The IE4 bug that just won't go away.......

The IE4 bug that just won't go away.......

Post by Richard M. Smit » Thu, 23 Apr 1998 04:00:00



Eric,

Maybe if Microsoft just would fix this bug, people would stop asking about it!
The bug seems like a real show-stopper to me and I am surprised that
IE4 ever went out the door with a bug this bad in it.  Even more surprising is
that the bug is still being "researched" according to the knowledge base article.

Mucking with the registry seems pretty dangerous to me.  I hope the problem has
been fixed in Win98.

Richard


Quote:> The answer is below.

> In the future, please check Dejanews and the FAQs before asking questions.
> I answered this question three times last week.

> Thanks!

> Eric

> > Here's the full info on getting window.open to work:

> > "Open In New Window" Does Not Work in Internet Explorer
> > Last reviewed: February 26, 1998
> > Article ID: Q180176
> > The information in this article applies to:
> > Microsoft Internet Explorer versions 4.0, 4.01 for Windows 95

> > SYMPTOMS
> > When you right-click a Web address on a Web page and then click Open In
> New
> > Window, the Web page may not be opened in a new window.

> > When you click a link on a Web page that uses scripting to open a new
> > window, the new window may not be opened, or you may receive the following
> > error message:

> >    Internet Explorer Script Error

> >    An error has occurred in the script on this page.

> >    Line:
> >    Char:
> >    Error:  No such interface supported
> >    Code:   0

> >    Do you want to continue running scripts on this page?

> > CAUSE
> > This issue can occur if either of the following conditions exists:

> > You install a program that does not properly register interfaces in the
> > following registry key:

> >       HKEY_CLASSES_ROOT\Interface
> > You uninstall Microsoft Visual Basic 5.0 after you use it to create a
> > program that uses the Webbrowser control provided by the Shdocvw.dll file.

> > STATUS
> > Microsoft is researching this problem and will post new information here
> in
> > the Microsoft Knowledge Base as it becomes available.

> > RESOLUTION
> > To resolve this issue, follow these steps:

> > Click Start, and then click Run.

> > In the Open box, type the following line:

> >       regsvr32 actxprxy.dll

> > Click OK, and then click OK again when you receive the following message:

> >       DllRegisterServer in actxprxy.dll succeeded.

> > Click Start, and then click Run.

> > In the Open box, type the following line:

> >       regsvr32 shdocvw.dll

> > Click OK, and then click OK again when you receive the following message:

> >       DllRegisterServer in shdocvw.dll succeeded.

> > Shut down and restart your computer.

 
 
 

The IE4 bug that just won't go away.......

Post by n.. » Fri, 24 Apr 1998 04:00:00


Quote:>Mucking with the registry seems pretty dangerous to me.  I hope the problem has
>been fixed in Win98.

Running regsvr32  is hardly "mucking with the registry".  It's not
like running regedit.

 
 
 

The IE4 bug that just won't go away.......

Post by MS Scripting D » Fri, 24 Apr 1998 04:00:00


Richard,

The bug is not in IE.  The bug is in whatever other program is messing up
IE's registry settings.  In one case, for example, you uninstall VB5 and VB5
messes up the SHDOCVW registry settings.  Now that the IE team is aware that
there is this bug in other programs that messes up the registry, I imagine
that they will find some way of working around the problem.

This is a general problem across the PC industry -- componentized software
depends heavily on other components being well-behaved.  Personally, I think
its a tragedy that one ill-behaved application can damage another
well-behaved application by a simple bug in their uninstall routine.  We're
working on it -- in the future, it will probably be a "sticker" requirement
that you use the much safer install-uninstall routines that we're building
into the operating system.

Eric


>Eric,

>Maybe if Microsoft just would fix this bug, people would stop asking about
it!
>The bug seems like a real show-stopper to me and I am surprised that
>IE4 ever went out the door with a bug this bad in it.  Even more surprising
is
>that the bug is still being "researched" according to the knowledge base
article.

>Mucking with the registry seems pretty dangerous to me.  I hope the problem
has
>been fixed in Win98.

>Richard


>> The answer is below.

>> In the future, please check Dejanews and the FAQs before asking
questions.
>> I answered this question three times last week.

>> Thanks!

>> Eric

>> > Here's the full info on getting window.open to work:

>> > "Open In New Window" Does Not Work in Internet Explorer
>> > Last reviewed: February 26, 1998
>> > Article ID: Q180176
>> > The information in this article applies to:
>> > Microsoft Internet Explorer versions 4.0, 4.01 for Windows 95

>> > SYMPTOMS
>> > When you right-click a Web address on a Web page and then click Open In
>> New
>> > Window, the Web page may not be opened in a new window.

>> > When you click a link on a Web page that uses scripting to open a new
>> > window, the new window may not be opened, or you may receive the
following
>> > error message:

>> >    Internet Explorer Script Error

>> >    An error has occurred in the script on this page.

>> >    Line:
>> >    Char:
>> >    Error:  No such interface supported
>> >    Code:   0

>> >    Do you want to continue running scripts on this page?

>> > CAUSE
>> > This issue can occur if either of the following conditions exists:

>> > You install a program that does not properly register interfaces in the
>> > following registry key:

>> >       HKEY_CLASSES_ROOT\Interface
>> > You uninstall Microsoft Visual Basic 5.0 after you use it to create a
>> > program that uses the Webbrowser control provided by the Shdocvw.dll
file.

>> > STATUS
>> > Microsoft is researching this problem and will post new information
here
>> in
>> > the Microsoft Knowledge Base as it becomes available.

>> > RESOLUTION
>> > To resolve this issue, follow these steps:

>> > Click Start, and then click Run.

>> > In the Open box, type the following line:

>> >       regsvr32 actxprxy.dll

>> > Click OK, and then click OK again when you receive the following
message:

>> >       DllRegisterServer in actxprxy.dll succeeded.

>> > Click Start, and then click Run.

>> > In the Open box, type the following line:

>> >       regsvr32 shdocvw.dll

>> > Click OK, and then click OK again when you receive the following
message:

>> >       DllRegisterServer in shdocvw.dll succeeded.

>> > Shut down and restart your computer.

 
 
 

The IE4 bug that just won't go away.......

Post by fl.. » Fri, 24 Apr 1998 04:00:00


Mr. Lippert, if the development team of one of Microsoft's most visible
applications can't sufficient access to VB5 devlopers to get them to stop
breaking their app, how is the rest of the world supposed to expect
Microsoft to be responsive to their needs?


: Richard,

: The bug is not in IE.  The bug is in whatever other program is messing up
: IE's registry settings.  In one case, for example, you uninstall VB5 and VB5
: messes up the SHDOCVW registry settings.  Now that the IE team is aware that
: there is this bug in other programs that messes up the registry, I imagine
: that they will find some way of working around the problem.

: This is a general problem across the PC industry -- componentized software
: depends heavily on other components being well-behaved.  Personally, I think
: its a tragedy that one ill-behaved application can damage another
: well-behaved application by a simple bug in their uninstall routine.  We're
: working on it -- in the future, it will probably be a "sticker" requirement
: that you use the much safer install-uninstall routines that we're building
: into the operating system.

: Eric


: >Eric,
: >
: >Maybe if Microsoft just would fix this bug, people would stop asking about
: it!
: >The bug seems like a real show-stopper to me and I am surprised that
: >IE4 ever went out the door with a bug this bad in it.  Even more surprising
: is
: >that the bug is still being "researched" according to the knowledge base
: article.
: >
: >Mucking with the registry seems pretty dangerous to me.  I hope the problem
: has
: >been fixed in Win98.
: >
: >Richard
: >

: >
: >> The answer is below.
: >>
: >> In the future, please check Dejanews and the FAQs before asking
: questions.
: >> I answered this question three times last week.
: >>
: >> Thanks!
: >>
: >> Eric
: >>
: >> > Here's the full info on getting window.open to work:
: >> >
: >> > "Open In New Window" Does Not Work in Internet Explorer
: >> > Last reviewed: February 26, 1998
: >> > Article ID: Q180176
: >> > The information in this article applies to:
: >> > Microsoft Internet Explorer versions 4.0, 4.01 for Windows 95
: >> >
: >> >
: >> > SYMPTOMS
: >> > When you right-click a Web address on a Web page and then click Open In
: >> New
: >> > Window, the Web page may not be opened in a new window.
: >> >
: >> > When you click a link on a Web page that uses scripting to open a new
: >> > window, the new window may not be opened, or you may receive the
: following
: >> > error message:
: >> >
: >> >
: >> >    Internet Explorer Script Error
: >> >
: >> >    An error has occurred in the script on this page.
: >> >
: >> >    Line:
: >> >    Char:
: >> >    Error:  No such interface supported
: >> >    Code:   0
: >> >
: >> >    Do you want to continue running scripts on this page?
: >> >
: >> > CAUSE
: >> > This issue can occur if either of the following conditions exists:
: >> >
: >> > You install a program that does not properly register interfaces in the
: >> > following registry key:
: >> >
: >> >
: >> >       HKEY_CLASSES_ROOT\Interface
: >> > You uninstall Microsoft Visual Basic 5.0 after you use it to create a
: >> > program that uses the Webbrowser control provided by the Shdocvw.dll
: file.
: >> >
: >> > STATUS
: >> > Microsoft is researching this problem and will post new information
: here
: >> in
: >> > the Microsoft Knowledge Base as it becomes available.
: >> >
: >> >
: >> >
: >> > RESOLUTION
: >> > To resolve this issue, follow these steps:
: >> >
: >> >
: >> > Click Start, and then click Run.
: >> >
: >> > In the Open box, type the following line:
: >> >
: >> >
: >> >       regsvr32 actxprxy.dll
: >> >
: >> > Click OK, and then click OK again when you receive the following
: message:
: >> >
: >> >
: >> >       DllRegisterServer in actxprxy.dll succeeded.
: >> >
: >> > Click Start, and then click Run.
: >> >
: >> > In the Open box, type the following line:
: >> >
: >> >
: >> >       regsvr32 shdocvw.dll
: >> >
: >> > Click OK, and then click OK again when you receive the following
: message:
: >> >
: >> >
: >> >       DllRegisterServer in shdocvw.dll succeeded.
: >> >
: >> > Shut down and restart your computer.
: >> >
: >
: >
: >

--

 Ben

<Just Another System Administrator>

 
 
 

The IE4 bug that just won't go away.......

Post by Christopher Thomps » Fri, 24 Apr 1998 04:00:00



: This is a general problem across the PC industry -- componentized software
: depends heavily on other components being well-behaved.  Personally, I think
: its a tragedy that one ill-behaved application can damage another
: well-behaved application by a simple bug in their uninstall routine.  We're
: working on it -- in the future, it will probably be a "sticker" requirement
: that you use the much safer install-uninstall routines that we're building
: into the operating system.

<grin>  Store application settings in ini files in that application's
directory.  Then, to uninstall the app, you simply delete the entire
directory and remove the shortcut.

Hey wait... didn't Win 3.1 do something like that?

All things considered, I'm glad I develop applications instead of
operating systems.

--

-=Christopher Thompson=-

comp.lang.javascript meta-FAQ at:
http://ugweb.cs.ualberta.ca/~thompson/programming/javascript/meta-FAQ...
All email is spam-filtered:  http://www.internz.com/SpamBeGone/
My minimalistic web site:  http://ugweb.cs.ualberta.ca/~thompson/

 
 
 

The IE4 bug that just won't go away.......

Post by Craig Kell » Fri, 24 Apr 1998 04:00:00




->This is a general problem across the PC industry -- componentized software
->depends heavily on other components being well-behaved.  Personally, I think
->its a tragedy that one ill-behaved application can damage another
->well-behaved application by a simple bug in their uninstall routine.  We're
->working on it -- in the future, it will probably be a "sticker" requirement
->that you use the much safer install-uninstall routines that we're building
->into the operating system.

... or just use an intelligent operating system which doesn't allow this
to happen.

--
Wheel is turning, but the hamster is dead.


 
 
 

The IE4 bug that just won't go away.......

Post by Roedy Gree » Fri, 24 Apr 1998 04:00:00


Eric Lippert MS Scripting Dev asked, wrote, or quoted:

Quote:>This is a general problem across the PC industry -- componentized software
>depends heavily on other components being well-behaved.  Personally, I
>think
>its a tragedy that one ill-behaved application can damage another
>well-behaved application by a simple bug in their uninstall routine

Every few weeks I write a rant on the Microsoft registry on BIX, telling
yet another tale to prove the registry is the most fiendish invention since
musak.

It is putting all your eggs in one basket.  If that rergistry file goes
EVERYTHING goes and you have to reinstall a zillion apps from scratch.  You
can't backup restore the parts of it appying to a single app.  Apps aren't
even smart enough to do a mini-reinstall that just repairs the registry and
checks for missing shared DLLs.  DLLs are the second greatest blunder in
computing history -- or rather  tossing them into a common directory shared
by all apps and muddling DLLs from private directories if they
"accidentally" have the same 8+3 name.

In contrast, in upgrading Linux last night, I had to reformat the system
partitions to recover. NOT ONE APPLICATION APP HAD TO BE REINSTALLED.  They
all continued to work perfectly.  The key is, each app keeps all the files
it needs in its own private directory/folder and no other app meddles with
them. PERIOD.

Granted, under linux there is a shared directory called /etc (similar to
the Win 3.1 C:\WINDOWS\*.INI fdiles) where apps often collect text-format
configuration files.  However, each app has separate files so it is easy to
backup/restore them individually.

If we got rid of that flipping registry, you could install a app under NT
and access it under Win95 with no problem, and easily restore a single
corrupted application.  If we got rid of sloppily shared DLLs we would get
rid of 50+% of the problems with Windows apps failing.

It registry would be fine IF each app had its own private piece of it
stored with the rest of its files.  Apps should be in air tight boxes where
they can't interfere with each other.  Communication should be via very
rigid protocols to prevent malicious or incompetent  meddling with other
apps.

For the JAVA GLOSSARY and the CMP Utilities: <http://oberon.ark.com/~roedy>
--
Roedy Green                          Canadian Mind Products

Sponsored by: www.athena.com, makers of Integer, a multiuser
spreadsheet JavaBean. Opinions expressed are  not necessarily
those of Athena Design.
-30-

 
 
 

The IE4 bug that just won't go away.......

Post by Bill Hous » Fri, 24 Apr 1998 04:00:00



>Eric Lippert MS Scripting Dev asked, wrote, or quoted:
>>This is a general problem across the PC industry -- componentized software
>>depends heavily on other components being well-behaved.  Personally, I
>>think
>>its a tragedy that one ill-behaved application can damage another
>>well-behaved application by a simple bug in their uninstall routine

>Every few weeks I write a rant on the Microsoft registry on BIX, telling
>yet another tale to prove the registry is the most fiendish invention since
>musak.

Musak!?  That's merely insipid and annoying.  The Windows registry has
probably drained billions of dollars out of the worldwide economy since its
introduction.  A MS tech support engineer once confided to me that fully 30%
of all MS tech support calls were related to incorrect/corrupted registry
entries. He went on to confess that they routinely FDISKed their machines
each month to get rid of accumulating reg errors.  I think the Windows
registry is the stupidest idea that anyone was foolish enough to implement
in software.

Bill House

Quote:

>For the JAVA GLOSSARY and the CMP Utilities: <http://oberon.ark.com/~roedy>
>--
>Roedy Green                          Canadian Mind Products

>Sponsored by: www.athena.com, makers of Integer, a multiuser
>spreadsheet JavaBean. Opinions expressed are  not necessarily
>those of Athena Design.
>-30-

 
 
 

The IE4 bug that just won't go away.......

Post by Rob Eamo » Fri, 24 Apr 1998 04:00:00



>Eric Lippert MS Scripting Dev asked, wrote, or quoted:
>>This is a general problem across the PC industry -- componentized software
>>depends heavily on other components being well-behaved.  Personally, I
>>think
>>its a tragedy that one ill-behaved application can damage another
>>well-behaved application by a simple bug in their uninstall routine

>Every few weeks I write a rant on the Microsoft registry on BIX, telling
>yet another tale to prove the registry is the most fiendish invention since
>musak.

>It is putting all your eggs in one basket.  If that rergistry file goes
>EVERYTHING goes and you have to reinstall a zillion apps from scratch.  You
>can't backup restore the parts of it appying to a single app.  Apps aren't
>even smart enough to do a mini-reinstall that just repairs the registry and
>checks for missing shared DLLs.  DLLs are the second greatest blunder in
>computing history -- or rather  tossing them into a common directory shared
>by all apps and muddling DLLs from private directories if they
>"accidentally" have the same 8+3 name.

>In contrast, in upgrading Linux last night, I had to reformat the system
>partitions to recover. NOT ONE APPLICATION APP HAD TO BE REINSTALLED.  They
>all continued to work perfectly.  The key is, each app keeps all the files
>it needs in its own private directory/folder and no other app meddles with
>them. PERIOD.

>Granted, under linux there is a shared directory called /etc (similar to
>the Win 3.1 C:\WINDOWS\*.INI fdiles) where apps often collect text-format
>configuration files.  However, each app has separate files so it is easy to
>backup/restore them individually.

>If we got rid of that flipping registry, you could install a app under NT
>and access it under Win95 with no problem, and easily restore a single
>corrupted application.  If we got rid of sloppily shared DLLs we would get
>rid of 50+% of the problems with Windows apps failing.

>It registry would be fine IF each app had its own private piece of it
>stored with the rest of its files.  Apps should be in air tight boxes where
>they can't interfere with each other.  Communication should be via very
>rigid protocols to prevent malicious or incompetent  meddling with other
>apps.

Excellent points. I also believed the centralized registry concept to
be one that was just asking for trouble. There are some cool things
about the registry (per user settings, all user settings) but
keeping apps from messing each other up is not one of them.

Interesting point about the mini-reinstall to repair the registry for
a specific app. Perhaps a tool of some sort is in order--certainly
the demand would be there.

 
 
 

The IE4 bug that just won't go away.......

Post by Rob Eamo » Fri, 24 Apr 1998 04:00:00





>->This is a general problem across the PC industry -- componentized
software
>->depends heavily on other components being well-behaved.  Personally, I
think
>->its a tragedy that one ill-behaved application can damage another
>->well-behaved application by a simple bug in their uninstall routine.
We're
>->working on it -- in the future, it will probably be a "sticker"
requirement
>->that you use the much safer install-uninstall routines that we're
building
>->into the operating system.

>... or just use an intelligent operating system which doesn't allow this
>to happen.

Ok, so MS can add some facilities to track this sort of thing...

...then you'll * about software bloat.

 
 
 

The IE4 bug that just won't go away.......

Post by Rob Eamo » Fri, 24 Apr 1998 04:00:00



>Mr. Lippert, if the development team of one of Microsoft's most visible
>applications can't sufficient access to VB5 devlopers to get them to stop
>breaking their app, how is the rest of the world supposed to expect
>Microsoft to be responsive to their needs?

It's not a perfect world. How could they possibly be expected to
automatically catch this issue? The need for this level of
exhaustive testing is not practical nor necessary. Or are
you really willing to pay for this level of assurance??
 
 
 

The IE4 bug that just won't go away.......

Post by Mats Olss » Fri, 24 Apr 1998 04:00:00




Quote:>Richard,
>This is a general problem across the PC industry -- componentized software
>depends heavily on other components being well-behaved.  Personally, I think
>its a tragedy that one ill-behaved application can damage another
>well-behaved application by a simple bug in their uninstall routine.

    Yes, this is one of the core, unnecessary weaknesses in the whole
registry idea. Collect all cruical information in one file, with little
protection against corruption, and then act suprised when the inevitable
happens.

Quote:>We're
>working on it -- in the future, it will probably be a "sticker" requirement
>that you use the much safer install-uninstall routines that we're building
>into the operating system.

    Of course, this won't help with old software, or software which is
supposed to work on older variants of Windows who doesn't have these
safer install/uninstall, right? Yes, yet another reason to buy a new
OS and all new programs - otherwise the registry can get hurt...

    /Mats

 
 
 

The IE4 bug that just won't go away.......

Post by David Pearc » Fri, 24 Apr 1998 04:00:00


Quote:> Hey wait... didn't Win 3.1 do something like that?

> All things considered, I'm glad I develop applications instead of
> operating systems.

Its much worse than that.... what about all the DLL's that come with
program, and all the ones of the same name but disimilar versions that
come with other programs? I think a much more involved process would be
needed, one that completely blocks off direct access to the system
directory via the software's install program (or any program come to
think of it). Instead each package would have to authenticate to the
registry (or create a account on install). Then the install would
register each of its DLLs and each version number for each DLL. The OS
would actually perform the copy. If the version about to be installed is
different from any other DLL on the system, then the name of the DLL is
mangled and copied to system directory. When the application actually
runs and it loaos DLL's, it authenticates itself to the OS (or at least
some method is used for it to alert the operating system as to which set
of DLLs it will use), and the OS then only loads DLL's for that program.
As long as agreements between companies about compatibilities between
versions of DLL's, this scheme would allow the same kind of
interoperability that Win95/NT allows now, but without the *
crashing problems, and if companies don't agree, then they at least will
not be in a position to affect each other. To preserve backwards
compatibility a public system directory for non-authenticated programs
could be created, where as it would be like before, first come first
serve, and all old install programs would work, well at least as well as
they did before, anyhow.

David

 
 
 

1. Those Damn Users' Caches Won't Go Away ...

Hi,
I posted here a while back, asking for help on a script to delete
users' CACHE files, the location of these being $HOME/USERS/DEFAULT/CACHE

What I wanted to do, wasn't to delete the CACHE directory, but to remove
the files *within* the directory.

After a lot of help from various folk around here, I came up with this:

#!/bin/sh
LOCATION=USERS/DEFAULT/CACHE/
for LOCATION in `ls /home` ; do
rm /home/$LOCATION/USERS/DEFAULT/CACHE/*
   done

I installed that into /usr/local/bin, ran it, and hey presto, saw
/home go down to 75% in a matter of seconds.

Today, I come in, and find a problem report, saying that /home is
once more at 100%.  I check the disk, and sure enough, it's full.

I find this strange, as the script is in root's crontab, but just
to be sure, I connect, and run the script manually.

I then run a find for any files still in the users' CACHE, and see that
some of the them have not been deleted !!

Could someone shed some light on why a script that worked before,
should suddenly stop working ..?

Thanks.

--
Desmond Coughlan    Network Engineer    Forum des Images    Paris    France
***************************************************************************
The views expressed in these articles are my own, and do not necessarily
reflect the views of the Forum des Images.
***************************************************************************

2. Vi question

3. ext2 file system -- PERMENATELY screwed -- why won't they go away??

4. LXNY invited to Meeting of NYACC: 10 September 1998

5. Help: "boot:" won't go away

6. Windows network file access to server across the Internet ?

7. KDE Splash Won't go Away

8. VTOC or partition table of mirrored boot disks using VxVM

9. [Fwd: LILO won' t go away.]

10. CD player won't go away!

11. /tmp/hpnpf. files won't go away

12. /tmp directory - files won't go away (continued)

13. Processes won't go away