What's the deal with MS05-002 (KB891711.EXE) and Windows 98?

Anita Bryant?

PCR wrote:
COLORADO, RUN! There is an MSFT on the prowl, going by the name of
Bryant!


--
Thanks or Good Luck,
There may be humor in this post, and,
Naturally, you will not sue,
should things get worse after this,
PCR

"Bill in Co." wrote in message
...
cquirke (MVP Windows shell/user) wrote:
On Fri, 25 Mar 2005 15:34:13 -0500, "Rick Chauvin"
cquirke (MVP Windows shell/user) wrote:

More on that, please? Which two .DLL files?

cquirke, if you can find a way to solve this problem without having to
swap
to the old 5.5 dlls and what that means, you will be W9x's official
grand
hero! vbg

I would sooo like to fix this problem.

It is "fixed" - just swap the damn files! Well, at least it's fixed
over
here! And I'm in heaven for it now!! There is NO WAY I'll put
those
files back, unless something major comes up - and it hasn't!

You have two choices now:
1) live with the problem, and suffer needlessly, or
2) join me in full contentment. :-)

(your call)

It drives me mad, on my own
system - it's pathetic when you have to spwn a C: (command prompt)
just to delete or rename files and create new dirs, because Windows is
too incompetent to do it properly. I haven't seen that degree of core
file management uselessness since Win3.yuk's UnDelete screw-up.

The problem is real and is a royal p.i.t.a ..I get around it by
holding
the Shift key when I want to delete anything out of the ordinary - and
that seems to keep me out of the problem

I routinely do that (it has the effect of bypassing Recycle Bin
directly to trash) and it doesn't fix the problem for me. What fixes
the problem (for a while) is Ctl+Alt+Del and killing Explorer, so that
the shell restarts. I've seen resource heaps go down with this
problem, but I've also seen the problem when the heaps were OK.

At one time I thought it had to do with damaged \Recyced folders, i.e.
those that lack the "special" Desktop.ini and thus look like regular
yellow folders in the drive's root dir. I played with that, and for a
while it looked like I could initiate the problem that way, but the
results were inconclusive and I abandoned testing.

I've seen the work you did with the TS-BATS setup you made for parallel
installs and what that represented in know-how for you to do that, and
other things, and so I'm complementing you forward to see if you can
resolves this issue for us too.

I'm flattered, but I doubt if I'll lick this - I'll certainly give it
a go, though. If time permitted, I might get medieval with FC and
pasting between .DLLs to find the rotten function call :-)



---------- ----- ---- --- -- - - - -
Gone to bloggery: http://cquirke.blogspot.com
---------- ----- ---- --- -- - - - -

Bill in Co. wrote:
Rick Chauvin wrote:
Bill in Co. wrote:

[...snips for brevity..]

Not a big guy, but a stubborn ornery one, one not easily swayed by "the
crowd".

Me neither, but a genuine right balance is the key.

Remember Emerson? "whoso would be a man, must be a non-conformist", and
that comes naturally for me. :-)

smile

Yeah, but do you really think it's gonna happen in our lifetime? Come on
now, let's not be naive.

Yes, and I'm far from being naive or a dreamer g

(Of course I never thought I'd live to see the Berlin Wall come down
either)

exactly

Good enough, Rick! :-)

Take good care

Rick

Bill in Co. wrote:
Rick Chauvin wrote:
Bill in Co. wrote:

[...snips for brevity..]

Not a big guy, but a stubborn ornery one, one not easily swayed by "the
crowd".

Me neither, but a genuine right balance is the key.

Remember Emerson? "whoso would be a man, must be a non-conformist", and
that comes naturally for me. :-)

smile

Yeah, but do you really think it's gonna happen in our lifetime? Come on
now, let's not be naive.

Yes, and I'm far from being naive or a dreamer g

(Of course I never thought I'd live to see the Berlin Wall come down
either)

exactly

Good enough, Rick! :-)

Take good care

Rick

cquirke (MVP Windows shell/user) wrote:

On Fri, 25 Mar 2005 15:34:13 -0500, "Rick Chauvin"
cquirke (MVP Windows shell/user) wrote:

cquirke, if you can find a way to solve this problem without having to swap
to the old 5.5 dlls and what that means, you will be W9x's official grand
hero! vbg

I would sooo like to fix this problem. It drives me mad, on my own
system - it's pathetic when you have to spwn a C: (command prompt)
just to delete or rename files and create new dirs, because Windows is
too incompitent to do it properly. I haven't seen that degree of core
file management uselessness since Win3.yuk's UnDelete screw-up.

The problem is real and is a royal p.i.t.a ..I get around it by holding
the Shift key when I want to delete anything out of the ordinary - and
that seems to keep me out of the problem

I routinely do that (it has the effect of bypassing Recycle Bin
directly to trash) and it doesn't fix the problem for me. What fixes
the problem (for a while) is Ctl+Alt+Del and killing Explorer, so that
the shell restarts. I've seen resource heaps go down with this
problem, but I've also seen the problem when the heaps were OK.

At one time I thought it had to do with damaged \Recyced folders, i.e.
those that lack the "special" Desktop.ini and thus look like regular
yellow folders in the drive's root dir. I played with that, and for a
while it looked like I could initiate the problem that way, but the
results were inconclusive and I abandoned testing.

I've seen the work you did with the TS-BATS setup you made for parallel
installs and what that represented in know-how for you to do that, and
other things, and so I'm complementing you forward to see if you can
resolves this issue for us too.

I'm flattered, but I doubt if I'll lick this - I'll certainly give it
a go, though. If time permitted, I might get medieval with FC and
pasting between .DLLs to find the rotten function call :-)

Okay, thank you Chris.
Also with your connections at MS maybe you can forward the problem to the
appropriate people who have the power to fix it, or least help you fix it.
If the original coder sees the problem, I'm sure he will right away know
exactly what to do, click click type type here and there adding a few lines
of code in the right spot to neutralize the problem; and we'll be free at
last, free at last.

Thank you,
Rick
( I'll re-read your post better Monday Chris for some above details I wanted
to read again, but I'm off to Conn now for the weekend to visit my mom in the
nursing home, bless her! )




---------- ----- ---- --- -- - - - -
Gone to bloggery: http://cquirke.blogspot.com
---------- ----- ---- --- -- - - - -

On Sat, 26 Mar 2005 10:37:27 -0700, "Bill in Co."

Speaking of "there's still no way of knowing"..., the same logic applies to
taking in some of those updates!

Aye, and that's the problem - with fixes like this, as well as lack of
acknoledgement and fixing of the "fix", patches become just another
risky type of software that are best avoided in the interests of
stability. Patches fix real issues, and you may have only days before
these issues blow up (Lovesan vs. RPC, 30 days; Sasser vs. LSASS, 14
days; some recent events have negative lead times).

The problem is, you can't tell in advance which patches will turn out
to be as uber-crucial as RPC and LSASS (both affecting NT, not Win9x,
it's worth noting) and which will be a snore (remember ASN.1?).

With so many defects and patches - if MS didn't limit these to
once-a-month release, it's almost as real-time as av updates - it
becomes difficult to keep your eye on the details. It beciomes:

"My PC keeps getting malware'd, and I don't click anything!"
' Are you up to date with your patches? '

So we're between a rock (need to patch) and a hard place (dubious
quality control and follow-up of patches). Dumping the whole platform
and moving to (say) Linux isn't a solution, because any
similarly-complex platform that attracts significant market share
*will* have the same problem with one-in-a-thousand defects coming up
regularly, given there are so many million lines of code.

For example, notice how Firefox subversions have been a revolving door
lately, remeniscent of AdAware SE 1.00 through 1.05? The plan was for
Firefox 1.1 to come out this month, but we've had 1.0.01 and now 1.0.2
instead. Firefox is getting popular (I love it!) and that means it's
beginning to attract malware attention. Patches inevitably follow.

Gary S. Terhune wrote:
And there's *still* no knowing what got broken by employing this "fix".

My specific worry is the MIME-spoofing defects that are prevelant in
SP 5.5 SP0 and SP1. That's why I want to know whether using the SP2
versions of those .DLL will fix the problem (my hunch is, they may
not). That defect is IMO too dangerous to leave in the water, given
how pervasive HTML and the IE HTML renderer are, and how malware have
routinely exploited that hole since BadTrans.B in 2001.

The fix says; keep the fixed .DLL local in IE's dir, and have the old
ones visible system-wide (to keep Windows Explorer happy) in
%WinDir%\System. Which .DLL gets used if IE's renderer gets called
upon to read HTML content in an arbitrary location - say, a .TXT that
is internally HTML, or a .CHM in some free app's base dir, or HTML
passed through %Temp% by an email app, or a toxic "ReadMe" that you
clicked on to view from the middle of a .Zip in WinZip?

There may be more than one devil in the details.



---------- ----- ---- --- -- - - - -
Gone to bloggery: http://cquirke.blogspot.com
---------- ----- ---- --- -- - - - -

On Sat, 26 Mar 2005 10:44:05 -0700, "Bill in Co."
Rick Chauvin wrote:
Bill in Co. wrote:

Today's fun link:

http://c2.com/cgi/wiki?WhiteBicycleTechnology

That's how MS approached "HTML everywhere!", and we still feel the
echoes of that pain today.

Remember Emerson? "whoso would be a man, must be a non-conformist", and
that comes naturally for me. :-)

Oh great, thanks for reminding me I'm a nut-swinger. Mind you, I'm
happy if this puts me in the same company as so many virtual
nut-swingers, from Susan B Anthony, the Pankhursts etc. to the more
recent Atwoods, Piercys and Moorheads of the world.

I agree that the way IE6 is stock with this problem is unacceptable.

TOTALLY UNACCEPTABLE!!!.

Agreed - and thesre's no face-saving PR solution for MS on this one.
"We decided to let Win9x bleed because we're promoting a move to XP"
may well be the truth, but that's just another scandal - the leverage
of product defects to compel sales. I don't think MS would like legal
and regulatory eyes to dwell too long on that.

I vote to get it fixed properly though so that those two dll's are
inclusive of all Numerous Updates that are Missing if swapped

If I had resources to fix this (as one presumes MS has) I'd do this:
- get documentation on what's changed in those two .DLL
- any crucial exploits that depend on those files to fix?
- determine most recent .DLL that work via the fix (IE 5.5 SP2?)
- FC these against the IE 6 ones that don't work
- see if what FC finds can be mapped to particular functions
- if offset dependencies allow:
- paste across functions from IE 6 to old until old breaks
- paste across functions from old to IE 6 until IE 6 works
- zoom in and disassemble the problemetic function
- see if a logic error etc. can be found
- fix the function and issue fixed .DLLs as on-request hotfix
- after testing, issue as downloadable hotfix, then WinUpate

MS wouldn't have to hack around with FC, given they'd have the source
code. It would also be easier from them to read up on what changes
were made to these .DLL, and whether any exploits need these changes
to remain blocked. It would be a sad commentary on MS, if within a
few months of reading this post, somene ITW came up with a fix, even
though they lacked MS's advantage of source code and documentation.

Well, there's the gauntlet. Anyone care to pick it up? :-)



---------- ----- ---- --- -- - - - -
Gone to bloggery: http://cquirke.blogspot.com
---------- ----- ---- --- -- - - - -

cquirke (MVP Windows shell/user) wrote:
On Sat, 26 Mar 2005 10:44:05 -0700, "Bill in Co."
Rick Chauvin wrote:
Bill in Co. wrote:

Today's fun link:

http://c2.com/cgi/wiki?WhiteBicycleTechnology

That's how MS approached "HTML everywhere!", and we still feel the
echoes of that pain today.

Remember Emerson? "whoso would be a man, must be a non-conformist", and
that comes naturally for me. :-)

Oh great, thanks for reminding me I'm a nut-swinger. Mind you, I'm
happy if this puts me in the same company as so many virtual
nut-swingers, from Susan B Anthony, the Pankhursts etc. to the more
recent Atwoods, Piercys and Moorheads of the world.

I agree that the way IE6 is stock with this problem is unacceptable.

TOTALLY UNACCEPTABLE!!!.

Agreed - and thesre's no face-saving PR solution for MS on this one.
"We decided to let Win9x bleed because we're promoting a move to XP"
may well be the truth, but that's just another scandal - the leverage
of product defects to compel sales. I don't think MS would like legal
and regulatory eyes to dwell too long on that.

I vote to get it fixed properly though so that those two dll's are
inclusive of all Numerous Updates that are Missing if swapped

If I had resources to fix this (as one presumes MS has) I'd do this:
- get documentation on what's changed in those two .DLL
- any crucial exploits that depend on those files to fix?
- determine most recent .DLL that work via the fix (IE 5.5 SP2?)

IE 5.5 SP2 does work (that's what I've been using). However, whether or
not "browseui" and "browselc" changed between IE 5.5, IE 5.5 SP1, and IE 5.5
SP2, I don't know for sure, but if I were a betting man, I would guess at
least one of them did.

Well let's see, I notice the file dates on those two files are 12/03 and
7/01, respectively (in my particular case), so presumably the "browseui" one
was upgraded a tad with SP2.

- FC these against the IE 6 ones that don't work
- see if what FC finds can be mapped to particular functions
- if offset dependencies allow:
- paste across functions from IE 6 to old until old breaks
- paste across functions from old to IE 6 until IE 6 works
- zoom in and disassemble the problemetic function
- see if a logic error etc. can be found
- fix the function and issue fixed .DLLs as on-request hotfix
- after testing, issue as downloadable hotfix, then WinUpate

MS wouldn't have to hack around with FC, given they'd have the source
code. It would also be easier from them to read up on what changes
were made to these .DLL, and whether any exploits need these changes
to remain blocked. It would be a sad commentary on MS, if within a
few months of reading this post, somene ITW came up with a fix, even
though they lacked MS's advantage of source code and documentation.

Well, there's the gauntlet. Anyone care to pick it up? :-)

Oh yeah. I can see all of them running right over and picking up that
gauntlet! (LOL).

cquirke (MVP Windows shell/user) wrote:
On Sat, 26 Mar 2005 10:44:05 -0700, "Bill in Co."
Rick Chauvin wrote:
Bill in Co. wrote:

Today's fun link:

http://c2.com/cgi/wiki?WhiteBicycleTechnology

That's how MS approached "HTML everywhere!", and we still feel the
echoes of that pain today.

Remember Emerson? "whoso would be a man, must be a non-conformist", and
that comes naturally for me. :-)

Oh great, thanks for reminding me I'm a nut-swinger. Mind you, I'm
happy if this puts me in the same company as so many virtual
nut-swingers, from Susan B Anthony, the Pankhursts etc. to the more
recent Atwoods, Piercys and Moorheads of the world.

I agree that the way IE6 is stock with this problem is unacceptable.

TOTALLY UNACCEPTABLE!!!.

Agreed - and thesre's no face-saving PR solution for MS on this one.
"We decided to let Win9x bleed because we're promoting a move to XP"
may well be the truth, but that's just another scandal - the leverage
of product defects to compel sales. I don't think MS would like legal
and regulatory eyes to dwell too long on that.

I vote to get it fixed properly though so that those two dll's are
inclusive of all Numerous Updates that are Missing if swapped

If I had resources to fix this (as one presumes MS has) I'd do this:
- get documentation on what's changed in those two .DLL
- any crucial exploits that depend on those files to fix?
- determine most recent .DLL that work via the fix (IE 5.5 SP2?)

IE 5.5 SP2 does work (that's what I've been using). However, whether or
not "browseui" and "browselc" changed between IE 5.5, IE 5.5 SP1, and IE 5.5
SP2, I don't know for sure, but if I were a betting man, I would guess at
least one of them did.

Well let's see, I notice the file dates on those two files are 12/03 and
7/01, respectively (in my particular case), so presumably the "browseui" one
was upgraded a tad with SP2.

- FC these against the IE 6 ones that don't work
- see if what FC finds can be mapped to particular functions
- if offset dependencies allow:
- paste across functions from IE 6 to old until old breaks
- paste across functions from old to IE 6 until IE 6 works
- zoom in and disassemble the problemetic function
- see if a logic error etc. can be found
- fix the function and issue fixed .DLLs as on-request hotfix
- after testing, issue as downloadable hotfix, then WinUpate

MS wouldn't have to hack around with FC, given they'd have the source
code. It would also be easier from them to read up on what changes
were made to these .DLL, and whether any exploits need these changes
to remain blocked. It would be a sad commentary on MS, if within a
few months of reading this post, somene ITW came up with a fix, even
though they lacked MS's advantage of source code and documentation.

Well, there's the gauntlet. Anyone care to pick it up? :-)

Oh yeah. I can see all of them running right over and picking up that
gauntlet! (LOL).

No! I'll do my best to protect you, though, soon as I return from
Saskatchewan!

--
Thanks or Good Luck,
There may be humor in this post, and,
Naturally, you will not sue,
should things get worse after this,
PCR

"Bill in Co." wrote in message
...
| Anita Bryant?
|
| PCR wrote:
| COLORADO, RUN! There is an MSFT on the prowl, going by the name of
| Bryant!
|
|
| --
| Thanks or Good Luck,
| There may be humor in this post, and,
| Naturally, you will not sue,
| should things get worse after this,
| PCR
|
| "Bill in Co." wrote in message
| ...
....snip

No! I'll do my best to protect you, though, soon as I return from
Saskatchewan!

--
Thanks or Good Luck,
There may be humor in this post, and,
Naturally, you will not sue,
should things get worse after this,
PCR

"Bill in Co." wrote in message
...
| Anita Bryant?
|
| PCR wrote:
| COLORADO, RUN! There is an MSFT on the prowl, going by the name of
| Bryant!
|
|
| --
| Thanks or Good Luck,
| There may be humor in this post, and,
| Naturally, you will not sue,
| should things get worse after this,
| PCR
|
| "Bill in Co." wrote in message
| ...
....snip