Has MS Crippled ActiveX in dotnet?

[Michael Culley]

This is a repost from the vb.dotnet.technical group (sorry phil). I've been
using dot net with a control that I wrote in VB6. There seem to be so many
bugs with using activeX controls in dotnet. It is like MS has intentially
crippled activeX controls to the point where they are completely un-usuable.
The worst bug is that HResult errors are ignored, making debugging
practically impossible.

What does everyone think? Has everyone just gone over to using dotnet
controls?

--
Michael Culley
www.vbdotcom.com

[Zane Thomas]

On Mon, 25 Mar 2002 08:38:00 +1100, "Michael Culley" <mike@vbdotcom.com>
wrote:

>It is like MS has intentially crippled activeX controls to the point
>where they are completely un-usuable.

Why would they do that?

>The worst bug is that HResult errors are ignored, making debugging
>practically impossible.

I haven't used COM interop, but I find it inconceivable that hresults are
ignored. A quick trip to the helpfile and a search for "hresult and
interop" turns up a number of things you might like to read:

"When the COM object returns a failure HRESULT, the .NET client catches a
corresponding exception."



--
Turn on, tune in, download.
zane@mvps.org

[Michael Culley]

> Why would they do that?

To make dot net controls more successful. If it worked really well then why
would someone upgrade GridX to GridX.net? This applies to both the author
and user of the control.

> "When the COM object returns a failure HRESULT, the .NET client catches a
> corresponding exception."

I know it is meant to do this but in the project I have the error is ignored
and code execution in the current stack is terminated. The interesting thing
is that I created a small sample project with the same activeX control and
it did turn the HResult into an exception.

Additionally, some properties are missing, others return errors for no
apparent reason, some events don't fire and the control goes invisible in
design. It is also messy to use - I am using my own grid control written in
VB6 and I get Cell, CellClass and _Cell (was just Cell in VB6), no big deal
but kinda confusing. The end result is that ActiveX controls are practically
unusable in dot net.

I really do like dot net but it just pisses me off that MS screw people for
their own gains.

--
Michael Culley
www.vbdotcom.com



"Zane Thomas" <zane@mabry.com> wrote in message
news:3cc34b11.353877218@news.devx.com...
> On Mon, 25 Mar 2002 08:38:00 +1100, "Michael Culley" <mike@vbdotcom.com>
> wrote:
>
> >It is like MS has intentially crippled activeX controls to the point
> >where they are completely un-usuable.
>
> Why would they do that?
>
> >The worst bug is that HResult errors are ignored, making debugging
> >practically impossible.
>
> I haven't used COM interop, but I find it inconceivable that hresults are
> ignored. A quick trip to the helpfile and a search for "hresult and
> interop" turns up a number of things you might like to read:
>
> "When the COM object returns a failure HRESULT, the .NET client catches a
> corresponding exception."
>
>
>
> --
> Turn on, tune in, download.
> zane@mvps.org

[Zane Thomas]

"Michael Culley" <mike@vbdotcom.com> wrote:

>> Why would they do that?
>
>To make dot net controls more successful. If it worked really well then why
>would someone upgrade GridX to GridX.net? This applies to both the author
>and user of the control.

That - pardon my french - is ridiculous. There are a lot of great new
features that can be implemented in .Net components and that's the reason
for upgrading. Furthermore, I think it's amazing that MS went to all the
trouble of providing for com interop since it actually has the effect of
*not* encouraging components to be rewritten for the .net framework.

You're engaging in unsubstantiated conspiracy theory-think of the
black-helicopter variety.


>> "When the COM object returns a failure HRESULT, the .NET client catches a
>> corresponding exception."
>
>I know it is meant to do this but in the project I have the error is ignored
>and code execution in the current stack is terminated.

Sounds like a bug in your use of the object.

>The interesting thing
>is that I created a small sample project with the same activeX control and
>it did turn the HResult into an exception.

Sounds even more like a bug.

>Additionally, some properties are missing, others return errors for no
>apparent reason, some events don't fire ...

Etc ... try dropping the conspiracy-theory stuff and post some specific
questions in the technical newsgroup. You might also try
microsoft.public.dotnet.framework.interop

>I really do like dot net but it just pisses me off that MS screw people for
>their own gains.

You have no facts upon which to base a claim that MS screwed people by
"intentially [sic] [crippling] activeX controls to the point where they
are completely un-usuable [sic]".


--
Turn on, tune in, download.
zane@mvps.org

[Michael Culley]

> Itr is highly unlikely that this the case.

I thinks it is extremely likely. MS are happy to kill any technology that
doesn't suit their purposes. Are you saying they have never done this
before?

> Assuming
> foul play without a shred of evidence implies a bit more malice in your
> heart due to unproven malice that you believe is in theirs?

OK, heres what I have. I have been using dot net for a couple of months now
and start to see a few bugs. I think the total is 6. Out of the six bugs I
found they all relate to activeX controls. I have not found one other bug in
all of dot net but 6 using activeX controls. These are the bugs:

Form_Load does not fire if it contains an activeX control when shown
modally.
HResults are not always turned into exceptions.
Some properties/methods are missing
Some properties give errors for no reason but other properties declared
exactly the same work.
ActiveX controls goes invisible at design time.
Keypreview property does not work on forms with activeX control on it.
Upgrade wizard fails to upgrade some projects that reference activeX
controls.

> So why not take a step back and then in a technical newsgroup you could
> REPORT the case where it seems like it is not working properly? Perhaps
> someone could either show you your error or prove Microsoft's (whichever
the
> case may be).

I have done this and found that the Form_Load not firing bug is in fact a
bug. I posted a question regarding some of the other bugs but did not get a
reply. I figured no-one is using com controls.

Believe me I am not a dot-notter, but it seems that anyone that critises dot
net gets lumped into the basket and flamed. If you tried using activeX
control you would find that it is pretty much unusable.


--
Michael Culley
www.vbdotcom.com



"Michael (michka) Kaplan" <former_mvp@nospam.trigeminal.spamless.com> wrote
in message news:3c9e60fa$1@10.1.10.29...
> "Michael Culley" <mike@vbdotcom.com> wrote...
>
> > I really do like dot net but it just pisses me off that MS screw people
> for
> > their own gains.
>
> Itr is highly unlikely that this the case. I can imagine there can be a
bug
> where in certain situations specific errors may not be properly reported
> (either for architectural limitation reasons or just because there is a
bug)
> but when you hear hoofbeats, you should expect HORSES, not ZEBRAS.
Assuming
> foul play without a shred of evidence implies a bit more malice in your
> heart due to unproven malice that you believe is in theirs?
>
> So why not take a step back and then in a technical newsgroup you could
> REPORT the case where it seems like it is not working properly? Perhaps
> someone could either show you your error or prove Microsoft's (whichever
the
> case may be).
>
>
> --
> MichKa
>
> Michael Kaplan
> Trigeminal Software, Inc. -- http://www.trigeminal.com/
>
> International VB? -- http://www.i18nWithVB.com/
> C++? MSLU -- http://msdn.microsoft.com/msdnmag/issues/01/10/
>
>
>

[Tom Bennet]

Sounds as though something you are doing has corrupted the stack. The com
interop is not perfect, but it has seemed to work with every control that
I have attempted to use it with and many of those are created in VB. Though
I admit I have yet to push any .Net programs into production.

"Michael Culley" <mike@vbdotcom.com> wrote:
>This is a repost from the vb.dotnet.technical group (sorry phil). I've
been
>using dot net with a control that I wrote in VB6. There seem to be so many
>bugs with using activeX controls in dotnet. It is like MS has intentially
>crippled activeX controls to the point where they are completely un-usuable.
>The worst bug is that HResult errors are ignored, making debugging
>practically impossible.
>
>What does everyone think? Has everyone just gone over to using dotnet
>controls?
>
>--
>Michael Culley
>www.vbdotcom.com
>
>
>
>
>

[Tom Bennet]

Yep, you are only entitled to the popular opinion and the rest of us be damned.
They call us fools and trolls and send us on our way.

I think some of these guys have made a career of defending this thing to
the bitter end. Many on both sides have written more lines of BS in here
than they have lines of code.

There is a lot of emotion on both sides of the fence.

"Michael Culley" <mike@vbdotcom.com> wrote:
>> That - pardon my french - is ridiculous.
>
>Yes and MS didn't intentionally kill netscape.
>
>> You have no facts upon which to base a claim that MS screwed people by
>> "intentially [sic] [crippling] activeX controls to the point where they
>> are completely un-usuable [sic]".
>
>Believe me Zane, I love dot net, as I get into it more and more I am amazed
>at all the incredible features they have added. But it seems that if anyone
>here critises dot net in the slightest that they get flammed to the childish
>extent of critising their spelling.
>
>--
>Michael Culley
>www.vbdotcom.com
>
>
>
>"Zane Thomas" <zane@mabry.com> wrote in message
>news:3cc46b9d.362209000@news.devx.com...
>> "Michael Culley" <mike@vbdotcom.com> wrote:
>>
>> >> Why would they do that?
>> >
>> >To make dot net controls more successful. If it worked really well then
>why
>> >would someone upgrade GridX to GridX.net? This applies to both the author
>> >and user of the control.
>>
>> That - pardon my french - is ridiculous. There are a lot of great new
>> features that can be implemented in .Net components and that's the reason
>> for upgrading. Furthermore, I think it's amazing that MS went to all
the
>> trouble of providing for com interop since it actually has the effect
of
>> *not* encouraging components to be rewritten for the .net framework.
>>
>> You're engaging in unsubstantiated conspiracy theory-think of the
>> black-helicopter variety.
>>
>>
>> >> "When the COM object returns a failure HRESULT, the .NET client catches
>a
>> >> corresponding exception."
>> >
>> >I know it is meant to do this but in the project I have the error is
>ignored
>> >and code execution in the current stack is terminated.
>>
>> Sounds like a bug in your use of the object.
>>
>> >The interesting thing
>> >is that I created a small sample project with the same activeX control
>and
>> >it did turn the HResult into an exception.
>>
>> Sounds even more like a bug.
>>
>> >Additionally, some properties are missing, others return errors for no
>> >apparent reason, some events don't fire ...
>>
>> Etc ... try dropping the conspiracy-theory stuff and post some specific
>> questions in the technical newsgroup. You might also try
>> microsoft.public.dotnet.framework.interop
>>
>> >I really do like dot net but it just pisses me off that MS screw people
>for
>> >their own gains.
>>
>> You have no facts upon which to base a claim that MS screwed people by
>> "intentially [sic] [crippling] activeX controls to the point where they
>> are completely un-usuable [sic]".
>>
>>
>> --
>> Turn on, tune in, download.
>> zane@mvps.org
>
>

[Mike Mitchell]

On Mon, 25 Mar 2002 10:28:32 +1100, "Michael Culley"
<mike@vbdotcom.com> wrote:

>> Why would they do that?
>
>To make dot net controls more successful. If it worked really well then why
>would someone upgrade GridX to GridX.net? This applies to both the author
>and user of the control.

Hey, this sounds like a job for Kunle's Redeemers, the legal guys he
was talking about who would go in and deem something was
anti-competitive! Like Guardian Angels for software.

>I really do like dot net but it just pisses me off that MS screw people for
>their own gains.

Oh, surely not! I cannot believe you could even suggest such a thing!

MM

[Michael Culley]

> That - pardon my french - is ridiculous.

Yes and MS didn't intentionally kill netscape.

> You have no facts upon which to base a claim that MS screwed people by
> "intentially [sic] [crippling] activeX controls to the point where they
> are completely un-usuable [sic]".

Believe me Zane, I love dot net, as I get into it more and more I am amazed
at all the incredible features they have added. But it seems that if anyone
here critises dot net in the slightest that they get flammed to the childish
extent of critising their spelling.

--
Michael Culley
www.vbdotcom.com



"Zane Thomas" <zane@mabry.com> wrote in message
news:3cc46b9d.362209000@news.devx.com...
> "Michael Culley" <mike@vbdotcom.com> wrote:
>
> >> Why would they do that?
> >
> >To make dot net controls more successful. If it worked really well then
why
> >would someone upgrade GridX to GridX.net? This applies to both the author
> >and user of the control.
>
> That - pardon my french - is ridiculous. There are a lot of great new
> features that can be implemented in .Net components and that's the reason
> for upgrading. Furthermore, I think it's amazing that MS went to all the
> trouble of providing for com interop since it actually has the effect of
> *not* encouraging components to be rewritten for the .net framework.
>
> You're engaging in unsubstantiated conspiracy theory-think of the
> black-helicopter variety.
>
>
> >> "When the COM object returns a failure HRESULT, the .NET client catches
a
> >> corresponding exception."
> >
> >I know it is meant to do this but in the project I have the error is
ignored
> >and code execution in the current stack is terminated.
>
> Sounds like a bug in your use of the object.
>
> >The interesting thing
> >is that I created a small sample project with the same activeX control
and
> >it did turn the HResult into an exception.
>
> Sounds even more like a bug.
>
> >Additionally, some properties are missing, others return errors for no
> >apparent reason, some events don't fire ...
>
> Etc ... try dropping the conspiracy-theory stuff and post some specific
> questions in the technical newsgroup. You might also try
> microsoft.public.dotnet.framework.interop
>
> >I really do like dot net but it just pisses me off that MS screw people
for
> >their own gains.
>
> You have no facts upon which to base a claim that MS screwed people by
> "intentially [sic] [crippling] activeX controls to the point where they
> are completely un-usuable [sic]".
>
>
> --
> Turn on, tune in, download.
> zane@mvps.org

[Tom Bennet]

If it is a stack corruption, it is probably due to a parameter that you are
passing the ActiveX control. It's pretty hard, if not impossible to corrupt
the stack using managed code alone.


"Michael Culley" <mike@vbdotcom.com> wrote:
>> Sounds as though something you are doing has corrupted the stack.
>
>This problem has persisted after a restart but maybe it is corrupting the
>stack every time it runs. But similar code works in another project.
>
>--
>Michael Culley
>www.vbdotcom.com

[Michael Culley]

> Sounds as though something you are doing has corrupted the stack.

This problem has persisted after a restart but maybe it is corrupting the
stack every time it runs. But similar code works in another project.

--
Michael Culley
www.vbdotcom.com



"Tom Bennet" <fdsfsd@fdsfs.com> wrote in message
news:3c9e71cd$1@10.1.10.29...
>
> Sounds as though something you are doing has corrupted the stack. The com
> interop is not perfect, but it has seemed to work with every control that
> I have attempted to use it with and many of those are created in VB.
Though
> I admit I have yet to push any .Net programs into production.
>
> "Michael Culley" <mike@vbdotcom.com> wrote:
> >This is a repost from the vb.dotnet.technical group (sorry phil). I've
> been
> >using dot net with a control that I wrote in VB6. There seem to be so
many
> >bugs with using activeX controls in dotnet. It is like MS has intentially
> >crippled activeX controls to the point where they are completely
un-usuable.
> >The worst bug is that HResult errors are ignored, making debugging
> >practically impossible.
> >
> >What does everyone think? Has everyone just gone over to using dotnet
> >controls?
> >
> >--
> >Michael Culley
> >www.vbdotcom.com
> >
> >
> >
> >
> >
>

[Kunle Odutola]

"Mike Mitchell" <kylix_is@yahoo.co.uk> wrote in message
news:3c9e7270.27952017@news.devx.com...

> Hey, this sounds like a job for Kunle's Redeemers, the legal guys he
> was talking about who would go in and deem something was
> anti-competitive! Like Guardian Angels for software.

Gross misrepresentation. I said nothing of the sort!
Ya pants on fire yet Mike?

Kunle

[Michael Culley]

Wrong, I have posted 5 out of six of the bugs in devx technical. I don't use
ms newsgroups because I can generally get the answer I need here.

> Clearly your priorities are not on getting fixes to behavior; perhaps once
> you analyze why you made these three choices you will determine what your
> motives are? :-)

My motives for posting the question here was to find out if activeX controls
had been crippled. If they had then I would know not to persist. Please
don't presume you know my motives.

--
Michael Culley
www.vbdotcom.com

[Michael Culley]

could you explain what you will say if any of those ix
> bugs trun out to tbe YOUR bugs?

One of the bugs has already been confirmed to be microsoft's. Maybe some
could be something I'm doing.

> Will you publicly apologize for accusing MS
> the way you did? :-)

Depends if you will publicly apoligize to me if what I say comes out to be
true.

> As I mentioned before, you are not showing any indicaion that you are
> looking for solutions....

As I said, I was looking for an answer to the question of whether or not com
controls worked ok in dot net. If someone said "yep, they work fine for me"
then I would persist with finding bugs. Notice no-one has said this yet. If
someone said they didn't work then I would start re-writing my controls.

> maybe you could change that?

I have posted the list of bugs.

--
Michael Culley
www.vbdotcom.com



"Michael (michka) Kaplan" <former_mvp@nospam.trigeminal.spamless.com> wrote
in message news:3c9e7547@10.1.10.29...
> If you8 hav a stack corrupting bug, I doube a restart of the app or a
reboot
> of the machine would have any effect.
>
> But you hav now STATED that your attempt at simple repro fails and
> everything works. So somewhere between your simple repro and your app, the
> bug lies.
>
> Now that you have accused MS of stunting a technology just because you
could
> not make it work, could you explain what you will say if any of those ix
> bugs trun out to tbe YOUR bugs? Will you publicly apologize for accusing
MS
> the way you did? :-)
>
> As I mentioned before, you are not showing any indicaion that you are
> looking for solutions.... maybe you could change that?
>
>
> --
> MichKa
>
> Michael Kaplan
> Trigeminal Software, Inc. -- http://www.trigeminal.com/
>
> International VB? -- http://www.i18nWithVB.com/
> C++? MSLU -- http://msdn.microsoft.com/msdnmag/issues/01/10/
>
>
> "Michael Culley" <mike@vbdotcom.com> wrote in message
> news:3c9e7320@10.1.10.29...
> > > Sounds as though something you are doing has corrupted the stack.
> >
> > This problem has persisted after a restart but maybe it is corrupting
the
> > stack every time it runs. But similar code works in another project.
> >
> > --
> > Michael Culley
> > www.vbdotcom.com
> >
> >
> >
> > "Tom Bennet" <fdsfsd@fdsfs.com> wrote in message
> > news:3c9e71cd$1@10.1.10.29...
> > >
> > > Sounds as though something you are doing has corrupted the stack. The
> com
> > > interop is not perfect, but it has seemed to work with every control
> that
> > > I have attempted to use it with and many of those are created in VB.
> > Though
> > > I admit I have yet to push any .Net programs into production.
> > >
> > > "Michael Culley" <mike@vbdotcom.com> wrote:
> > > >This is a repost from the vb.dotnet.technical group (sorry phil).
I've
> > > been
> > > >using dot net with a control that I wrote in VB6. There seem to be so
> > many
> > > >bugs with using activeX controls in dotnet. It is like MS has
> intentially
> > > >crippled activeX controls to the point where they are completely
> > un-usuable.
> > > >The worst bug is that HResult errors are ignored, making debugging
> > > >practically impossible.
> > > >
> > > >What does everyone think? Has everyone just gone over to using dotnet
> > > >controls?
> > > >
> > > >--
> > > >Michael Culley
> > > >www.vbdotcom.com
> > > >
> > > >
> > > >
> > > >
> > > >
> > >
> >
> >
>
>

[Michael Culley]

I really don't understand why you are so touchy about me criticising
microsoft. They do alot of great stuff but also do alot of dodgy stuff and
sometimes deserve critisism. Note that you are labelling my post a
"conspiracy theory" without knowing yourself that it is not true. In other
words you are doing exactly what you are acusing me of.

--
Michael Culley
www.vbdotcom.com



"Michael (michka) Kaplan" <former_mvp@nospam.trigeminal.spamless.com> wrote
in message news:3c9e760b@10.1.10.29...
> "Michael Culley" <mike@vbdotcom.com> wrote...
> > Wrong, I have posted 5 out of six of the bugs in devx technical. I don't
> use
> > ms newsgroups because I can generally get the answer I need here.
>
> Um, you are the peson say you had six bugs and you post on one of them.
> Don't get all shirty with me if I follow your lead? :-)
>
> So, your goal is not to have bugs reported but only to get things working
> for you. Very selfless of you! <g>
>
> > > Clearly your priorities are not on getting fixes to behavior; perhaps
> once
> > > you analyze why you made these three choices you will determine what
> your
> > > motives are? :-)
> >
> > My motives for posting the question here was to find out if activeX
> controls
> > had been crippled. If they had then I would know not to persist. Please
> > don't presume you know my motives.
>
> Microsoft has not crippled activex controls. Conspiracy theorists (such as
> yourself?) can believe otherwise, but maybe you should track down the bugs
> before you presume to ascribe motives to them!!!
>
> --
> MichKa
>
> Michael Kaplan
> Trigeminal Software, Inc. -- http://www.trigeminal.com/
>
> International VB? -- http://www.i18nWithVB.com/
> C++? MSLU -- http://msdn.microsoft.com/msdnmag/issues/01/10/
>
>
>