspect you might be ill-informed.
I guess I should acknowledge your kindness here. Thanks.
[ precursor to later challenge cut]
> You said you were from Applications and Languages group, right?
I almost implied that. Those are separate groups. I
worked on a component that is used by both of them.
> Visual C++ versions 4.0, 4.1, 4.2 and 5.0 (maybe earlier ones
> too, I can't remember) all include a program, pstat.exe, which
> is installed only on NT, not on 95. This program uses
> highly undocumented APIs from ntdll.dll (no API from this
> DLL has ever been documented by Microsoft; it's actually the
> most crucial DLL in NT) to print lots of information about
> processes, threads, and general system info. Some of this
> information is available by using documented APIs (special
> registry keys); however, due to either deliberate mucking or
> extremely faulty design (personally, I would immediately
> fire the person who came up with it) these it is much much slower
> than these undocumented APIs and provides much less information.
It is quite a stretch to treat pstat.exe (or pview95.exe)
as an example of "secret API" exploitation, especially
since it is available in a variety of places where it is
obvious that Microsoft is not trying to sell it. It is
logically part of the O.S. (In my view of things cyber,
it should always be installed with the O.S.) I find it
on my Windows NT CD-ROM. By its nature, it has to reach
into regions of the O.S. internals that have no other
reason to be exposed, and certainly not as a stable API.
Your comments about ntdll.dll indicate that you seriously
misunderstand its role. While it is indeed a crucial DLL
(equal in cruciality to several others), it provides the
foundation upon which the API-providing DLL's rest. Its
interfaces are strictly internal and change from release
to release. (but not every interface at every release)
> While we're at it, the whole mechanism by which Visual
> C/C++ supports C and C++ exceptions depends on
> undocumented OS interface of supporting exception
> handlers chains. The interface is undocumented quite
> deliberately; for example, the C runtime library
> source lacks exactly the source assembly file which
> is needed to support them (only .obj is provided). Noone
> has made too much fuss over it, probably because
> the mechanism is very simple and was easily reverse-engineered
> by other compilers' authors.
Let's treat C and C++ exceptions separately.
The SEH mechanism, first built into NT and long published
in the Win32 SDK, is what you must mean by "C exceptions."
If by "needed to support them", you mean to throw and
catch them, I cannot see what more is needed. In C, or
C-ish C++) they are caught, handled and possibly rethrown.
There is no reason to chain them.
The C++ exception mechanism is purely language supported
(meaning under the control of generated code among a
set of compatibly compiled and linked translation units)
except for the use of a known mechanism for getting at
thread-local storage. I guess you are asserting that this
mechanism has not been sufficiently documented to fall
into the "published API" category.
I think you may be technically right on that. Since I
cannot find a published promise that Tls{Alloc,SetValue,
GetValue,Free} can be safely inlined into applications,
but Visual C++ generates code as if they could, and if
we take as given that Win32 applications compiled with
Visual C++ will work on future versions of the O.S., it
appears that the Visual C++ code-generator authors have
special knowledge as to how stable the implementation
of thread-local storage is. A similar case can be made
with respect to MSVC's __declspec(thread) attribute.
Does this establish a pattern? Can it prop up the whole
"secret channel" conspiracy theory? I find it hard to
spot any evil intent here, or even a concerted effort
to create and exploit an unfair advantage. But clearly
Microsoft should make the promise of stability of that
mechanism public, and should have done so as soon as it
was determined that it was needed to support efficient
C++ exception handling.
> About a year ago there was a story about special APIs
> put into NT by Microsoft especially to greatly speedup
> Web servers and similar applications. The MS web server
> of course exploited the APIs, and Microsoft refused
> to supply their details to Netscape, O'Reilly, and
> other Win32 Web servers vendors. (MS eventually supplied
> the details half a year later).
I recall this controversy, (dimly), and my memory of it
is that the web server group developed a DLL, (founded
on the published API), to support repeated transactions
efficiently, and somebody decided it should be sold as
part of the operating system. Since they could have
kept it to themselves without exploiting "secret API's",
and since it is not a core O.S. component, it does not
qualify as "part of the O.S. tardily revealed".
> These are three random examples which immediately sprang
> to my mind. There are more of these (especially Win95-related).
> Some of them are documented in several books about 95 internals
> available.
>
> This is my answer to your challenge.
I doubt your marginal and idiosyncratic examples are what
people have in mind when they get all indignant about
exploitation of the "secret channel". I had asked:
Can anybody find a single connection between a
Microsoft application and the O.S. that does
not go thru a published API?
I grant that you have provided a "shred of evidence"
suggesting a trickle of information flow thru the
"secret channel", but I still see nothing to suggest
significant or unfair exploitation of it.
> [argument about unwillingness of OS designers
> to release undocumented APIs because it prevents further
> OS changes snipped]
>
> The argument is valid, but it misses one point. Some
> of these APIs (for example, the convention that fs:[0]
> holds a pointer to the first handler record in
> exceptions handlers list on x86) are as "documented"
> as others to the designers. In other words, they're
> fixed and will not changed and designers consider
> them valid and documented; but another authority in
> MS prevents them from being released to the public.
> Then you have an API which is as good as a documented
> one (the OS team is committed to support it) and
> is only available to MS teams.
Well, now that MSVC has relied on the underpinnings of
TlsWhatnot(), they are probably "fixed" as you say,
but I would claim they are available to non-MS teams
also. (That is a tardy availability, IMO.)
I think your point requires a modification of my
argument. In its simple form, it addressed random
use of unpublished interfaces. Things are a little
more complicated near the surface of the API set,
where nominally internal features not originally
intended to become part of the API are later added
to it, with the attendant documentation and ongoing
commitment. My argument would seem to apply to this
creep, since I did not distinguish it. The issues
cannot be so simply stated in that boundary zone.
> >by using secret O.S. features, especially ones that
> >have to be used so slyly as to be invisible to folks
> >willing to examine DLL linkages.
>
> Not necessarily. Many of these hidden APIs
> are exported through (undocumented) DLLs such
> as ntdll.dll, psapi.dll, (some of)Toolhelp-32
> DLL on 95, and some more.
There are many hidden PI's, but no hidden API's.
I think you would learn from Andrew Schulman's
discussion of these issues in "Undocumented DOS"
(Addison Wesley, 1990) There is no reason for
applications to link against ntdll.dll, and I
have yet to see a Microsoft application do so.
> >As far as I'm concerned, "the secret API's" are no
> >more than fodder for the same sort of conspiracy
> >theorists that maintain aliens crashed at Roswell.
>
> Your irony is misplaced. It seems that you are the
> conspiracy theorist - you dismiss every mention
> of hidden APIs as a product of "MS-hater" mindset.
> Well, you're blind to the facts then.
I do not see any conspiracy. (I see the conjecture
more as urban legend that best takes root in the
minds of straw-grasping MS-haters.) You overstate
the case with your "every mention". I do dismiss
conjecture without credible evidence. But I am not
about to consider that as "blind to the facts." I
am open to real evidence. What you have provided
is evidence of something, but I do not see it as
going to the heart of the controversy. With bias,
some folks may see it as the tip of "the" iceberg,
but I have not seen any iceberg yet. Is that blind?
> >I read endless bald statements that Microsoft
> >exploits this "secret channel", but I have yet
> >to see anybody offer a shred of evidence that
> >would suggest such exploitation to an unbiased
>
> You had to avoid reading some of the
> Windows internals books then. Would you also
> claim that MS officials have never issued deliberate
> lies about MS products (such as integration
> of '95 and DOS etc.)? There are many examples of
> such proven lies in these books as well.
Actually, I have followed the "Is Win95 really
DOS in disguise?" tizzy with great interest. I
think the whole issue is complex enough that it
is inappropriate to label a few simplifications
of it uttered by MS "officials" as deliberate
lies. I have concentrated mainly on technical
descriptions and overviews published by MS-Press
and I see no "lies" there.
But let us agree to disagree there. That subject
could consume many thousands of words and never
get anywhere, as has already been demonstrated.
> >technical observer. Do the "secret channel"
> >kooks think that Microsoft applications call
> >directly into O.S. locations, bypassing the
> >well-established DLL linkage conventions? Or
> >is it done via mysterious software interrupts?
>
> No and no. You're setting up a strawman
> argument.
You misunderstand my point. I will clarify it.
If these alleged illicit linkages go thru the
normal mechanisms, they should be easy to find.
To postulate other linkages is so far out on the
fringe as to be in kook territory. Why have we
not seen solid, verifiable evidence of illicit
linkages of the expectable (DLL) kind when they
are easy to find? Because they do not exist.
> >Can anybody find a single connection between a
> >Microsoft application and the O.S. that does
> >not go thru a published API? Not yet. I guess
>
> I have offered you three examples before. I could
> probably offer another three easily, or
> another thirteen after a few
> days of working through books, articles and
> monitoring all calls of some applications
> (_especially_ in 95) but I do not get payed for this.
I guess I will have to accept your word on the
count. I certainly have to reserve judgment as
to the import of whatever you are counting.
> >and "Microsoft is evil enough to do it." Again, I
> >chuckle at what passes for thinking by some folks.
>
> Your patronizing is noted and not appreciated.
> Apparently there is no place in your world
> to developers who do not hate MS fanatically,
> yet recognize its weaknesses and questionable
> practics. If that is so, I suggest that you read
> a good book on critical thinking.
I patronize conjecturers everywhere who spout their
fantasy as truth when it represents nothing more
than an emotionally-pleasing excuse to believe what
they want to believe for their own purposes. I dare
not speculate upon those; I am not a psychologist.
If you think my low esteem of conjecturers ought
to apply to you, then I am sorry, for you.
In fact, I have worked with developers inside and
outside of Microsoft who recognize various poor
practices, and I share some of those opinions. I
have attempted to develop critical thinking skills
throughout my life. If you think you have a good
reading suggestion in that realm, please share it.
> >An interesting question for those who believe in the
> >"secret channel" hypothesis is: If the secret API's
> >are so valuable for application performance, and if
> >Microsoft personnel are so immoral, why is there no
> >active black market for that information? I would
>
> You're evidently smart enough to answer that yourself.
> Because only MS programmers can be sure the OS group won't
> change the interfaces in the next OS release. Also,
> they are rarely really critical to justify illegal black
> market operations. When they _are_ critical, the company can
> usually afford to pay a good programmer to
> reverse-engineer them.
I was suggesting that MS personnel in a position
to know some hidden PI is actually a hidden API,
(whether they are programmers or not), should be
able to sell that information if it is valuable.
The issue is not reverse-engineering but knowing
which "secret API's" are stable enough that they
ought to have been published.
Remember, the allegation is that Microsoft apps
gain significant (or even critical) competitive
advantages by using stable hidden API's. You do
not get to defuse my point by going into API's
that do not offer significant advantages. (You
would agree, I think, that using hidden API's
for trifling advantages would be stupid for the
"flexibility" reasons I mentioned earlier.)
I admit ignorance regarding the exact economic
tradeoffs between buying illicit information
and reverse-engineering. But I would think
that providing information that should have
been published would be safer from a legal
standpoint. (Even getting fired for letting
that sort of information out would create an
interesting, big-money legal situation.)
> In fact, there are many companies whose products
> need undocumented abilities of 95 or NT so desperately
> that they're ready to risk having to redo it
> when the next OS version comes out. A good portion
> of my paycheck often comes from such companies, so
> I would know.
I guess there will always be software relying
on unspecified behavior. So what? How does
that bear on what Microsoft does?
> Also (this isn't really relevant, but amusing) I hear
> that NT sources are quite well distributed in
> some circles of hackers' community.
That is amusing and relevant, I think. Would not
the availability of those sources improve the odds
of someone antagonistic towards Microsoft being
able to expose more of the alleged shenanigans?
> Anatoly Vorobey,
> mel...@pobox.com http://pobox.com/~mellon/
--
-- Larry Brasfield
The aforementioned views are mine alone.
(Convert under to dot for e-mail reply.)
From: jer...@netcom.com (Jeremy Allison)
Subject: Re: MS's secret OS/App channel (was Re: Microsoft monopoly threatens ...)
Date: 1997/09/06
Message-ID: <jeremyEG2ABL.EGF@netcom.com>#1/1
X-Deja-AN: 270097018
Sender: jer...@netcom13.netcom.com
References: <5sukls$lpm$1@news.netrail.net> <5svm9f$igb$1@scully.new-era.net>
<slrn5v75gb.tb.kturvey@pug1.sprocketshop.com> <01bca982$07c5d6e0$8d396478@johnpo_x2>
<3408B3A9.4E55@ici.net> <340D8A10.2B41@world.std.com> <340DBF66.5C9B@se80.dseg.com>
<340E008F.2781E494@SsPqAiM.com> <01bcb904$a4a6e4e0$8101019d@bit_blaster>
Organization: Netcom On-Line Services
Newsgroups: comp.lang.java.advocacy
"Larry Brasfield" <SpamGuard_larr...@earthlink.net> writes:
>Can anybody find a single connection between a
>Microsoft application and the O.S. that does
>not go thru a published API? Not yet. I guess
>that is too much work. It is far easier and a
>lot more satisfying to just (mindlessly) repeat
>the "secret channel" conjecture rather than try
>to build it upon a factual foundation.
Larry,
Under Windows NT 3.1 and 3.5 one of the most requested
APIs was that which enabled an application to take
a users name and plaintext password, and allow
that application to determine if the plaintext
password was valid. Many people wanted such an API
to allow such things as telnet daemons to be ported
to Windows NT - I was one of them. My need was for
a commercial vendor who wished to integrate into
the Windows NT security system so we did not have
to maintain our own password database.
Microsofts position on this was that such APIs
were available interally to the OS group but not
generally to application developers, as they were
not yet 'finished' or it had not been completely
decided how this was to be made available to
external developers.
Fair enough, I thought - at least none of the
competition can use this functionality.
Then MS SQL Server 4.21 came out.
Lo and behold - it was *completely*
integrated into the Windows NT security
system, once a user had logged onto NT,
they did not have to log on again in order
to be authenticated to SQL server. It was
using the very functions that the MS OS
group had told everyone were 'not ready'
(LsaLogonUser, if you really want to know).
So, are you going to tell me that SQL Server
4.21 was 'part of the OS group' ?
This is a perfect example of the 'secret
channel' between Microsoft applications
group and OS group.
I'm sure there are others, but this is
the only one I have personal knowledge of.
Jeremy Allison.
jer...@netcom.com
From: Matti S Poutanen <pouta...@beta.hut.fi>
Subject: Re: MS's secret OS/App channel (was Re: Microsoft monopoly threatens ...)
Date: 1997/09/15
Message-ID: <mcrlo0y321h.fsf@beta.hut.fi>#1/1
X-Deja-AN: 272728032
References: <5sukls$lpm$1@news.netrail.net> <5svm9f$igb$1@scully.new-era.net>
<slrn5v75gb.tb.kturvey@pug1.sprocketshop.com> <01bca982$07c5d6e0$8d396478@johnpo_x2>
<3408B3A9.4E55@ici.net> <340D8A10.2B41@world.std.com> <340DBF66.5C9B@se80.dseg.com>
<340E008F.2781E494@SsPqAiM.com> <01bcb904$a4a6e4e0$8101019d@bit_blaster>
<slrn61132q.30d.mellon+usenet@sasami.jurai.net> <01bcc092$47648720$8101019d@bit_blaster>
Distribution: inet
Organization: Helsinki University of Technology, Finland
Newsgroups: comp.lang.java.advocacy
Larry Brasfield wrote:
>up using one. That is not the "myth" I refute. The
>myth I keep seeing is the proposition that the O.S.
>group has created API's with similar commitment and
>documentation to published ones, but they are kept
>secret and only published internally for the benefit
>of Microsoft non-O.S. products, to give them unfair
>competitive advantage. (I readily grant that such a
>practice would be unfair and would support a breakup
>of Microsoft if that was what it took to preclude
>such unfair tactics.)
Larry, I remember reading an article at www.ntinternals.com about
APIs that appeard in NT4 Service Pack 2 (?). The APIs were fast
disk IO functions that supported scatter loading/saveing. The authors
were positive that these APIs were put into place for one MS product:
MS SQL server.
--
Matti Poutanen | EMAIL: matti.pouta...@hut.fi
Student of Physics | "And on a mission over China, the lady opens up her
Helsinki University | arms. The flowers bloom where you have placed them,
of Technology Finland, Europe| and the lady smiles just like mom."
From: ken...@nospam.lyapunov.ucsd.edu (Matt Kennel (Remove 'NOSPAM' to reply))
Subject: Re: MS's secret OS/App channel (was Re: Microsoft monopoly threatens ...)
Date: 1997/09/15
Message-ID: <slrn61rff4.317.kennel@lyapunov.ucsd.edu>#1/1
X-Deja-AN: 272784081
References: <5sukls$lpm$1@news.netrail.net> <5svm9f$igb$1@scully.new-era.net>
<slrn5v75gb.tb.kturvey@pug1.sprocketshop.com> <01bca982$07c5d6e0$8d396478@johnpo_x2>
<3408B3A9.4E55@ici.net> <340D8A10.2B41@world.std.com> <340DBF66.5C9B@se80.dseg.com>
<340E008F.2781E494@SsPqAiM.com> <01bcb904$a4a6e4e0$8101019d@bit_blaster>
<slrn61132q.30d.mellon+usenet@sasami.jurai.net> <01bcc092$47648720$8101019d@bit_blaster>
<mcrlo0y321h.fsf@beta.hut.fi>
Organization: University of California at San Diego
Reply-To: ken...@NOSPAMlyapunov.ucsd.edu
Newsgroups: comp.lang.java.advocacy
On 15 Sep 1997 20:44:10 +0300, Matti S Poutanen <pouta...@beta.hut.fi> wrote:
:
:Larry Brasfield wrote:
:
:>up using one. That is not the "myth" I refute. The
:>myth I keep seeing is the proposition that the O.S.
:>group has created API's with similar commitment and
:>documentation to published ones, but they are kept
:>secret and only published internally for the benefit
:>of Microsoft non-O.S. products, to give them unfair
:>competitive advantage. (I readily grant that such a
:>practice would be unfair and would support a breakup
:>of Microsoft if that was what it took to preclude
:>such unfair tactics.)
:
:Larry, I remember reading an article at www.ntinternals.com about
:APIs that appeard in NT4 Service Pack 2 (?). The APIs were fast
:disk IO functions that supported scatter loading/saveing. The authors
:were positive that these APIs were put into place for one MS product:
:MS SQL server.
Let's all remember that it matters NOT whether the 'applications' group
knows about secret API's that others may not, but that the applications
group can *ASK* the operating systems group for API's, even public ones,
which just happen to be perfect for the new release of their applications.
(And they get to beta-test them as secret API's).
Their request will be taken much more seriously than anybody else's.
Furthermore, a request by an application competitor to add any particular OS
feature can be used as 'tactical intelligence' and passed on to the
applications group in order to discern their competitor might be planning on
doing. And then the feature rejected, or more likely, delayed until the
MS version (as they have a version of *EVERYTHING* now) is relased with
a hyped-to-be-equivalent feature.
--
* Matthew B. Kennel/Institute for Nonlinear Science, UCSD
*
* According to California Assembly Bill 3320, it is now a criminal offense
* to solicit any goods or services by email to a CA resident without
* providing the business's legal name and complete street address.
*
From: "Carlos Blanco" <my.email.is.carlo...@microsoft.com>
Subject: Re: MS's secret OS/App channel (was Re: Microsoft monopoly threatens ...)
Date: 1997/09/16
Message-ID: <5vn7bv$a30@news.microsoft.com>#1/1
X-Deja-AN: 273097627
References: <5sukls$lpm$1@news.netrail.net> <5svm9f$igb$1@scully.new-era.net>
<slrn5v75gb.tb.kturvey@pug1.sprocketshop.com> <01bca982$07c5d6e0$8d396478@johnpo_x2>
<3408B3A9.4E55@ici.net> <340D8A10.2B41@world.std.com> <340DBF66.5C9B@se80.dseg.com>
<340E008F.2781E494@SsPqAiM.com> <01bcb904$a4a6e4e0$8101019d@bit_blaster>
<slrn61132q.30d.mellon+usenet@sasami.jurai.net> <01bcc092$47648720$8101019d@bit_blaster>
<mcrlo0y321h.fsf@beta.hut.fi> <slrn61rff4.317.kennel@lyapunov.ucsd.edu>
X-MimeOLE: Produced By Microsoft MimeOLE V4.71.1712.3
Organization: Microsoft
Newsgroups: comp.lang.java.advocacy
>Furthermore, a request by an application competitor to add any particular
OS
>feature can be used as 'tactical intelligence' and passed on to the
>applications group in order to discern their competitor might be planning
on
>doing. And then the feature rejected, or more likely, delayed until the
>MS version (as they have a version of *EVERYTHING* now) is relased with
>a hyped-to-be-equivalent feature.
No. I guarantee you that the group that interfaces between 3rd parties and
the system services groups (as well as the systems services groups
themselves) are very careful not to propagate 3rd party information onto
applications groups. In many cases there are physical legal contracts that
explicitly prevent and delineate this kind of behavior in both directions.
--Carlos (carlosbl at microsoft.com)
From: jer...@netcom.com (Jeremy Allison)
Subject: Re: MS's secret OS/App channel (was Re: Microsoft monopoly threatens ...)
Date: 1997/09/20
Message-ID: <jeremyEGs7rr.F47@netcom.com>#1/1
X-Deja-AN: 274153800
Sender: jer...@netcom13.netcom.com
References: <5sukls$lpm$1@news.netrail.net> <5svm9f$igb$1@scully.new-era.net>
<slrn5v75gb.tb.kturvey@pug1.sprocketshop.com> <01bca982$07c5d6e0$8d396478@johnpo_x2>
<3408B3A9.4E55@ici.net> <340D8A10.2B41@world.std.com> <340DBF66.5C9B@se80.dseg.com>
<340E008F.2781E494@SsPqAiM.com> <01bcb904$a4a6e4e0$8101019d@bit_blaster>
<slrn61132q.30d.mellon+usenet@sasami.jurai.net> <01bcc092$47648720$8101019d@bit_blaster>
<mcrlo0y321h.fsf@beta.hut.fi> <slrn61rff4.317.kennel@lyapunov.ucsd.edu>
<5vn7bv$a30@news.microsoft.com> <3421E222.43E0@earthlink.net>
Organization: Netcom On-Line Services
Newsgroups: comp.lang.java.advocacy
Carlos Blanco wrote:
> No. I guarantee you that the group that interfaces between 3rd parties and
> the system services groups (as well as the systems services groups
> themselves) are very careful not to propagate 3rd party information onto
> applications groups. In many cases there are physical legal contracts that
> explicitly prevent and delineate this kind of behavior in both directions.
>
> --Carlos (carlosbl at microsoft.com)
What I really find funny is that this whole thread started
with someone who claimed there was no 'secret channel'
between MS OS & Apps groups challenging people to come
up with examples.
And indeed they did (up to 5 including my own if I
haven't lost count).
The original poster went very quiet at that point :-).
And now we find someone from MS repeating the same
old stuff.....
Carlos, it won't become true just because you say it
a lot. You're talking to software *developers* here,
people who have to deal with the MS-NDA-API-of-the-week.
We *know* what we're talking about so please
don't come out with the same 'official' statements
you have to give to the DoJ. We know better. :-).
Jeremy Allison.
jer...@netcom.com
From: "Carlos Blanco" <my.email.is.carlo...@microsoft.com>
Subject: Re: MS's secret OS/App channel (was Re: Microsoft monopoly threatens ...)
Date: 1997/09/24
Message-ID: <60ca5q$59p@news.microsoft.com>#1/1
X-Deja-AN: 275301135
References: <5sukls$lpm$1@news.netrail.net> <5svm9f$igb$1@scully.new-era.net>
<slrn5v75gb.tb.kturvey@pug1.sprocketshop.com> <01bca982$07c5d6e0$8d396478@johnpo_x2>
<3408B3A9.4E55@ici.net> <340D8A10.2B41@world.std.com> <340DBF66.5C9B@se80.dseg.com>
<340E008F.2781E494@SsPqAiM.com> <01bcb904$a4a6e4e0$8101019d@bit_blaster>
<slrn61132q.30d.mellon+usenet@sasami.jurai.net> <01bcc092$47648720$8101019d@bit_blaster>
<mcrlo0y321h.fsf@beta.hut.fi> <slrn61rff4.317.kennel@lyapunov.ucsd.edu>
<5vn7bv$a30@news.microsoft.com> <3421E222.43E0@earthlink.net>
<jeremyEGs7rr.F47@netcom.com>
X-MimeOLE: Produced By Microsoft MimeOLE V4.71.1712.3
Organization: Microsoft
Newsgroups: comp.lang.java.advocacy
>What I really find funny is that this whole thread started
>with someone who claimed there was no 'secret channel'
>between MS OS & Apps groups challenging people to come
>up with examples.
>
>And indeed they did (up to 5 including my own if I
>haven't lost count).
>
>The original poster went very quiet at that point :-).
>
>And now we find someone from MS repeating the same
>old stuff.....
>
>Carlos, it won't become true just because you say it
>a lot. You're talking to software *developers* here,
>people who have to deal with the MS-NDA-API-of-the-week.
>
>We *know* what we're talking about so please
>don't come out with the same 'official' statements
>you have to give to the DoJ. We know better. :-).
With all due respect, I *do* know better. This was about propagating
confidential 3rd party info into app groups. Which is not true. And I
know because, I've been a contact to several companies and have not
propagated their confidential info. And neither does anyone I know, and I
know and have known just about all the people who've worked in that area
over the past few years.
--Carlos (carlosbl at microsoft.com)
From: jer...@netcom.com (Jeremy Allison)
Subject: Re: MS's secret OS/App channel (was Re: Microsoft monopoly threatens ...)
Date: 1997/09/25
Message-ID: <jeremyEH2vMz.Fp4@netcom.com>#1/1
X-Deja-AN: 275513631
Sender: jer...@netcom13.netcom.com
References: <5sukls$lpm$1@news.netrail.net> <5svm9f$igb$1@scully.new-era.net>
<slrn5v75gb.tb.kturvey@pug1.sprocketshop.com> <01bca982$07c5d6e0$8d396478@johnpo_x2>
<3408B3A9.4E55@ici.net> <340D8A10.2B41@world.std.com> <340DBF66.5C9B@se80.dseg.com>
<340E008F.2781E494@SsPqAiM.com> <01bcb904$a4a6e4e0$8101019d@bit_blaster>
<slrn61132q.30d.mellon+usenet@sasami.jurai.net>
<01bcc092$47648720$8101019d@bit_blaster> <mcrlo0y321h.fsf@beta.hut.fi>
<slrn61rff4.317.kennel@lyapunov.ucsd.edu> <5vn7bv$a30@news.microsoft.com>
<3421E222.43E0@earthlink.net> <jeremyEGs7rr.F47@netcom.com>
<60ca5q$59p@news.microsoft.com>
Organization: Netcom On-Line Services
Newsgroups: comp.lang.java.advocacy
"Carlos Blanco" <my.email.is.carlo...@microsoft.com> writes:
>With all due respect, I *do* know better. This was about propagating
>confidential 3rd party info into app groups. Which is not true. And I
>know because, I've been a contact to several companies and have not
>propagated their confidential info. And neither does anyone I know, and I
>know and have known just about all the people who've worked in that area
>over the past few years.
>--Carlos (carlosbl at microsoft.com)
This was *not* about propagating confidential 3rd party info into
app groups. Please re-read the thread. I believe you when you
say that propagating 3rd party info into app groups is not done
by Microsoft. I also don't care about that !
This is about the secret OS API's that Microsoft uses *internaly*
to give it's own apps group an advantage over external apps.
Several examples of these were given in this thread (one from
me).
Could you please address these *specific* examples (given
previously in this thread) of secret Microsoft OS API's used
internaly by Microsoft apps groups (my example was the
Windows NT LogonUser API used by the MS-SQLServer team before
it was publicly available to external developers).
This is precisely the practivce that Microsoft has denied
for many years now. Having seen it in action *personally*
that is why I said I 'know better'.
If you could address these particular issues I'd be very
grateful.
If you don't respond then I would say that my original
accusation stands.
Regards,
Jeremy Allison,
jer...@netcom.com
From: "Carlos Blanco" <my.email.is.carlo...@microsoft.com>
Subject: Re: MS's secret OS/App channel (was Re: Microsoft monopoly threatens ...)
Date: 1997/10/01
Message-ID: <60v1sk$hct@news.microsoft.com>#1/1
X-Deja-AN: 277197815
References: <5sukls$lpm$1@news.netrail.net> <5svm9f$igb$1@scully.new-era.net>
<slrn5v75gb.tb.kturvey@pug1.sprocketshop.com> <01bca982$07c5d6e0$8d396478@johnpo_x2>
<3408B3A9.4E55@ici.net> <340D8A10.2B41@world.std.com> <340DBF66.5C9B@se80.dseg.com>
<340E008F.2781E494@SsPqAiM.com> <01bcb904$a4a6e4e0$8101019d@bit_blaster>
<slrn61132q.30d.mellon+usenet@sasami.jurai.net>
<01bcc092$47648720$8101019d@bit_blaster> <mcrlo0y321h.fsf@beta.hut.fi>
<slrn61rff4.317.kennel@lyapunov.ucsd.edu> <5vn7bv$a30@news.microsoft.com>
<3421E222.43E0@earthlink.net> <jeremyEGs7rr.F47@netcom.com>
<60ca5q$59p@news.microsoft.com> <jeremyEH2vMz.Fp4@netcom.com>
X-MimeOLE: Produced By Microsoft MimeOLE V4.71.1712.3
Organization: Microsoft
Newsgroups: comp.lang.java.advocacy
Jeremy Allison wrote in message ...
>This was *not* about propagating confidential 3rd party info into
>app groups. Please re-read the thread.
I disagree on this--see my very first, original reply. It was directed at
the following statement that I'm including below:
>Furthermore, a request by an application competitor to add any particular
OS
>feature can be used as 'tactical intelligence' and passed on to the
>applications group in order to discern their competitor might be planning
on
>doing. And then the feature rejected, or more likely, delayed until the
>MS version (as they have a version of *EVERYTHING* now) is relased with
>a hyped-to-be-equivalent feature.
Regardless, to address your other point:
>Could you please address these *specific* examples (given
>previously in this thread) of secret Microsoft OS API's used
>internaly by Microsoft apps groups (my example was the
>Windows NT LogonUser API used by the MS-SQLServer team before
>it was publicly available to external developers).
>
>If you could address these particular issues I'd be very
>grateful.
We have an entire group of people whose only job is to make sure
information is properly propagated from the OS groups to 3rd parties. If
they didn't do this in the case you mention, both the OS people and this
group of people screwed up, and I'd have to agree that you have a valid
complaint. The OS groups don't necessarily have an obligation to propagate
information, but they are well aware that it's in their best interests to
do so and have acted accordingly in the cases I've been involved with
personally.
--Carlos (carlosbl at microsoft.com)
From: jer...@netcom.com (Jeremy Allison)
Subject: Re: MS's secret OS/App channel (was Re: Microsoft monopoly threatens ...)
Date: 1997/10/07
Message-ID: <jeremyEHpEqy.JsF@netcom.com>#1/1
X-Deja-AN: 278546244
Sender: jer...@netcom13.netcom.com
References: <5sukls$lpm$1@news.netrail.net> <5svm9f$igb$1@scully.new-era.net>
<slrn5v75gb.tb.kturvey@pug1.sprocketshop.com> <01bca982$07c5d6e0$8d396478@johnpo_x2>
<3408B3A9.4E55@ici.net> <340D8A10.2B41@world.std.com> <340DBF66.5C9B@se80.dseg.com>
<340E008F.2781E494@SsPqAiM.com> <01bcb904$a4a6e4e0$8101019d@bit_blaster>
<slrn61132q.30d.mellon+usenet@sasami.jurai.net>
<01bcc092$47648720$8101019d@bit_blaster> <mcrlo0y321h.fsf@beta.hut.fi>
<slrn61rff4.317.kennel@lyapunov.ucsd.edu> <5vn7bv$a30@news.microsoft.com>
<3421E222.43E0@earthlink.net> <jeremyEGs7rr.F47@netcom.com>
<60ca5q$59p@news.microsoft.com> <jeremyEH2vMz.Fp4@netcom.com>
<60v1sk$hct@news.microsoft.com>
Organization: Netcom On-Line Services
Newsgroups: comp.lang.java.advocacy
"Carlos Blanco" <my.email.is.carlo...@microsoft.com> writes:
>Regardless, to address your other point:
>>Could you please address these *specific* examples (given
>>previously in this thread) of secret Microsoft OS API's used
>>internaly by Microsoft apps groups (my example was the
>>Windows NT LogonUser API used by the MS-SQLServer team before
>>it was publicly available to external developers).
>>
>>If you could address these particular issues I'd be very
>>grateful.
>We have an entire group of people whose only job is to make sure
>information is properly propagated from the OS groups to 3rd parties. If
>they didn't do this in the case you mention, both the OS people and this
>group of people screwed up, and I'd have to agree that you have a valid
>complaint. The OS groups don't necessarily have an obligation to propagate
>information, but they are well aware that it's in their best interests to
>do so and have acted accordingly in the cases I've been involved with
>personally.
>--Carlos (carlosbl at microsoft.com)
Thanks Carlos. I agree - a group of people screwed up, and
they did so in a way that gave an uncompetitive advantage
to Microsoft's applications division over 3rd party application
developers. This is more than a little suspicious.
That's why I have a 'valid complaint' and *thats* why
I (and many other application developers for MS Windows)
want to see Microsoft broken into an OS division and an
Applications division who can only communicate by *published*
API's. I hope you'd agree that this is called for. If it
were just a case of this happening once, then it would
be a valid mistake, unfortunately this happens too
often to be just 'a bunch of people screwing up'.
I have no animosity towards Microsoft's success, whenever
I get to compete with their App groups fairly. It's when
the OS division stacks the deck in their favour that I get
annoyed.
Jeremy Allison
jer...@netcom.com