Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Win63K

0 views
Skip to first unread message

Joe Bednorz

unread,
Feb 17, 2000, 3:00:00 AM2/17/00
to
(Additions & corrections welcome, I don't feel I really did the
subject justice.)
Original at: <http://www.techweb.com/wire/story/TWB20000217S0008>

>Microsoft's development team is fuming about a published report
>stating that Windows 2000 contains more than 63,000 bugs.

Especially because the source for the report is a company with
little credibility in the computing industry...Microsoft.


>This Friday, as Microsoft officially releases Windows
>2000, group vice president Jim Allchin is expected to
>issue a statement reassuring the public that Windows
>2000 is not bug-infested, Microsoft developers said
>Wednesday at the Windows 2000 Conference and Expo.

Translation: "Sacrificial veep...excuse me, GROUP veep,
that no-one ever heard from before (or will again) is expected
to lie at the Win2K Conference & Expo."

Note that this prediction makes it unnecessary for the unknown
veep...<ahem> GROUP veep to actually tell the lie.

Note also that the unknown vee...GROUP veep was only going
to "reassure the public,", not deny the report (whose source
was who again, Mr. Group Veep?)

>The report contains excerpts from an internal
>Microsoft memo stating that Windows 2000 contains
>more than 63,000 problems, and roughly one-third of
>them are potential bugs. However, Microsoft developers
>said the memo is exaggerated.

After all, it came from us, and we're notorious liars.


>"It's a complete mischaracterization," said John Gray, a
>Windows 2000 developer who said none of the
>reported 20,000 potential "real problems" outlined in the
>Microsoft memo will cause system crashes or
>blue-screen fiascos.

Although the non-"real problems" might. The "real problems"
will just cause illegal operations, program freezes, the ever
popular system hangs, smoke to emit from your computer, or
large angular demonic forms with huge fangs dripping unholy
ichor to rip your face off. But no crashes or BSOD's.


>"I'm absolutely positive [of this].
>That report doesn't represent a bug list."

Right. It's nowhere near complete since it just lists KNOWN
bugs...63,000 of them. (That's 172 bugs/day for a *year*, folks.)

And you can trust him. He works for the same company that
compiled the bug list.

>Another developer said some issues remain with the code,

Like who's going to be blamed for this, and how is he going to
get a new job after this fiasco.

> but said end users would not run into problems.

Trip, stagger & fall maybe, but not run, not with Win2K.
(Excuse ME!) Win63K.


>He said Microsoft's internal Prefix tool, which identified the
>63,000 items, often flags inconsequential issues such as
>misspellings and sets off false alarms.

Misspellings in code are inconsequential? It's okay because
their applications are too buggy to be trusted?


>"There are still some low-priority issues, but we believe people
>won't hit them," said another Windows 2000 developer.

Low priority for M$, that is. And people won't hit them.
Their computers will.

>"It's like running a grammar checker."
That you wrote yourself and were a) too incompetent to do right
and are b) too incompetent to fix.


>Mark Minasi, a Windows NT author and speaker at the conference this week,
>said Windows 2000 is very stable, but users may run into problems running
>their applications with Active Directory and IntelliMirror.

So this guys definition of "very stable" is "users may run into problems."


>"The kernel is rock solid," Minasi said.

And runs like one.

>"You'll rarely see blue screens, but there will be more visits from Dr.
> Watson."

Well, *that*'s an improvement, right?

>Dr. Watson is a system tool in Windows 98 and Windows 2000 that gives
>users a picture of their software environment so that when a problem
>occurs, they can use this information in conjunction with tech support
>to identify the source of the error.

Which will be Win2k^H^H63K.

>Minasi said the biggest groans about Windows 2000 out of the gate will
>stem from application incompatibilities and the lack of drivers following
>the initial release of Windows 2000, not bugs.

So "application incompatibilities" are "not bugs." And just because
there are bigger problems doesn't mean that M$'s *own report* of 63K
bugs is not a problem.


>"The biggest problem is the lack of drivers," Minasi said.
>"And as always, it won't be 100 percent compatible with all legacy
>applications."

Like MS Word, MS Excel, MS Powerpoint, MS Exchange, and MS Lookout
Express. After all, M$ wants to sell you upgrades.

>At least one source noted problems with Adaptec's 2940UW SCSI card
>http://www.techweb.com/encyclopedia/defineterm.cgi?sstring=SCSI"
>Proxim Symphony and RangeLan wireless LAN cards, and the Diamond
>Multimedia Systems Monster Sound MX300 card.

After all, M$ can't be expected to know how to keep an interface
downwardly compatible, now can they? They can't even write
a decent bug-spotting tool.

>And at least one Windows 2000 VAR said he has run into a driver problem
>related to the new Windows file protection feature.

>"We received an updated driver to fix a buggy one for a major vendor's
>multiport serial card, and yet when I installed the updated driver, per
>their instructions, [Windows 2000] forcibly replaced the fixed driver with
>the buggy one, without asking," said the VAR, who managed to figure out a
>workaround after several days. "The problem is that none of us really
>got to test this feature during the beta because so few third-party
>drivers were available. Many of us yelled and screamed at vendors to
>release beta drivers, but they routinely gave the response: 'Win2K is a
>beta OS and we don't support beta OSes.'

(A blanket statement that MS OSes will never be supported.)

Meaning that the vendors have tried supporting MS official beta releases,
then found that the final release was completely different from the betas.
So that for the first few months after the release, only MS's (newer,
more expensive) software would work with it.


>"It has the potential of being very ugly out there for some time to come.
>There's no easy workaround. There's no way for an administrator to disable
>it or to override the feature."

Courtesy of Microsoft. How thoughty of them.


>He said vendors are required to get digital signatures from Microsoft.

For lots & lots of money.

>"This is going to be an administrative nightmare in about three weeks," the
>VAR said. "Every major and minor vendor is going to have new drivers out
>there. Some will update them weekly until they get them right. Does
>Microsoft think vendors have time to get every driver signed?"

Microsoft thinks vendors have money to get every driver signed.

>Microsoft, Redmond, Wash., declined to comment on that specific driver
>issue.

And the reporter who asked the question will never be accredited by MS
again.

>At the show, Microsoft officials said many drivers already have digital
>signatures and they do not expect many problems.

A pioneer of the Somebody Else's Problem school of support.


>They also maintained that most applications will run seamlessly on the new
>platform.

(I.e., crash as a unit, instead of dying one piece at a time.)


--
Joe Bednorz
"The Accursed will be advised of the absence of his rights under
the uniform code..." FST-DCTDHMTP

Christian Bauernfeind

unread,
Feb 17, 2000, 3:00:00 AM2/17/00
to
In article <38ac5f50...@news.mindspring.com>,

jbed...@mindspring.com (Joe Bednorz) writes:
>
>>He said Microsoft's internal Prefix tool, which identified the
>>63,000 items, often flags inconsequential issues such as
>>misspellings and sets off false alarms.
>
> Misspellings in code are inconsequential? It's okay because
> their applications are too buggy to be trusted?
>

That takes the cake right there. I mean, they spot a spelling error,
and instead of starting a flamewar^W^W^Wfixing the error, they enter
a scandisking problem report?


Christian
--
Christian Bauernfeind <v2ba...@fishkill.ibm.com>
Not speaking for Infineon, not even working for IBM

RFC 882 put the dot in .com.

Robert Crawford

unread,
Feb 17, 2000, 3:00:00 AM2/17/00
to
Christian Bauernfeind <v2ba...@fishkill.ibm.com> wrote:

> jbed...@mindspring.com (Joe Bednorz) writes:
>>>He said Microsoft's internal Prefix tool, which identified the
>>>63,000 items, often flags inconsequential issues such as
>>>misspellings and sets off false alarms.
>> Misspellings in code are inconsequential? It's okay because
>> their applications are too buggy to be trusted?
>That takes the cake right there. I mean, they spot a spelling error,
>and instead of starting a flamewar^W^W^Wfixing the error, they enter
>a scandisking problem report?

Calling a bug a "misspelling" is Microsoft's attempt to make
morons think it's OK. After all, the morons misspell words all the time,
so it's not a big deal, right?

Unless, of course, that misspelling results in your registry
being corrupted.

--
craw...@iac.net

"The bullets are just his way of saying 'Keep it down, I've got a
hangover.'"


Joe Bednorz

unread,
Feb 19, 2000, 3:00:00 AM2/19/00
to
On Sat, 19 Feb 2000 00:07:30 GMT, pmca...@iol.ie (Paul Mc Auley) wrote:

>Joe Bednorz <jbed...@mindspring.com> wrote on Thu, 17 Feb 2000 21:04:03 GMT:
>| (Additions & corrections welcome, I don't feel I really did the
>| subject justice.)
>| Original at: <http://www.techweb.com/wire/story/TWB20000217S0008>
>
>| >Microsoft's development team is fuming about a published report
>| >stating that Windows 2000 contains more than 63,000 bugs.
>

>Right, hands up who saw that number and thought:
> "Hmm, 16 bit int!"
>

Wonder how many times it rolled over?

(That completely slipped by me.)

--
Joe Bednorz "kill -9 them all. init will know its own."

Richard Gadsden

unread,
Feb 20, 2000, 3:00:00 AM2/20/00
to
In article <88hqni$c7q$1...@news.btv.ibm.com>, v2ba...@fishkill.ibm.com
(Christian Bauernfeind) wrote:

> In article <38ac5f50...@news.mindspring.com>,


> jbed...@mindspring.com (Joe Bednorz) writes:
> >
> >>He said Microsoft's internal Prefix tool, which identified the
> >>63,000 items, often flags inconsequential issues such as
> >>misspellings and sets off false alarms.
> >
> > Misspellings in code are inconsequential? It's okay because
> > their applications are too buggy to be trusted?
> >
>
> That takes the cake right there. I mean, they spot a spelling error,
> and instead of starting a flamewar^W^W^Wfixing the error, they enter
> a scandisking problem report?

And. Remember that the vast majority of M$'s developers don't even see
the source code to the OS, and majority of those who do can't change it.

Look at the build number. That's the number of times that NT has been
compiled. Ever. Suckage from hell.

--
Richard Gadsden
"[T]he secret to high uptimes is no one to use the network, no
one to manage the network and no one to maintain the network"
Chris Hacking, the Scary Devil Monastery

Christian Bauernfeind

unread,
Feb 21, 2000, 3:00:00 AM2/21/00
to
In article <memo.2000022...@richard-tga.u-net.com>,

ric...@tga.u-net.com (Richard Gadsden) writes:
>
> And. Remember that the vast majority of M$'s developers don't even see
> the source code to the OS, and majority of those who do can't change it.
>

This can't possibly be UI. So please elaborate.

> Look at the build number. That's the number of times that NT has been
> compiled. Ever. Suckage from hell.
>

Flippant answer: Since their development mantra is "It compiles,
ship it," it must be full build #5 hitting the shelves last week.

Semi-serious answer: At a rate of one per day, 2195 still accounts for
6 years of development. This says nothing about the number of compiles
on individual modules.

Richard Gadsden

unread,
Feb 21, 2000, 3:00:00 AM2/21/00
to
In article <88rj84$cme$1...@news.btv.ibm.com>, v2ba...@fishkill.ibm.com
(Christian Bauernfeind) wrote:

> In article <memo.2000022...@richard-tga.u-net.com>,
> ric...@tga.u-net.com (Richard Gadsden) writes:
> >
> > And. Remember that the vast majority of M$'s developers don't even
> > see the source code to the OS, and majority of those who do can't
> > change it.
> >
>
> This can't possibly be UI. So please elaborate.

[According to an acquaintance on the IE development team]

Unless you're on the NT development team, you don't get close to the
source for NT. The upshot is that the apps developers find bugs in the
OS, kick bug reports to the OS developers and the OS developers then
have to find the cause of the bug. So IE developers code a workaround
for an OS bug into IE because the OS developers are too busy to fix the
bug. Large chunks of the bloat is one set of M$ developers writing
workarounds for another set's bugs.

Equally, if you're an NT developer, you can't change anything other than
the bit you're working on. You probably can't see the source to the
rest, and even if you can, you can't change it and test the effects of
the change, because you're not allowed to compile anyone else's bits.
All you can do is say "I think that this diff will work, but I can't be
sure because I've never compiled it".

One proposal that's been knocking around is "internal open source" to
allow M$ developers to hack M$ source and recompile to test.

Mark Atwood

unread,
Feb 21, 2000, 3:00:00 AM2/21/00
to
ric...@tga.u-net.com (Richard Gadsden) writes:
<snipped>

Cheese and rice on a toasted bun!

Who set this up? Some ex-spook?

--
Mark Atwood | It is the hardest thing for intellectuals to understand, that
m...@pobox.com | just because they haven't thought of something, somebody else
| might. <http://www.friesian.com/rifkin.htm>
http://www.pobox.com/~mra

Robert Crawford

unread,
Feb 21, 2000, 3:00:00 AM2/21/00
to
On 22 Feb 2000 02:55:30 GMT, Peter da Silva <pe...@abbnm.com> wrote:

>Mark Atwood <m...@pobox.com> wrote:
>> Who set this up? Some ex-spook?
>Bill Gates is apparently a bit of a control freak.

<sarcasm>
Hmmmm... Ya know, I would never have guessed that.
</sarcasm>

Alexander Viro

unread,
Feb 21, 2000, 3:00:00 AM2/21/00
to
In article <88str2$g...@web.nmti.com>, Peter da Silva <pe...@abbnm.com> wrote:
>In article <m3n1oua...@flash.localdomain>,

>Mark Atwood <m...@pobox.com> wrote:
>> Who set this up? Some ex-spook?
>
>Bill Gates is apparently a bit of a control freak.

... and they probably got fed up with:
1.427: $SOME_DATE by $LUSER1
foo() was checking for bar being below MAX_BAR. I need more, so I've
removed the check.
1.428: $SOME_DATE by $LUSER2
It didn't compile. Fixed it removing the junk and reformatting the
rest^Wend of foo() - from some point it was indented one level too deep.
I'm not too good in C, but this one was quite obvious.
[...]
1.929: $YEAR_LATER by $KERNEL_GUY
It was unfsckingbelievably broken. Turns out that the first imbecil
(see 1.427) removed the opening brace when he was busily breaking foo()
(fuckwit, if we check for something we may have a reason for that. Like checking
for potential stack smashing five lines below). Guess what had the next idiot
done? Just look at 1.428. And then a dozen of guys unfamiliar with the code
in foobar.c had been applying bandaids for a FUCKING YEAR, DAMNIT. Removed
the crap from foobar.c, restored foo() into something that looks sane (if I
understood the changes they tried to apply in 1.694 and 1.833 - somebody,
check this). IWBNI somebody could find a time to remove the bandaids outside
of foobar.c... WHY THE FUCK DO WE HAVE TO SPEND TIME CLEANING UP AFTER THE
CLUELESS WANKERS WHO SHOULD NOT BE ALLOWED TO TOUCH THE CODE? Not too good
in C, indeed... Aaarrrrgh...
[...]

--
All that blue light from Orthanc at night? That was Saruman, trying to
moderate news.admin.palantir-abuse.sightings.
Mike Andrews in the Monastery

Peter da Silva

unread,
Feb 22, 2000, 3:00:00 AM2/22/00
to
In article <m3n1oua...@flash.localdomain>,
Mark Atwood <m...@pobox.com> wrote:
> Who set this up? Some ex-spook?

Bill Gates is apparently a bit of a control freak.

--
In hoc signo hack, Peter da Silva <pe...@baileynm.com>
`-_-' Ar rug tú barróg ar do mhactíre inniu?
'U`
"I *am* $PHB" -- Skud.

Jasper Janssen

unread,
Feb 22, 2000, 3:00:00 AM2/22/00
to
al...@lspace.org (Alan Bellingham) wrote:

>Conceivably, the misspelling is in a comment. But WTF would the tool be
>doing looking at the comments?

CreatorCode: MicroSoft.

You have to *ask*?

Given that MS files localisation mispelings and mistranslations as
bugs, I wouldn't be surprised..

Jasper

Christian Bauernfeind

unread,
Feb 22, 2000, 3:00:00 AM2/22/00
to
In article <memo.20000221...@richard-tga.u-net.com>,

ric...@tga.u-net.com (Richard Gadsden) writes:
> In article <88rj84$cme$1...@news.btv.ibm.com>, v2ba...@fishkill.ibm.com
> (Christian Bauernfeind) wrote:
>
>> In article <memo.2000022...@richard-tga.u-net.com>,
>> ric...@tga.u-net.com (Richard Gadsden) writes:
>> >
>> > And. Remember that the vast majority of M$'s developers don't even
>> > see the source code to the OS, and majority of those who do can't
>> > change it.
>> >
>>
>> This can't possibly be UI. So please elaborate.
>
> [According to an acquaintance on the IE development team]
>
> Unless you're on the NT development team, you don't get close to the
> source for NT. The upshot is that the apps developers find bugs in the

That makes more sense now. Somehow I had parsed the above with
developers=NT group and envisioned a truly bizarre development process
in which the developers had no access to their own code
that would somehow explain everything.

Alas, no revelation. Just the usual suckiness all over the place.

Nick Manka

unread,
Feb 22, 2000, 3:00:00 AM2/22/00
to
In article <88t1na$3...@weyl.math.psu.edu>,
vi...@weyl.math.psu.edu (Alexander Viro) writes:

> ... and they probably got fed up with:
> 1.427: $SOME_DATE by $LUSER1

> 1.428: $SOME_DATE by $LUSER2
> 1.929: $YEAR_LATER by $KERNEL_GUY

This assumes one of the division in which they gave up using
SourceSafe for version control.

--
Network Samurai http://www.syncronym.org/~nick/

Devin L. Ganger

unread,
Feb 23, 2000, 3:00:00 AM2/23/00
to
<flashything!> Remember only that on 22 Feb 2000 17:42:26 GMT, light
refracted through a pocket of swamp gas in alt.sysadmin.recovery

made Nick Manka <ni...@faust.eng.baileynm.com> write:

> This assumes one of the division in which they gave up using
> SourceSafe for version control.

Microsoft doesn't use VSS internally. It's not nearly flexible enough,
let alone scalable enough, for their needs.

--
Devin L. Ganger, Chief Systems Administrator, Premier1 Internet Services
"I'm worried Usenet is making me the woman I'll be tomorrow."
Frossie in the Scary Devil Monastery
** Please don't email & post responses to my Usenet articles unless asked **

Christian Bauernfeind

unread,
Feb 23, 2000, 3:00:00 AM2/23/00
to
In article <slrn8b69tp...@aridhol.premier1.net>,

de...@premier1.net (Devin L. Ganger) writes:
>
> Microsoft doesn't use VSS internally. It's not nearly flexible enough,
> let alone scalable enough, for their needs.
>

Do they also use 6 different compiles on W2k because one simply doesn't
scale up?

Mike Andrews

unread,
Feb 23, 2000, 3:00:00 AM2/23/00
to
Mark Atwood <m...@pobox.com> wrote:

: ric...@tga.u-net.com (Richard Gadsden) writes:
: <snipped>

: Cheese and rice on a toasted bun!

: Who set this up? Some ex-spook?

The very thought I had. This makes military compartmentalization look
almost reasonable by comparison. Do they issue blinders to all personnel
so that they won't see things in their walks (following the blue line,
of course) down the hall to and from the john? Or do they just hood
the developers when they're out of their work areas, and golf-cart them
through the building?

I mean, these developers (almost started to call them people, but
caught myself in time) have almost certainly signed NDAs giving Big
Bad Bill their bones, blood, bank account, and firstborn if they blab.
What's all this internal compartmenting _about_?

--
"HTML's a cheap whore. Treating her with respect is possible, and even pref-
erable, because once upon a time she was a beautiful and virginal format, but
you shouldn't expect too much of her at this point." (Mark 'Kamikaze' Hughes)


Robert Crawford

unread,
Feb 23, 2000, 3:00:00 AM2/23/00
to
On Wed, 23 Feb 2000 17:54:09 GMT, Mike Andrews <mi...@mikea.ath.cx> wrote:
[MSFT's draconian _internal_ source restrictions]

>I mean, these developers (almost started to call them people, but
>caught myself in time) have almost certainly signed NDAs giving Big
>Bad Bill their bones, blood, bank account, and firstborn if they blab.
>What's all this internal compartmenting _about_?

Control. The fear is that, if people start fixing bugs
in other people's code, they'll never get 'round to adding
bugs^Wfeatures to their own. MSFT is still wrapped up in the
"feature shoot out" mentality that McConnel(?) described in one
of his books.

Christian Bauernfeind

unread,
Feb 24, 2000, 3:00:00 AM2/24/00
to
In article <bp25bskp54lrbjshi...@4ax.com>,

Jasper Janssen <jas...@janssen.dynip.com> writes:
>
> Given that MS files localisation mispelings and mistranslations as
> bugs, I wouldn't be surprised..
>

Don't call that odorous heap of organic fertilizer localisation.
The Amiga had localization, what they have is a scandisking separate
version for _British English_, and no hope in $HADES of keeping your
system "current"[1] unless you run the glorified US of A version.


Christian

[1] For values of current that include patches that are only available
through secret incantations in the "knowledge" base and that
that definitely not include stable.

Paul Tomblin

unread,
Feb 24, 2000, 3:00:00 AM2/24/00
to
In a previous article, mi...@mikea.ath.cx (Mike Andrews) said:
>of course) down the hall to and from the john? Or do they just hood
>the developers when they're out of their work areas, and golf-cart them
>through the building?

Now you know why Microsofties all get real offices with doors instead of
cubes. Bastards.


--
Paul Tomblin, not speaking for anybody.

'Usenet "belongs" to those who administer the hosts of which it is comprised'
- RFC 1036, draft revision

0 new messages