.NOT My Views

[Karl E. Peterson]

But Karl does have a point - the "Integer" in VBc is not an "int" it
is a 16-bit signed thingy. Had they allowed the Integer to 'float'
with VB4, they would have had to add a new data type. MS chose not
to, so they stuck with keeping an Integer a 'short'. With a 64-bit
'intergral data type' on the horizon, they would have had to do
something, anyway, had VBc stayed in existence.

To be fair to Karl, and to maintain true VBness, the "Integer" should
probably have remained 16-bit. A Long 32-bit, and they could have
come up with something new for a 64-bit thingy. Perhaps a 'Fred'. <g>

YES! Someone understands!!! :-D

 

[mayayana]

Maybe you could explain your point
for those of us who don't have a
broadband connection. Is there a point?

 

[Ralph]

Ralph,

I was refering to what you said:


How can you not be dependent on the size of the thing if you are reading
from a binary file? Even if you had been taught otherwise.

Gary

You are absolutely correct. I was mentally wandering and expressed myself
poorly. I was only trying to say - that the whole mess could have be
avoided if VB had adopted a "Int" - a creature that was the 'optimal'
integer type when size didn't matter and had specifically sized things when
it did, eg your binary file.

I don't think many VBers ever fully appreciated what that meant - they only
had very specifically size data types.

My apologies.

-ralph

 

[Ralph]

I am _well_ aware of that historical difference between languages. MSBasic had a
fixed-size Integer, unlike some others.

"If you need a different type of data, create a new datatype!"

We suggested they restore the original definition of the keywords, but they balked.
The ideal solution would've been to introduce new names for fixed sizes, and
something like SysInt for platform sized variables. That simple concession would've
done a _lot_ to show that they understood the issue of gratuitous incompatabilities.

Exactly. As you agree with me, you are obviously a man of fine distinctions
and high intelligence. <g>

"SysInt" would have been a great name. Simple and descriptive.

Ah what might have been.

Have you given any thought to what the 'managed' extensions to VBc would
have looked like? I was thinking that a new 'managed class-type' module
would have worked. With a separate Windows Forms designer that the managed
'classes' would have owned and served as the backend for.

Both would run in the same app. Simple property attributes similar to
'Instancing' would have allowed one or the other section to be the
'interopt' of the other. But must admit I start to confuse myself after
awhile.

Ah, again, what might have been. <g>

-ralph

 

[J French]

On Wed, 12 Oct 2005 09:26:52 -0500, "Ralph"

Although what Ralph said about 'Integers' is true in theory, in
practice C programmers know all about the size of their 'Integers'

True portability is a myth, unless one rigorously sticks to ANSI
standards.

Redefining the Integer was no big deal for C programmers as unlike VB
thay have a whole slew of variables

Just to nitpick - it isn't a matter of theory, it is a practical adaptation
to hardware variations. And the "int" is never 'redefined', but a slew of
specific types with defined sizes are created and are availble. [int is an
ANSI standard]

I wonder whether C will move to a 64 bit definition of Integer
- somehow I rather suspect not

But Karl does have a point - the "Integer" in VBc is not an "int" it is a
16-bit signed thingy. Had they allowed the Integer to 'float' with VB4, they
would have had to add a new data type. MS chose not to, so they stuck with
keeping an Integer a 'short'. With a 64-bit 'intergral data type' on the
horizon, they would have had to do something, anyway, had VBc stayed in
existence.

Well the VB Long was ... long established so they did not really need
to change the VB definition of Integer

To be fair to Karl, and to maintain true VBness, the "Integer" should
probably have remained 16-bit. A Long 32-bit, and they could have come up
with something new for a 64-bit thingy. Perhaps a 'Fred'. <g>

Totally agreed - a Long64 would be a useful addition to VB

Part of the reason why B# moved to a 32 bit Integer was probably
because the designer Anders Hejlsberg had already forced a similar
conversion in the move from Delphi1 to Delphi2

In that case it was no big deal as the benefits from upgrading from 16
to 32 bit were irresistable

Another reason was, I guess to keep things in line with C#

Personally I see no great problem in a variables type name differing
between languages ...

 

[Ralph]

On Wed, 12 Oct 2005 09:26:52 -0500, "Ralph"

Although what Ralph said about 'Integers' is true in theory, in
practice C programmers know all about the size of their 'Integers'

True portability is a myth, unless one rigorously sticks to ANSI
standards.

Redefining the Integer was no big deal for C programmers as unlike VB
thay have a whole slew of variables

Just to nitpick - it isn't a matter of theory, it is a practical adaptation
to hardware variations. And the "int" is never 'redefined', but a slew of
specific types with defined sizes are created and are availble. [int is an
ANSI standard]

I wonder whether C will move to a 64 bit definition of Integer
- somehow I rather suspect not

But Karl does have a point - the "Integer" in VBc is not an "int" it is a
16-bit signed thingy. Had they allowed the Integer to 'float' with VB4, they
would have had to add a new data type. MS chose not to, so they stuck with
keeping an Integer a 'short'. With a 64-bit 'intergral data type' on the
horizon, they would have had to do something, anyway, had VBc stayed in
existence.

Well the VB Long was ... long established so they did not really need
to change the VB definition of Integer

To be fair to Karl, and to maintain true VBness, the "Integer" should
probably have remained 16-bit. A Long 32-bit, and they could have come up
with something new for a 64-bit thingy. Perhaps a 'Fred'. <g>

Totally agreed - a Long64 would be a useful addition to VB

Part of the reason why B# moved to a 32 bit Integer was probably
because the designer Anders Hejlsberg had already forced a similar
conversion in the move from Delphi1 to Delphi2

In that case it was no big deal as the benefits from upgrading from 16
to 32 bit were irresistable

Another reason was, I guess to keep things in line with C#

Personally I see no great problem in a variables type name differing
between languages ...

Delphi had no impact. Hejlsberg was only dealing with the same issue that
face all languages once the 'native interger type' changes sizes. Of course
the difference is, he dealt with it.

With any C/C++ compiler there is a header (stddef.h usually) that contains
something to the effect...
#ifdef _WIN64
typedef int _int64
....
So yes C will change the value of int, depending on what it is compiled
against, it always has and always will. It is built into the language so to
speak.

While if VB would have addressed the issue long ago with a "SysInt" (a la
Karl) there would be not have been a problem.

So you prefer "Long64" over a "Fred"?

-ralph

 

[J French]

On Thu, 13 Oct 2005 07:15:58 -0500, "Ralph"

So you prefer "Long64" over a "Fred"?

Int64 is pragmatic, QWORD is also nice
- they would have been useful in an extended VB6

Also the tit-heads could have released a mega DLL that was self
checking and acted as a 'slice'

My view is that MS have been taken over by a bunch of tossers

 

[Ralph]

On Thu, 13 Oct 2005 07:15:58 -0500, "Ralph"



Int64 is pragmatic, QWORD is also nice
- they would have been useful in an extended VB6

Also the tit-heads could have released a mega DLL that was self
checking and acted as a 'slice'

My view is that MS have been taken over by a bunch of tossers

"Tossers" - meaning the Irish music group loaded with attitude and alcohol?
Or as "a stupid or unpleasant person"?

Both seem to fit. However, based on episodes from my own past, I can some
what forgive the former.

-ralph

 

[Abubakar]

Don is an architect on Indigo team and makes sure the next gen performance
critical apps built on windows using indigo live upto the promise of indigo
performance. Jeff is a miracle guy who, among other things, is focusing on
testing indigo and writes applications inside microsoft company that takes
indigo to its limits. Everything here revolves around .net. Don is a
microsoft employee and Jeff is a consultant at microsoft for a long time.
They dont merely teach and they know what performance is.

-Ab.
http://joehacker.blogspot.com

 

[Gary Nelson]

Ab,

They dont merely teach and they know what performance is.

But they don't have thousands of clients and hundreds of thousands of lines
of legacy code to maintain.

Gary

 

[Gary Nelson]

Personally I see no great problem in a variables type name differing
between languages ...

The problem comes up when you are trying to maintain similar code in both,
in particular if you read and write data in a binary file. It is very easy
to make a mess of things.

Gary

 

[Abubakar]

That reminds me of another reason I don't move to

.Net: The IDE won't run on Win98SE and I don't want
to switch to Windows Xtra Problems. I find it ironic

wow, you did not shift to windows xp, so how can i expect you'll move to
..net. Fine with me.

to switch to Windows Xtra Problems. I find it ironic
that most Java and .Net developer software runs only
in very limited environments. Yet both systems aim to
be thoroughly cross-platform.

they need power to run!

is that there is a difference. He thinks .Net is fine for
web-based software. The point he's making is that
it's basically Microsoft's version of Java and doesn't
belong on the desktop. It may be true that new hardware

"Microsoft's version of Java" have heard this a lot. If this is what he
says, I'm sure its not worth reading.

that as progress. Needing 4 GB of RAM to write letters
and send email is a pitiful state of affairs.

totally untrue, figures are not correct!

I think you guys are perfect with vb6. Carry on. And good luck.

Cheers,

Ab.
http://joehacker.blogspot.com
there are better things to discuss like :
http://www.msnbc.msn.com/id/7328143/

 

[mayayana]

"Microsoft's version of Java" have heard this a lot. If this is what he
says, I'm sure its not worth reading.

That reminds me of a famous story about
Galileo. As the story goes, Galileo had found
craters on the moon by looking through his
telescope. At the time, the Catholic church had
a monopoly on knowledge and truth in Europe,
and they didn't like Galileo coming up with new
discoveries and ideas because that could lead to
the development of a whole new truth franchise in
Europe, which could threaten Catholic power.
So a Catholic bishop was
sent to see what Galileo was up to. Galileo
offered to demonstrate his claim about craters on
the moon by letting the bishop see for himself
through the telescope. The bishop responded
something like, "I don't need to look through
your telescope because I already know that
God would not make the moon as anything other
than a perfect sphere."

 

[Ralph]

That reminds me of a famous story about
Galileo. As the story goes, Galileo had found
craters on the moon by looking through his
telescope. At the time, the Catholic church had
a monopoly on knowledge and truth in Europe,
and they didn't like Galileo coming up with new
discoveries and ideas because that could lead to
the development of a whole new truth franchise in
Europe, which could threaten Catholic power.
So a Catholic bishop was
sent to see what Galileo was up to. Galileo
offered to demonstrate his claim about craters on
the moon by letting the bishop see for himself
through the telescope. The bishop responded
something like, "I don't need to look through
your telescope because I already know that
God would not make the moon as anything other
than a perfect sphere."

So your saying that .Net is a perfect sphere, you need a telescope to see
its features, or we should all just join the Catholic church?

-ralph

 

[Ralph]

The problem comes up when you are trying to maintain similar code in
both, in particular if you read and write data in a binary file. It is very
easy to make a mess of things.

Gary

I have to agree with you there.

If the French, for example, didn't foolishly create a different word for
everything, communication would be a lot easier with them.

As it is now one has to do a lot of hand waving and speak in a loud voice.
It is very easy to make a mess things, if one isn't careful.

-ralph

 

[Karl E. Peterson]

Exactly. As you agree with me, you are obviously a man of fine
distinctions and high intelligence. <g>

"SysInt" would have been a great name. Simple and descriptive.

Ah what might have been.

Indeed. Could've been _so_ clean. But redefined datatypes and recycled keywords put
the kibosh on ever being able to trust the designers again. Failure to see what's
wrong with that means the languages they produce aren't suitable for investment,
other than truly (not their interpretation) "disposable" code.

Have you given any thought to what the 'managed' extensions to VBc
would have looked like?

Not really. Without *some* indication Microsoft was willing to see the fundamental
problem for what it was, it seemed like a waste of mental bandwidth.

Ah, again, what might have been. <g>

Yes...

 

[DanS]

OK, so I've read through ALL of this thread.

My 2¢......

Why does anyone argue about VBc vs VB.NET ? Johnny .Net is always going
to think that it is better and tell you to come out of the dark ages, and
Mr. VBc has loads of VB6 apps to maintain.

EVERYTHING has it's good points and it's bad points.

Truthfully, a programming language is nothing more than a tool to get to
the end result. As a programmer, or a motorhead, or a communications
tech, or a construction worker, you use the tool that gives the best
result for the task at hand.

Only those that have an equal amount of knowledge and experience in VBc
and VB.NET can have a valid opinion of which is 'better', but to which I
will always ask, 'better to do exactly what ?'. A big wrench will drive
in a nail, but a hammer was made to do that.

Noone that strictly does .NET, and only has a small knowledge (or none at
all) of VBc has the right to say .NET is way better than VBc. (Which is
what I perceive when these threads pop up will people bashing VBc, as
indicated by 'well how do you do this in VB6?')

I myself have not gone to .NET yet because programming is only a hobby
that I am fortunate enough to sometimes get paid to do for a few local
small companies, either of which if I asked what language they want it
in, I know the answer would be 'whatever gets the job done'.

In closing, and this is not a .NET bash, just a personal peeve, if RR was
correct in his 'one-line text file read', I CAN'T BELIEVE M$ MADE A
NAMESPACE CALLED 'MY.(anything here)'. That is _totally_ gay, which I
thought they realized when they decided to drop the 'My' moniker for the
user folders in Vista.

(And whoever said 'Basic is mainly a M$ thing' is either naive, never
paid attention in CS classes, or just blinded by the .net.)

Regards,

DanS

 

[mayayana]

So your saying that .Net is a perfect sphere, you need a telescope to see
its features, or we should all just join the Catholic church?

If you like.

 

[Ken Halter]

In closing, and this is not a .NET bash, just a personal peeve, if RR was
correct in his 'one-line text file read', I CAN'T BELIEVE M$ MADE A
NAMESPACE CALLED 'MY.(anything here)'. That is _totally_ gay, which I
thought they realized when they decided to drop the 'My' moniker for the
user folders in Vista.

LOL! I agree.... seems like a bunch of selfish cry-babies adding "My" to
everything eh? <g> My pictures, My documents, etc. The one that really ticks
me off is "My Network Places". I'd like to rip that one off of every PC
known to man. Free clue to who ever's in charge of the next OS.... My
Network Places should either be gone for good, show only valid connections
or have the ability to detect invalid connections without hanging for 10
minutes.

I also agree with the folder names. The "My" prefix ruins a users ability to
open a folder in windows explorer, click in the files pane and have files
selected as they type. It also make sorted names less obvious.

(And whoever said 'Basic is mainly a M$ thing' is either naive, never
paid attention in CS classes, or just blinded by the .net.)

Regards,

DanS

Finally, I couldn't care less if a specific "chore" takes 1 line, 10 lines
or 1000 lines, as long as it works and I don't have to write it again every
6 months.

 

[Dragon]

Excuse me?

We suggested they restore the original definition of the keywords, but they balked.
The ideal solution would've been to introduce new names for fixed sizes, and
something like SysInt for platform sized variables. That simple

concession would've

AFAIK, IntPtr and UIntPtr are exactly that, i.e. platform-sized
integers! Aren't them?

Roman