Is COM going away?

[Matt Markus]

Someone told me that... is it true?

[ian drake]

"Matt Markus" <matt_markus@hotmail.com> wrote:
>Someone told me that... is it true?

False. COM is just not the mechanism of choice for components to "describe"
themselves anymore. Which means no more hasles of un/registering DLLs and
rebooting. There's a new way of versoning that is better and doesn't involve
the registry!

You CAN create a TLB for your .NET components so they can be used w/COM+
or other environments that need a COM interface. You need to run a utility
on it though, it's not done by the compiler.

I would do some reading at the MSDN site for more in-depth info. This subject
has been visited on many threads by now, and I don't think a lot people will
go into detail on it again, but I've been known to be wrong.

Ian Drake

[Michael \(michka\) Kaplan]

COM itself is not, as the system heavily relies on it and none of THAT is
changing.

..NET does go to some lengths to hide COM from us, though.

--
MichKa

a new book on internationalization in VB at
http://www.i18nWithVB.com/

"Matt Markus" <matt_markus@hotmail.com> wrote in message
news:3a34fc2d@news.devx.com...
> Someone told me that... is it true?
>
>

[Richard Purchas]

Whats happening with the whole COM+ / MTS concept then, and all those
wonderful things they told us about JIT instantiation ? .... not to mention
"transactions" ! Will .NET components be able to be hosted in COM+ / MTS ?

Can anyone point me to some good reading of how one now builds MTS/COM+
enabled components in .NET.

Thanks

"Michael (michka) Kaplan" <former_mvp@spamfree.trigeminal.nospam.com> wrote
in message news:3a350863@news.devx.com...
> COM itself is not, as the system heavily relies on it and none of THAT is
> changing.
>
> .NET does go to some lengths to hide COM from us, though.
>
> --
> MichKa
>
> a new book on internationalization in VB at
> http://www.i18nWithVB.com/
>
> "Matt Markus" <matt_markus@hotmail.com> wrote in message
> news:3a34fc2d@news.devx.com...
> > Someone told me that... is it true?
> >
> >
>
>

[Anthony Boudouvas]

Michael (michka) Kaplan <former_mvp@spamfree.trigeminal.nospam.com> wrote in
message news:3a350863@news.devx.com...
> COM itself is not, as the system heavily relies on it and none of THAT is
> changing.
>
> .NET does go to some lengths to hide COM from us, though.
>
> --
> MichKa


It does not only hide COM but also NOT heavily rely on it...


anthonyb

[Michael \(michka\) Kaplan]

Oh sure, things like OLE DB (which they are now saying is used in
conjunction with ADO.NET), the Shell, and all the rest of the system
services that VS.NET has to use somehow to do not constitute a heavy
reliance.

Well, I won't argue. The .NET framework and VS setup install so many
COM-related reg keys I think the jury will have no problems deciding which
of us is right when they look at the evidence, Anthony. :-)

--
MichKa

a new book on internationalization in VB at
http://www.i18nWithVB.com/

"Anthony Boudouvas" <mediatel@ath.forthnet.gr (anthonyb)> wrote in message
news:3a35f628@news.devx.com...
> Michael (michka) Kaplan <former_mvp@spamfree.trigeminal.nospam.com> wrote
in
> message news:3a350863@news.devx.com...
> > COM itself is not, as the system heavily relies on it and none of THAT
is
> > changing.
> >
> > .NET does go to some lengths to hide COM from us, though.
> >
> > --
> > MichKa
>
>
> It does not only hide COM but also NOT heavily rely on it...
>
>
> anthonyb
>
>

[Maxime Bombardier]

Sure, read the post COM Components? - complete* from myself on December 11th...
you had a complete COM example using a COM+ application... but the explanation
is quite big.

.NET Application can be hosted by COM+ but so far, I can't call them from
a .NET application, only from a COM application (i.e.: VBScript, VB6, etc.)
if you are creating them with .NET. For a .NET application to talk to a
different .NET application, you have to put it inside the Global Assembly
Cache.



"Richard Purchas" <rpurchas@factotum.com.au> wrote:
>Whats happening with the whole COM+ / MTS concept then, and all those
>wonderful things they told us about JIT instantiation ? .... not to mention
>"transactions" ! Will .NET components be able to be hosted in COM+ / MTS
?
>
>Can anyone point me to some good reading of how one now builds MTS/COM+
>enabled components in .NET.
>
>Thanks
>
>"Michael (michka) Kaplan" <former_mvp@spamfree.trigeminal.nospam.com> wrote
>in message news:3a350863@news.devx.com...
>> COM itself is not, as the system heavily relies on it and none of THAT
is
>> changing.
>>
>> .NET does go to some lengths to hide COM from us, though.
>>
>> --
>> MichKa
>>
>> a new book on internationalization in VB at
>> http://www.i18nWithVB.com/
>>
>> "Matt Markus" <matt_markus@hotmail.com> wrote in message
>> news:3a34fc2d@news.devx.com...
>> > Someone told me that... is it true?
>> >
>> >
>>
>>
>
>

[Joe \Nuke Me Xemu\ Foster]

"Anthony Boudouvas" <mediatel@ath.forthnet.gr (anthonyb)> wrote in message news:3a35f628@news.devx.com...

> Michael (michka) Kaplan <former_mvp@spamfree.trigeminal.nospam.com> wrote in
> message news:3a350863@news.devx.com...
> > COM itself is not, as the system heavily relies on it and none of THAT is
> > changing.
> >
> > .NET does go to some lengths to hide COM from us, though.
> >
> > --
> > MichKa
>
>
> It does not only hide COM but also NOT heavily rely on it...

I've mainly been lurking to try to figure out just what DOTNET will do
for/to us that COM/COM+ cannot, but either the bottom line is hidden
under too many layers of marketing-speak, or all the good stuff is under
NDA. =( However, it does look like DOTNET really is different and not
just COM in a cape and tights, a la OLE and ActiveX. Why didn't MS just
fix COM, or would that run into Total Cost of Ownership issues? (Can't
fix the bugs, because that would break the workarounds...)

--
Joe Foster <mailto:jfoster@ricochet.net> Space Cooties! <http://www.xenu.net/>
WARNING: I cannot be held responsible for the above They're coming to
because my cats have apparently learned to type. take me away, ha ha!

[Michael \(michka\) Kaplan]

Joe,

They have entirely broken almost everything from prior versions. Viosual
Fred (aka VB.Net) is a whole new language.

It has cool features, but there is no backwards compatibility whatsoever.

--
MichKa

a new book on internationalization in VB at
http://www.i18nWithVB.com/

"Joe "Nuke Me Xemu" Foster" <joe@bftsi0.UUCP> wrote in message
news:3a36cfc7$1@news.devx.com...
> "Anthony Boudouvas" <mediatel@ath.forthnet.gr (anthonyb)> wrote in message
news:3a35f628@news.devx.com...
>
> > Michael (michka) Kaplan <former_mvp@spamfree.trigeminal.nospam.com>
wrote in
> > message news:3a350863@news.devx.com...
> > > COM itself is not, as the system heavily relies on it and none of THAT
is
> > > changing.
> > >
> > > .NET does go to some lengths to hide COM from us, though.
> > >
> > > --
> > > MichKa
> >
> >
> > It does not only hide COM but also NOT heavily rely on it...
>
> I've mainly been lurking to try to figure out just what DOTNET will do
> for/to us that COM/COM+ cannot, but either the bottom line is hidden
> under too many layers of marketing-speak, or all the good stuff is under
> NDA. =( However, it does look like DOTNET really is different and not
> just COM in a cape and tights, a la OLE and ActiveX. Why didn't MS just
> fix COM, or would that run into Total Cost of Ownership issues? (Can't
> fix the bugs, because that would break the workarounds...)
>
> --
> Joe Foster <mailto:jfoster@ricochet.net> Space Cooties!
<http://www.xenu.net/>
> WARNING: I cannot be held responsible for the above They're
coming to
> because my cats have apparently learned to type. take me away,
ha ha!
>
>

[Joe \Nuke Me Xemu\ Foster]

"Michael (michka) Kaplan" <former_mvp@spamfree.trigeminal.nospam.com> wrote in message news:3a36d364$1@news.devx.com...

> Joe,
>
> They have entirely broken almost everything from prior versions. Viosual
> Fred (aka VB.Net) is a whole new language.
>
> It has cool features, but there is no backwards compatibility whatsoever.

Sorry, I should have been clearer. If MS had decided to base Fred and C+
on COM, would the changes to COM to support all the new buzzwords have
broken VS6?

--
Joe Foster <mailto:jfoster@ricochet.net> Space Cooties! <http://www.xenu.net/>
WARNING: I cannot be held responsible for the above They're coming to
because my cats have apparently learned to type. take me away, ha ha!

[Michael \(michka\) Kaplan]

Oh, probably not. But they hired a Java/Delphi man, who convinced them to
change it all.

--
MichKa

a new book on internationalization in VB at
http://www.i18nWithVB.com/

"Joe "Nuke Me Xemu" Foster" <joe@bftsi0.UUCP> wrote in message
news:3a36d90a$1@news.devx.com...
> "Michael (michka) Kaplan" <former_mvp@spamfree.trigeminal.nospam.com>
wrote in message news:3a36d364$1@news.devx.com...
>
> > Joe,
> >
> > They have entirely broken almost everything from prior versions. Viosual
> > Fred (aka VB.Net) is a whole new language.
> >
> > It has cool features, but there is no backwards compatibility
whatsoever.
>
> Sorry, I should have been clearer. If MS had decided to base Fred and C+
> on COM, would the changes to COM to support all the new buzzwords have
> broken VS6?
>
> --
> Joe Foster <mailto:jfoster@ricochet.net> Space Cooties!
<http://www.xenu.net/>
> WARNING: I cannot be held responsible for the above They're
coming to
> because my cats have apparently learned to type. take me away,
ha ha!
>
>

[Rob Teixeira]

"Anthony Boudouvas" <mediatel@ath.forthnet.gr (anthonyb)> wrote:
>Michael (michka) Kaplan <former_mvp@spamfree.trigeminal.nospam.com> wrote
in
>message news:3a350863@news.devx.com...
>> COM itself is not, as the system heavily relies on it and none of THAT
is
>> changing.
>>
>> .NET does go to some lengths to hide COM from us, though.
>>
>> --
>> MichKa
>
>
>It does not only hide COM but also NOT heavily rely on it...
>
>
>anthonyb
>


If you have a few spare days on your hands, run depends.exe on the DLL's
in your system/system32/shared componentes directory. Then count all the
ones that import from olepro, oleaut, and stdole. Then count all the ones
that import from them.

When you use parts of the framework that wrap some of these functions in
these DLL's, guess what... you're using COM

I agree with Michka, COM isn't going away ANY time soon. At some point, MS
*might* make an OS that is completely 1:1 with the framework with no wrappers
and built around/with/for the CLR. But look at all the work that would go
into making a .NET version of windows compatible with all the existing (non-.NET)
stuff.

-Rob

[Joe \Nuke Me Xemu\ Foster]

"Michael (michka) Kaplan" <former_mvp@spamfree.trigeminal.nospam.com> wrote in message news:3a36db62$1@news.devx.com...

> Oh, probably not. But they hired a Java/Delphi man, who convinced them to
> change it all.

This was all the idea of Anders Hejlsberg, AKA Creutzfeldt-Delphi Patient
Zero?! Awww geez, the DOJ isn't trying to kill MS, it's already dead, and
Janet Reno's merely one of the hyenas trying to fill its belly before the
corpse stops twitching and gets all gamy. Looks like the Borland/Inprise
hydrophobiacs finally found their VB-Killer -- VB.NET! Too bad they didn't
try it by actually making a competitive development system...

--
Joe Foster <mailto:jfoster@ricochet.net> Space Cooties! <http://www.xenu.net/>
WARNING: I cannot be held responsible for the above They're coming to
because my cats have apparently learned to type. take me away, ha ha!

[David Bayley]

MichKa,

> They have entirely broken almost everything from prior versions. Viosual
> Fred (aka VB.Net) is a whole new language.

The language is pretty much the same, although not backwards compatible.
What has changed is the underlying architecture and frameworks. IMO, the
language had little choice but to break compatibility, but it IS still the
same language.

Joe: Yes, COM is going away. Slowly but surely, it will be killed off IMO.
This isn't going to happen any time soon, just as DOS is still around in
Win9x today. But some time in the not so distant future, a new Win.NET
platform will arrive (maybe Blackcomb, maybe the one after, who knows) that
will make WinNT legacy, and will merely emulate COM for backwards
compatibility. "WinCE.NET" however, sans-COM, may arrive a bit sooner.

--
David.

[Michael \(michka\) Kaplan]

"David Bayley" <dbayley@aebacus.com> wrote in message
news:3a380429@news.devx.com...

> The language is pretty much the same, although not backwards compatible.
> What has changed is the underlying architecture and frameworks. IMO, the
> language had little choice but to break compatibility, but it IS still the
> same language.

An interesting opinion.... the language is not backwards compatible and not
even a conversion tool can make old code work in most cases, the basic
language semantics have changed, yet you believe it to be the same language.
No worries, mate... you're assimilated into the collective!

> Joe: Yes, COM is going away. Slowly but surely, it will be killed off
IMO.
> This isn't going to happen any time soon, just as DOS is still around in
> Win9x today. But some time in the not so distant future, a new Win.NET
> platform will arrive (maybe Blackcomb, maybe the one after, who knows)
that
> will make WinNT legacy, and will merely emulate COM for backwards
> compatibility. "WinCE.NET" however, sans-COM, may arrive a bit sooner.

Again, your opinion. COM is still very mjuch a part of the OS and no one is
even considering rewriting code here. They are just hiding it from the
developer, thats all. An interesting opinion you have there, thankfully it
is not one that the people *at* Microsoft are claiming (as they would not
want their developers to burn them at the stake? <g>).

--
MichKa

a new book on internationalization in VB at
http://www.i18nWithVB.com/