Good for PB bad for VB !!

[Paul Dwyer]

PB users will be interested to hear that the next version of VB basically
will break every piece of VB code written to date. I'm glad I found PB when I did.

For those interested, this is a link to VB World where I hang around less and less
and their article on VB.NET. I can almost hear the screams and Outcrys from here in Tokyo
http://www.vb-world.net/misc/vbdotnet/index2.html

Very amusing

------------------

Paul Dwyer
Network Engineer
Aussie in Tokyo
(Paul282 at VB-World)

[Scott Turchin]

I like this part, take the BASIC out of Visual *BASIC*:

Miscellaneous – Initialise variable values on the same line as you declare them.
No more fixed-length strings. No fixed-size arrays when using UDTs.
No GoSub/Return.
No VarPtr, ObjPtr, StrPtr and such.
No more LSet.
No more 'As Any' declarations.
Setting a Timer to 0 will no longer disable it.
Windows forms will not support DDE.
The Name property will no longer by available at runtime.
True will no longer equal 1, it will equal –1.
In VB.Net, adding a Null to another string will generate a mismatch error, whereas currently the whole concatenation operation would return a Null.
And, Or, Not and Xor will no longer be Boolean operators – they will be replaced with BitAnd, BitOr, BitNot and BitXor.

------------------
Scott
mailto:[email protected][email protected]</A>

[Peter P Stephensen]

You are talking about VB.NET not VB7.
VB7 is the next version of VB.

VB7 will have inheritance, overloading and polymorphism as well
as structured error handling (try - catch). Cool!

Regards
Peter



------------------

[Jim Karabatsos]

Peter,

The next version of Visual Studio will NOT be VS7, it will in fact
be Visual Studio .NET. It follows that the next version of
VB will be VB.NET.

The features that you mentioned (inheritance, overloading and
structured exception handling) are all VB.NET features.

Polymorphism already exists in VB through interfaces; VB.NET will
have additional support for polymorphism through inheritance.

VB.NET will indeed break a LOT of code. This is distressing a
lot of VB developers. I am a bit undecided on where I stand on
this one. While I am *big* on backwards compatibility (and have
lambasted MS in the past in my articles in AVDF), I recognise that
the kinds of changes that MS is trying to introduce now really
require that the language be reigned in a bit. They are taking a
bit (well, actually a lot) of license to change the way things
work so that the language can be evolved to fit the requirements
of modern development practices. Remember, VB is not the right
choice if you want small, tight executables (we all now that
PowerBasic is the best tool for that). VB plays in a very
different space, and that space (corporate development) is not
very tolerant of upgrades that break old code. MS wants to have all
languages in VS compile to a common runtime, and that means that
VB is changing so that data types and so on match the other
languages.

In our team, we are basically going to leave old code under VB6.
There is no compelling reason to convert the code -- VB6 will work
as well next year as it does now, and the use of COM as the basis
for all new Windows APIs means that it will not age any time
soon.

However, we are re-evaluating whether we should buy in to the
.NET mindset at this stage. Frankly, I'd rather teach my people
Delphi than a new flavour of VB, and I would get the ability to
target Linux with little or no additional effort. I also get a
version 6 (or 7 soon) product that is mature and stable. Contrast
that to VB.NET which, make no mistake, is a V1 product.

PB? Well, I don't want to start the old COM argument yet again,
but IN MY ENVIRONMENT PB will only ever be a bit player until it
gets built-in COM support. Obviously, this is not true for
everybody but it is the case in corporate development. For ME,
COM support rates above all other requirements, including a
visual designer and Linux support. But as I said, that argument
has been had and I don't want it to erupt again.

I think the real problem with VB.NET is not so much that old code
is broken, it is that old gurus are shot down in flames. VB.NET is
SO different that the people who are seen as the experts suddenly
find themselves at the same level as everyone else -- that is,
pretty much clueless. That is why so many people are grumbling,
and that might be a real problem for MS -- it is not every day
that the very people who are seen as advocates for a company
turn around and complain (very vocally) about a new version.

-- Jim

[Mark Newman]

I think the real problem with VB.NET is not so much that old code
is broken, it is that old gurus are shot down in flames. VB.NET is
SO different that the people who are seen as the experts suddenly
find themselves at the same level as everyone else -- that is,
pretty much clueless. That is why so many people are grumbling,
and that might be a real problem for MS -- it is not every day
that the very people who are seen as advocates for a company
turn around and complain (very vocally) about a new version.

I've heard this argument and I don't agree. The VB gurus are doing
their job, pointing out the incredible amount of change in VB.NET
and that most of these changes are purely gratuitous. Besides,
the qualities and skills that made the gurus once will make them
gurus again in .NET: Study, research, practical application and
pushing the language to the limits.

To me, C# is the star of the .NET show and VB was just brought
along for the ride. Once again, VB is viewed as a second-place
language by MS.

------------------
Mark Newman

[This message has been edited by Mark Newman (edited January 31, 2001).]

[Rich Brockway]

1. Miscellaneous – Initialise variable values on the same line as you declare them.
2. No more 'As Any' declarations.
3.True will no longer equal 1, it will equal –1.
4. And, Or, Not and Xor will no longer be Boolean operators – they will be replaced with BitAnd, BitOr, BitNot and BitXor.

These changes are positive. Item one makes the code cleaner. Item two allows for strict type checking. Item three, goes without saying. Using hard coded values instead of the supplied constants is simply bad practice or a failure to understand basic programming. Item four corrects an annoying problem. VB could never perform "short circuit" evaluation. This fixes it.

You are talking about VB.NET not VB7.
VB7 is the next version of VB.

No, VB, as you know it, is dead. There will be only VB.NET.

Rich

------------------




[This message has been edited by Rich Brockway (edited January 31, 2001).]

[Rich Brockway]

Jim,

I am interested in hearing why you believe "COM as the basis for all new Windows APIs means that it will not age any time soon." According to MS, COM is supported in .NET. .NET puts a wrapper around COM components. However, .NET has its own component architecture, assemblies. To me, that indicates COM is the legacy technology and assemblies are the "new" technology. Also, according what I have read at Borland's website, they are not going to support ActiveX in Delphi/C++ Builder 6.0. Your thoughts?

Rich


------------------




[This message has been edited by Rich Brockway (edited January 31, 2001).]

[Peter P Stephensen]

Maybe this link is outdated?
http://www.vb-world.net/articles/vb7/

Regards
Peter

------------------


[This message has been edited by Peter P Stephensen (edited January 31, 2001).]

[Paul Dwyer]

The part I really like is the lack of support for GOTO!

Up till now major keywords had a version to fade out but in VB6
it was THE ONLY way to do error handling so it's in all professional
code, now it doesn't work!!! MS Genius

Arrays HAVE TO start at zero! how much code with that break?
Dim MyArray(10) as long

used to be 1 to 10, now it's 0 to 9, so if code had IF MyArray(10) = x Then
...then you'll get an overflow!

Default is ByVal not ByRef! funny, that was a big feature for VB4 (I think it was 4)
They are trying to keep in line with C# having no pointers.

No Varptr!!! Well, there goes half the API

Not Lset (you'd think they'd introduce Unions to plug the gap)

Xor to BitXor! Well, my DES code won't compile then

I say Goodbye to the future of VB, The new OO is ok, the free threads are long overdue, too much web crap though.
Web scripting is not that hard! why move the language so high? it seems odd.

I think if VS7 is any good it'll be because of C# which looks okay from what I read.
VB6 has enough for me and new windows features that need compiler support I will look to PB for support as they listen to their customer base rather than try to "Shape the future"

Pity, I though VB should go lower lever but MS thought Higher was better.

So here I am at PB, probably give them a big boost in sales


------------------

Paul Dwyer
Network Engineer
Aussie in Tokyo
(Paul282 at VB-World)

[This message has been edited by Paul Dwyer (edited January 31, 2001).]

[Michael Meeks]

I wouldn't count VB3 through VB6 out just yet!

I have many clients who still use VB3 through VB6, AND, use it
with WindowsMe! without any problems! To say that VB7 or
VB.NET, will shake the world is leaping before you look! There
will be vendors and programmers creating conversion kits etc...
Micro$oft will know if they were right, because it's all in the
sales, and pricing...which leaves many people out of the loop!

This is where PB6 stands on it own. Price, Program Size and
Speed! Look at the changes of PowerBasic in the past 6 years!
I remember hearing a few screams back then....

What will PB7 look like or be called ? And it's amazing how
far many of us have come with PB. For example, Ezgui, WinLift
and many of you others who have contributed to this site with
your programs and modules! We won't be standing still when
VB7 is released, nor will be intimated!

As long as I got PB on my computer, I'm a happy camper!


Mike

[John Kovacich]

Microsoft NEVER made their products backwards compatible.

Version 3 to 4 required considerable conversion (thanks go to Crescent
for providing free conversion tools).

Version 4 to 5 broke some of my programs sufficiently that they remained on
VB4.

Today, I still use VB4 and VB5 as add-ons to PB6.

I can't endorse any of IBM's interfaces, but they know how to maintain compatibility
when releasing new versions of their compilers.

Microsoft has forged the way everything works for some time now, but I
don't think that trend will continue. None of my clients have migrated
from NT4 to 2000. The industry is coming to a point where they know what
they want, and blindly following Microsofts lead is not what they want.



------------------
Thanks,

John Kovacich

[Scott Turchin]

a bit TOO philisophical for me but I do have to comment on this one:

Microsoft has forged the way everything works for some time now, but I
don't think that trend will continue. None of my clients have migrated
from NT4 to 2000. The industry is coming to a point where they know what
they want, and blindly following Microsofts lead is not what they want.

That's YOUR loss if you don't upgrade to Win2k, it is by far SO SO SO Much more stable than NT could ever DREAM to be..

And I have a friend in charge of testing whistler, he said it is 10x better than Win2k is stability....


Microsoft has done a LOT of good, lets not start bashing them.
Their VB series, well there it goes if they start pulling stuff out.

And so if it is VB NET, so be it, where's VB7 then? In VB NET from what I've read and if it's NOT then it's still missing a lot of features that make BASIC BASIC.

Why not convert to C++ if they are going to go to all the trouble?


Scott


------------------
Scott
mailto:[email protected][email protected]</A>

[This message has been edited by Scott Turchin (edited January 31, 2001).]

[Jim Karabatsos]

Rich

re COM as the basis for all new APIs

I have yet to see any mention of new system extensions that do not have COM support. Sure, it is possible that MS will change that in the future, but somehow I doubt it, especially for the next couple of years. I do not profess to be an expert it .NET, but I don't think assemblies impact the way you would expose
functionality from an OS point of view at all.

I don't know what you read on Borland's web site, but I have not heard anything about dropping ActiveX support in Delphi 6. Of course, there will be no ActiveX support in Kylix -- what would be the point of ActiveX support on Linux?

I am not setting myself up as an expert here, and I stand to be corrected if necessary. However, I think that a lot of VB 6 programmers are going to need to have a real think about where they are going to go next, and I don't think it is a foregone conclusion that they will indeed go to .NET

That's not to say that I am predicting the demise of VB -- It has just too much presence on its home turg for that.

-- Jim


------------------

[Paul Dwyer]

That's YOUR loss if you don't upgrade to Win2k, it is by far SO SO SO Much more stable than NT could ever DREAM to be..
- I aggree with you there, I work in SysAdmin for AT&T Japan and the attitude is, Go Win2k, ban WindowsME and remove 98 where possible. 98 is a nightmare to support it's so buggie, no chance that we'll give ME a go.

Why not convert to C++ if they are going to go to all the trouble?
-They are converting C++, It's called C# (C Sharp) and it's their new flagship compiler in VS



------------------

Paul Dwyer
Network Engineer
Aussie in Tokyo
(Paul282 at VB-World)

[Michael Meeks]

We'll said, Scott...

Also, Paul Dwyer, about WindowsMe! You are right about Win98, but
WindowsMe has taken the Nightmare away from Windows, at least for
me! WindowsMe has been very stable, compared to all the Versions!

C++ or VB++, but in my view, they are trying to take the Basic out
of Visual Basic.

Wish PB (powers to be) would give us some more clues on the new
and improved PowerBasic 7 or (the new name), but, that would be
asking too much!

With that said, I bet it's more in line with our thinking, than
the new VB.Net will ever be!

Thanks
Mike

[Bern Ertl]

Having been literally on a desert island for the better part of the last three years, I'm a little bit behind the curve and trying to catch up. I have a general understanding that M$ is trying to gravitate everything to a Webcentric future which is of no use for me professionally, but I'm not savvy to the finer details of the differences between Win98, Win2K & WinMe. I was of the understanding that WinMe somehow incorporated a significant architectural difference (no Win32?) from the previous versions of Windows. Is this true?

Does anyone have any idea if M$ has plans to continue revising their Windows operating systems to the point where PB programs will no longer work on them?



------------------
Bernard Ertl

[mark smit]

I have some worries here for PB. If you look far enough down the line here folks it almost seems like Microsoft is trying to gain a monopoly in the compiler business. Pretty soon the only compiler that can keep up with new versions of windows will be Microsofts. I mean PBDLL doesn't even support COM/ActiveX and that seems to be going the way of the dodo. Looks like we're all going to turn into Microsmurfs!

------------------
Cheers

[Mike Comeau]

Lets not forget that, ALL PB programmers are M$ supporters.

We code to a target audiance of M$ systems. We just do it our way.

Please see my OOP topic, we need to develop code rippers to convert source code that will ensure our future in PB.

All c c++ code can be converted into compatable PB functions, so we just include the ability to rip down COM and NET resources.

Meanwhile PB will produce compatible compilers for the newer OS's.

Mike...

------------------

[Michael Mattias]

IIRC, Microsoft's "design and run-time relationship" between its compiler products and its operating system products (and business software, and browser software) was recently addressed by the U. S. Dept of Justice.

Should the court-ordered breakup actually occur and be enforced, I'm guessing the WinAPI will be a bit more open and developers will not be locked in to a Microsoft compiler, nor subject to the a new release of Internet Explorer "breaking" existing code because the common control API was changed.

If one stays outside the realm of ActiveX, OCX, etc and relies on the "core" API functions in USER, GDI and KERNEL, the Win32 API is actually a very nice, powerful tool with a lot of stability.

My $0.02, anyway.

MCM

[Guest]

What do any of you touting Windows 2000 as
being more stable than NT base your statements on - microsoft fud?
Windows NT 3.51 is the most stable microsoft windows OS they have ever cranked out.
Win2000 is NT4 with some new features added. It is not re-written from scratch as
you-know-who has so often stated (you know when Bill Gates is lying - his lips are moving!).

We have had NT4 running for over 2 1/2 years on a cluster, running sql server and it has
not flickered. Our experiences with Win2000 is that it is slower
unless you run apps written for it. They did a good job of scrambling the interface to
basically achieve the same functionality - stupid!
We don't subscribe to paying 10's of thousands of
dollars for licenses and doing the same thing again a couple years later when microsoft
says they have the Oracle "equalizer" product - if it runs on Win2000.

As for VB.Net - i don't like the idea based on my very limited knowledge. As for converting
vb6 and earlier apps, my approach would be to rewrite.
I don't think we would use VB.not. We would use CPound (C#).
It will have to earn the "sharp" designator.



------------------
Ron