Microsoft not content with "dissing" just the Classic VB Developer Army....

[Jim Hubbard]

Me to. Don't get me wrong. I like and use MS products - but, I also
like and use Linux. It's nice to have something that will work on both.

I have been a long time fan of Microsoft and their products (although these
posts may make that less than clear at times). But, with the direction they
have taken VB.Net (which is anti-RAD, IMHO) and the continued rising costs
Microsoft products that, let's be honest, really don't add that much value
for the average office user (like MS Office 2002 to 2003) and with the trend
towards intentionally breaking backwards compatibility - we need another
option.

For your enjoyment...

http://mono-project.com/FAQ:_General#The_Novell_Role_in_the_Mono_Project
Will Novell offer Mono commercially?

Novell will offer a commercial support and services for Mono. Mono
components are also available to be licensed commercially. For
licensing details, contact (e-mail address removed)

Thanks again!

To be totally up front... I haven't been all that impressed with SuSE.
It's not that it's bad - I just don't like RPM based packaging. I have
really fallen in love with my Gentoo system and it's portage package
manager.

I'll have to try Gentoo......but, they'll have a hurdle to make it easier
than Linspire - which I installed 4 times in the last 3 days. Linspire is
the easiest Linux system I have ever seen. The 5.0 version desktop is very
user friendly and the CNR (Click N Run) way that you can instll new
applications from their library by just clicking on them is just great.

The only problem I see is a lack of polished software when compared to
Windows. That's why I think that a RAD IDE like REALbasic would help Linux
a great deal. The more poeple we have coding for Linux, the more polished
apps we will have to choose from, the more viable the Linux desktop becomes.

I support any IDE that makes Linux easier to write polished, professional
applications on. I do so because I believe that competition is good for
consumers.

Pepsi is good for Coke (or we'd probably be still stuck with that "New Coke"
crap). Canon is good for Xerox - it keeps them on their toes. Many choices
in cars makes the manufacturers build better (and less expensive) cars. We
need that competition for the desktop too. And, if I can do anything to
helop bring it about - I will.

Like recommending that anyone reading this thread that has not downloaded a
FREE copy of REALbasic do so BEFORE THE APRIL 15, 2005 DEADLINE by clicking
http://www.realbasic.com/vb6/index.php?id=GVVDPQFY .

Thanks, Tom, for the Mono information.

Jim Hubbard

 

[Jim Hubbard]

Careful with that. They are a business too. Just because they say they
listen to the developer doesn't always make that the case. I've seen, and
heard that before.

This is where competition is helpful to consumers. We need more than one
option. Then, consumers can go where they are listened to.

I have spoken with the President of Real Software (Geoff Perlman) and he
seems to understand the need to keep the customers in charge of the
direction of the product. From the conversations and emails that we have
traded, I have to say that I have faith in him that this is true if him, his
company and team.

Just like every other linux project that crops up, there is always
community support, just like what you get with Microsoft. And I'm sure a
ton of support companies will crop up once it starts to get adopted more.

That's a good thing.

So, if it's so great, why are we all still arguing? Why don't the people
that have a big problem with Microsoft dumping classic VB just switch to
real basic?

Who says they're not?

Perhaps people are rushing to download their FREE copy of REALbasic BEFORE
THE APRIL 15, 2005 DEADLINE by clicking on
http://www.realbasic.com/vb6/index.php?id=GVVDPQFY . We'll see.

Jim Hubbard

 

[Tom Shelton]

I have been a long time fan of Microsoft and their products (although these
posts may make that less than clear at times). But, with the direction they
have taken VB.Net (which is anti-RAD, IMHO) and the continued rising costs
Microsoft products that, let's be honest, really don't add that much value
for the average office user (like MS Office 2002 to 2003) and with the trend
towards intentionally breaking backwards compatibility - we need another
option.



Thanks again!



I'll have to try Gentoo......but, they'll have a hurdle to make it easier
than Linspire - which I installed 4 times in the last 3 days.

Actually... Gentoo doesn't even try to make it easier Gentoo is a
source based distro. It has no graphical installer, like most distro's.
In fact, it doesn't really have what you would call an installer at all.
You basically, boot from a cd, manually partition your disks, download a
source tarball based on the level of install you want, and the run some
scripts to compile stuff. The documentation is pretty good - but you
HAVE to follow it. You can't just skip ahead because you think you know
something.

The thing I like about Gentoo is the fact that it is pretty minimal in
it's base form (it doesn't even install X by default). I can add stuff
as I need it with a simple command from the command line, and it works
out and installs all the dependencies... If you ever worked with a
FreeBSD system and there ports collection, portage is pertty similar.

Linspire is
the easiest Linux system I have ever seen. The 5.0 version desktop is very
user friendly and the CNR (Click N Run) way that you can instll new
applications from their library by just clicking on them is just great.

Does it still have you run as root by default? That was one of it's bad
points in the past...

The only problem I see is a lack of polished software when compared to
Windows. That's why I think that a RAD IDE like REALbasic would help Linux
a great deal. The more poeple we have coding for Linux, the more polished
apps we will have to choose from, the more viable the Linux desktop becomes.

I support any IDE that makes Linux easier to write polished, professional
applications on. I do so because I believe that competition is good for
consumers.

That I agree on. It's one of the main reasons I support Linux. It's
not that I dislike MS or that I want to see them go away... I just want
to see them have to compete a little

 

[Tom Shelton]

James - thanks. I'll have to check that out...

Just a follow up... I installed gambas immediately after posting this.
First impression - definately worth a closer look.

 

[Paul Clement]

¤ <snip>
¤
¤ > ¤ While still far behind Windows, the demand for Linux is growing by leaps and
¤ > ¤ bounds....if I may....
¤ > ¤
¤ >
¤ > Well that's what some folks having been saying for the last five years. You would have thought by
¤ > now that Linux would have passed up Windows by now. ;-)
¤ >
¤
¤ Do you read the news much? Linux is being adopted... Not so much on the
¤ desktop (though that is happening) - but on the server side it is
¤ growing pretty darn fast.
¤

Yes, it's being adopted by Unix users. Windows has typically lagged behind in
the server market but that is changing as well since it's continued to gain
market share in that venue.

I'm afraid the slaughter rule still applies on the desktop.

¤ > ¤ It's like the adoption of Firefox in place of IE. Firefox is making great
¤ > ¤ strides in the browser market, with no signs of stalling. People will adopt
¤ > ¤ the best technology for their enterprise, whether that is MAC, Windows or
¤ > ¤ Linux.
¤ > ¤
¤ >
¤ > Don't get me started on the Firefox issue. As market share increases it becomes a much bigger target
¤ > to hackers and those looking to exploit security holes. If probably won't help that MS is now
¤ > working on an updated version of IE.
¤ >
¤
¤ I agree that Firefox is not as big a target at the moment... But, you
¤ need to look at the fix rate as well:
¤
¤ http://www.tigertools.net/board/?topic=topic8&msg=7567
¤ According to Brussels-based ScanIT, users of Microsoft's Internet
¤ Explorer (IE) were "unsafe" 98 percent of the time during 2004, while
¤ Mozilla users -- which would include those using Mozilla and Firefox
¤ -- were "unsafe" only 15 percent of last year.
¤
¤ The fact is that Firefox IS more secure, not to mention it is just a
¤ better browser. I have high hopes for IE7, but right now Firefox just
¤ plain kicks IE's butt.
¤

We won't know how secure FireFox is until it becomes a much bigger target. I'm
sure we all thought IE was secure at one point, probably when it was overtaking
Netscape. But the fact the FireFox does not support some features targeted as
security risks you could say it's inherently more secure in that sense.

Personally I've found some compatibility issues with FireFox that annoy me. In
addition, it doesn't have the rendering smarts that IE does. If try FireFox
periodically but I always end up going back to IE when I hit a usability road
block.

¤ > ¤ The adoption of Linux will happen sooner than you think, in more places than
¤ > ¤ you think. There are things in the works right now that will make Linux the
¤ > ¤ premier desktop of small and mid-sized businesses worldwide. Add them to
¤ > ¤ the governments making the switch, and you have yourself a little
¤ > ¤ revolution.
¤ > ¤
¤ >
¤ > They way Linux has been hyped over the last several years I would have expected a significantly
¤ > higher adoption rate. Problem is there's literally no money to be made in this market in comparison
¤
¤ Tell that to Novell and IBM. Both major Linux players.
¤

Novell has been on life support for quite a while now. IBM is getting out of the
PC hardware business. They also both have a grudge against Microsoft.

¤ > to the Windows market so quality applications lag behind.
¤
¤ Some do, some don't.
¤

Well they either do or don't. ;-) It's typically true, although you can find
some good Linux apps that are competitive with Windows apps.

¤ > In addition, there's simply too many user
¤ > interfaces and variations for this OS so standardization becomes virtually impossible.
¤ >
¤
¤ To be honest, I actually like having the choices. But, this makes no
¤ sense really. It is quite simple for a distro maker to standardize on a
¤ particluar interface, if they so choose.

Nothing wrong with having choices, but the Linux market is over saturated so
there never is a settlement in the Linux community on a distro, which makes it
difficult to establish a standard UI.

Plus the vendors seem to come and go. That doesn't speak highly of the support
and stability of the product.


Paul
~~~~
Microsoft MVP (Visual Basic)

 

[Jim Hubbard]

I just bought 9.2 Pro a couple of weeks ago and have enjoyed playing

Actually... Gentoo doesn't even try to make it easier Gentoo is a
source based distro. It has no graphical installer, like most distro's.
In fact, it doesn't really have what you would call an installer at all.
You basically, boot from a cd, manually partition your disks, download a
source tarball based on the level of install you want, and the run some
scripts to compile stuff. The documentation is pretty good - but you
HAVE to follow it. You can't just skip ahead because you think you know
something.

The thing I like about Gentoo is the fact that it is pretty minimal in
it's base form (it doesn't even install X by default). I can add stuff
as I need it with a simple command from the command line, and it works
out and installs all the dependencies... If you ever worked with a
FreeBSD system and there ports collection, portage is pertty similar.

This level of control seems great for businesses and professional software
people, but the Linspire thing looks more like something that could make a
real difference on end user desktops. And, I think that's the market they
are shooting for. They have just done a huge deal with WalMart and are
selling new Linspire Laptops for $598 and Desktops for $318 to $348.

It is an affordable alternative for the masses to Microsoft.......we just
need more polished apps for the desktop.

Does it still have you run as root by default? That was one of it's bad
points in the past...

Yes it does. Why is that a bad point? Don't you need to be root to install
and tweak stuff? I think they may opt for this to make the OS as easy as
possible for the end user. The end user may have trouble learning the
difference between root and other levels of desktop users. Look at
Windows.....most users run as system administrator. Although this makes the
system less secure, it seems that end users would rather have a less secure
desktop with more power in their hands than a safe, limited use desktop.

That I agree on. It's one of the main reasons I support Linux. It's
not that I dislike MS or that I want to see them go away... I just want
to see them have to compete a little

Amen to that!

Jim Hubbard

 

[Aaron Smith]

While it is true that a lot of code written by task oriented developers does
not conform to proper use of the language (from the point of view of
professional programmers) and may require rewriting, the very fact that the
task oriented developer could write the program in the first place helped
the small business along and showed the usefulness of the task oriented
developers idea.

I have been called upon to enhance and "fix" these applications for most of
my career. But, I have also seen many ingenious methods of solving problems
that have streamlined businesses and departments that were not only adequate
for their purpose, but whose usage saved companies thousands of dollars each
year.

In my experience, the companies that needed rewrites were the companies that
rushed the employees instead of giving them the time to learn and use Visual
Basic correctly. These same companies frequently assign ridiculous
deadlines to professional developers while changing specs constantly and
will have the same problems as before (not every time, but most of the time
I see this as being the case).

So, if a company has to give the employee ample time to learn and
correctly use VB6, why not just hire it out? I realize not all software
shops are reasonably priced, and we may be an exception, but we can
crank out solutions for people faster than what it would take them to
let an employee learn the language and then write the app. Most
employees won't do this stuff on their own time, they want to get paid
to do it. So you have to pay the employee to sit there and read a book,
read a website, post messages, and "play" with the software while they
learn the process or write the app. Seems way cheaper to me if you just
call a software shop that specializes in writing custom software.

In all, I find that REALbasic offers more "bang for your buck" than the
other alternatives that I have looked at so far. If you have any suggested
alternatives, please post them here. I'd love to look at them.

VB.Net, C#

 

[Craig A. Finseth]

We won't know how secure FireFox is until it becomes a much bigger target. I'm
sure we all thought IE was secure at one point, probably when it was overtaking
Netscape. But the fact the FireFox does not support some features targeted as
security risks you could say it's inherently more secure in that sense.

...

Umm, no. I can't think of anyone I know knowledgable about security
that thought that IE was secure when it started. Even quick and dirty
analyses showed a lot of holes.

It just took a few years for the attackers to find them.

Craig

 

[Jim Hubbard]

So, if a company has to give the employee ample time to learn and
correctly use VB6, why not just hire it out? I realize not all software
shops are reasonably priced, and we may be an exception, but we can crank
out solutions for people faster than what it would take them to let an
employee learn the language and then write the app.

In our area, homesourcing costs $90-$120 an hour. Most businesses can
afford to put an internal employee on a project and spend less money even if
the employee takes considerably longer to finish the application. And, the
company then has an in-house developer that can be used for other small
projects or RAD designs that get passed on to professional developers.

Most employees won't do this stuff on their own time, they want to get
paid to do it. So you have to pay the employee to sit there and read a
book, read a website, post messages, and "play" with the software while
they learn the process or write the app. Seems way cheaper to me if you
just call a software shop that specializes in writing custom software.

Lots of small companies like the idea of having an in-house person that can
trouble-shoot apps on-the-spot rather than having to call in someone
whenever there is a problem. Training internal resources allows
this.....and using Visual Basic allows a developer to get up-to-speed a
great deal faster than using a language like C++.

VB.Net, C#

Intentionally breaking backwards compatibility and hiding the fact that .Net
patches are the DLL Hell of .Net are just 2 reasons I am seeking other
solutions to .Net.

Jim Hubbard

 

[Tom Shelton]

This level of control seems great for businesses and professional software
people, but the Linspire thing looks more like something that could make a
real difference on end user desktops. And, I think that's the market they
are shooting for. They have just done a huge deal with WalMart and are
selling new Linspire Laptops for $598 and Desktops for $318 to $348.

It is an affordable alternative for the masses to Microsoft.......we just
need more polished apps for the desktop.

Gentoo is really more targeted for power users...

Yes it does. Why is that a bad point?

Security. Linspire is the only distro that does this.

Don't you need to be root to install
and tweak stuff?
I think they may opt for this to make the OS as easy as
possible for the end user. The end user may have trouble learning the
difference between root and other levels of desktop users. Look at
Windows.....most users run as system administrator. Although this makes the
system less secure, it seems that end users would rather have a less secure
desktop with more power in their hands than a safe, limited use desktop.

It's a bad idea to run as admin on windows as well (though, I'm as
guilty as anyone...). The only reason I can see for people to run on
windows as admin is because there is so much software (games in
particluar) that just won't work as a regular user. This isn't the case
with linux. Sure, it is a slightly different mindset - but not
difficult. I think Linspire is doing a diservice to it's customers by
encouragin this practice. Do they even explain the difference between a
normal user and root?

You might want to check out Xandros. I here it's pretty windows user
friendly.

 

[Tom Shelton]

hiding the fact that .Netpatches are the DLL Hell of .Net

I'm not quit sure I get this one...

 

[Philip Hristov]

Hello,

To all VB6 developers, the things in the world are changing. Visual
Basic is evolving and if you are good developers you will evolve too.
Visual Basic 6 and previous versions are pale resemblances of the real
computer languages, Visual Basic.NET is a other story, real OOP
language, real framework. I like it.

Visual Basic 6, Delphi for Win32 and other Win32 languages are
obsolete. Now you have to think .NET style. Longhorn API will be .NET
based, Win32 is obsolete.

I know why you do not want to move to .NET...but I do not want to
offend some people here, the above stuff was for the serious
developers...

Best,

Philip.

 

[james]

Just a follow up... I installed gambas immediately after posting this.
First impression - definately worth a closer look.

Have fun with it!
james

 

[Herfried K. Wagner [MVP]]

Visual Basic 6, Delphi for Win32 and other Win32 languages are
obsolete. Now you have to think .NET style. Longhorn API will be .NET
based, Win32 is obsolete.

This would make Visual C/C++ (without Managed Extensions) and VFP at least
as obsolete as Classic VB. Sorry, but that's simply plain nonsense.

 

[Jim Hubbard]

I'm not quit sure I get this one...

Microsoft loudly and frequently touted the .Net framework as the end to DLL
Hell. But, this is not necessarily true.

If you subscribe to the free newsletter at www.KBAlertz.com (and I HIGHLY
recommend that you do) you can look up weird behavior in the .Net IDE and
framework. Last time I looked, there were 1,561 issues acknowledged with
the .Net 1.1 framework. (NOTE: I am NOT deriding Microsoft for not having
a perfect product. Nobody has a perfect product. Not Real Software. Not
Hubbard Software. Nobody. )

But, the .Net Patch Hell problem comes in like this.....

You find some buggy behavior in the .Net framework mentioned in the
Knowledge Base (most likely through a KBAlertz email). You can (as most do)
choose to work around it in code, or you can call Microsoft and request a
patch. (I will resist the temptation to rant about not being able to just
download the damned patches from the KB Articles.......for now.)

If you install the patch, it patches the .Net framework on your development
PC. Now, your development PC is out of sync with everyone that does not
have the patch. Not a problem really....just distribute the patch with your
program. Well, not a problem for you anyway.

If you decide to ship the patch with your program, there is a high
probability that other developers' programs on the target machines that used
the code work-around will no longer function correctly. Your patched code
and patch will break their code.

If you choose the common path of coding around the bugs where possible and
some else ships a patch, there goes your code!

So, it seems that we all either don't fix the problems and code around them,
or we all apply the patches. You and I both know that "all of us" will
never do anything. So, here we are.....in Patch Hell. And, how is this any
different than DLL Hell?

It's not.

There is a solution (albeit a pricey one for small ISVs) called Thinstall.
It wraps your entire .Net application (including all .Net framework
dependencies and any applied patches) into a single executable that will run
on any Windows PC from Win98 up. It's actually an incredible solution, but
it's out of the price range of most small companies and not suited for every
application (if you use hooks into the kernel for example).

If you can't afford Thinstall for all of your .Net apps, you'll be risking
Patch Hell with most of the rest of the world.

Jim Hubbard

 

[Jim Hubbard]

It's a bad idea to run as admin on windows as well (though, I'm as
guilty as anyone...). The only reason I can see for people to run on
windows as admin is because there is so much software (games in
particluar) that just won't work as a regular user. This isn't the case
with linux. Sure, it is a slightly different mindset - but not
difficult. I think Linspire is doing a diservice to it's customers by
encouragin this practice. Do they even explain the difference between a
normal user and root?

Not really.....at least not that I noticed during install. They do answer
questions about root on their site at
http://help.linspire.com/cgi-bin/li...cat_lvl1=&p_cat_lvl2=&p_new_search=1&p_page=1 .

You might want to check out Xandros. I here it's pretty windows user
friendly.

I'll try it tonight.

Jim Hubbard

 

[Sean Hederman]

[Snip]

If you install the patch, it patches the .Net framework on your
development PC. Now, your development PC is out of sync with everyone
that does not have the patch. Not a problem really....just distribute the
patch with your program. Well, not a problem for you anyway.

If you decide to ship the patch with your program, there is a high
probability that other developers' programs on the target machines that
used the code work-around will no longer function correctly. Your patched
code and patch will break their code.

I'm not sure I agree. I have a VMWare installed with .NET 1.0 SP0, when I
compile an application with it it runs on .NET 1.0 SP3, .NET 1.1 SP1, and
..NET 2.0 Beta 1. The evidence is pretty good that MS is doing a pretty good
job of avoiding Patch Hell. Since IL uses tokens rather than specific
addresses it's actually quite difficult to break a version.

If you choose the common path of coding around the bugs where possible and
some else ships a patch, there goes your code!

No, your code should be fine unless you write code that *depends* on the
bug, which let's face it is a pretty silly thing to do if you can avoid it.
You could find third-party libraries that do what you want or you could
write your own Interop library to perform the work for you.

So, it seems that we all either don't fix the problems and code around
them, or we all apply the patches. You and I both know that "all of us"
will never do anything. So, here we are.....in Patch Hell. And, how is
this any different than DLL Hell?

It's not.

There is a solution (albeit a pricey one for small ISVs) called Thinstall.
It wraps your entire .Net application (including all .Net framework
dependencies and any applied patches) into a single executable that will
run on any Windows PC from Win98 up. It's actually an incredible
solution, but it's out of the price range of most small companies and not
suited for every application (if you use hooks into the kernel for
example).

Breaking up the .NET Framework DLL's and linking them into your application
wipes out the entirety of the .NET CAS system. Since Thinstall cannot
re-sign the Framework DLLs, they cannot be checked for tampering. Not a
problem many people worry about, I know, but it is one of my main objections
to .NET linking programs.

If you can't afford Thinstall for all of your .Net apps, you'll be risking
Patch Hell with most of the rest of the world.

Since Patch Hell has not yet materialized after 5 years, you'll forgive me
for not getting too worried yet ;D

 

[Philip Hristov]

Hmm. And please tell me how you will develop the new applications for
Longhorn if VB 6 is not osbolete? How you will use the WinFX API? Do
you understand that .NET is the new standard of building applications?
As I see you are DotNet developer...so I do not see the point to
advocate to VB6? Also, I Visual C++ will not be obsolete when it comes
to building drivers and low level stuff - video games too. But it will
be obsolete when you are building desktop applications. So of what I
said is nonsense? Maybe that in NT 6 all Win32 code will be considered
as legacy and will be run in legacy mode.

Regards,

Philip.

 

[Jim Hubbard]

[Snip]

If you install the patch, it patches the .Net framework on your
development PC. Now, your development PC is out of sync with everyone
that does not have the patch. Not a problem really....just distribute
the patch with your program. Well, not a problem for you anyway.

If you decide to ship the patch with your program, there is a high
probability that other developers' programs on the target machines that
used the code work-around will no longer function correctly. Your
patched code and patch will break their code.

I'm not sure I agree. I have a VMWare installed with .NET 1.0 SP0, when I
compile an application with it it runs on .NET 1.0 SP3, .NET 1.1 SP1, and
.NET 2.0 Beta 1. The evidence is pretty good that MS is doing a pretty
good job of avoiding Patch Hell. Since IL uses tokens rather than specific
addresses it's actually quite difficult to break a version.

You are comparing different versions (which, BTW, are not always backwards
compatible) NOT the patches to which I refer.

Have you called Microsoft for a patch? Have you installed one? Have you
used the affected portions of the .Net framework and distributed this code
to non-patched systems?

No, your code should be fine unless you write code that *depends* on the
bug, which let's face it is a pretty silly thing to do if you can avoid
it.

If you are writing a "work-around" it will naturally depend on the bug.

You could find third-party libraries that do what you want or you could
write your own Interop library to perform the work for you.

Heck, you could write your own IDE. But this still does not address the
Patch Hell problem.

Breaking up the .NET Framework DLL's and linking them into your
application wipes out the entirety of the .NET CAS system.

For all of the ballyhoo surrounding the security built into the .Net
framework, I have yet to install an application or work on an enterprise
project that uses it. IMHO, lots of hype about nothing.

Since Thinstall cannot re-sign the Framework DLLs, they cannot be checked
for tampering. Not a problem many people worry about, I know, but it is one
of my main objections to .NET linking programs.

I think it really comes down to only installing applications from reputable
vendors. Anything else will not save you.

Since Patch Hell has not yet materialized after 5 years, you'll forgive me
for not getting too worried yet ;D

You are forgiven.

In fact, I never saw DLL Hell. But, there has always been a simple way to
avoid DLL Hell, and responsible shops used it for widely distributed
applications. Simply put your executable's DLLs in the exe directory.
Windows will use those DLLs before using shared DLLs.

The point I am making, is that the potential for "X Hell" has not dissapated
with the advent of .Net. .

Net brings many good things to the table. Protection from "X Hell" is not
one of them.

Jim Hubbard

 

[Herfried K. Wagner [MVP]]

Hmm. And please tell me how you will develop the new applications for
Longhorn if VB 6 is not osbolete? How you will use the WinFX API? Do
you understand that .NET is the new standard of building applications?

At least parts of this API will be made available to COM and/or Win32, as
far as I know. This has mainly practical and economical reasons: A rewrite
of existing applications in managed code is not accomplishable within a few
years. Take a look at the technologies supported by Windows XP today: Even
DDE is still supported, because applications rely on it, and DOS/Win16
applications can still be executed.

It's utopistic to think that Microsoft will pull the switch for Win32/COM
applications within the next ~10 years. By doing that Windows would loose
millions of customers who wrote software for Windows. It's compatibility
and interoperability which tie customers to a certain platform. If
compatibility is not one of the main goals any more and customers' assets
are rendered disposable, customers will consider turning to a more
compatible platform which preserves their assets. "Everything in .NET" is a
nice pipe dream.

As I see you are DotNet developer...so I do not see the point to
advocate to VB6?

I don't see the whole issue from a technical standpoint only. In fact I see
the economical perspective too which is the driving force for choosing a
platform over another. I don't advocate VB6 because I think it's better
than .NET. Sure, there are some things which were better in VB6, but there
are lots of things I like more in .NET. The reason why I don't think that
VB6 should be disposed are the billions of lines of code written in this
language which won't be converted within short time (some years). COBOL,
which is some decades old now, is still used for banking and financial
transactions. Programming languages are not coupled to a platform or
technology and in general their lifetime is much longer than the lifetime of
a technology. There are no technical reasons for not adding .NET support to
the Classic VB programming language like it was done with other programming
languages, for example, C++, EIFFEL, COBOL, FORTRAN, ...

Also, I Visual C++ will not be obsolete when it comes
to building drivers and low level stuff - video games too. But it will
be obsolete when you are building desktop applications.

Do you really think that Office, for example, will be fully managed when
Longhorn is released? Complete rewrites are hardly ever done because a new
technology is appearing. The costs of a rewrite are much higher than the
benefit gained by it.

Maybe that in NT 6 all Win32 code will be considered as legacy
and will be run in legacy mode.

I doubt that customers will accept that, because it will dispose assets
similar to how assets are/were disposed by discontinuing VB6. Microsoft has
lost customers and is about to loose customers by disposing VB6. Microsoft
may loose the Linux vs. Windows war by breaking compatibility. I like
Microsoft products and I use them, but I don't agree with all decisions made
by Microsoft in the last decade. I am afraid that Microsoft will go the
same way IBM and other companies did if it doesn't give their customers'
voice more importance.