Rumors for beta 2

[Gregor R. Peisker]

Hi Bill,

> What did they have to go and do that for? I suppose they could just
> insert parenthesis.

Who really knows? From what I know, it's uncommon for bitwise operator to
precede comparison operators. But this way, you can do:

If value BitAnd flag = flag Then

Parens for the Upgrade Wiz are fine, but if developers don't know the rules
have changed, someway might just go in and say, "Hey, what's that for, let's
remove it!".

Regards,
Gregor

[Dave Haskell]

Jonathan,

> You problem seems to be that you don't understand strongly typed operators.
> No problem, it is easy to learn.
<snip>

Uh, they were talking about what the upgrade wizard does. Does it do what you
stated?

[Jonathan Allen]

> Uh, they were talking about what the upgrade wizard does. Does it do what
you
> stated?

It might. Don't forget, the upgrade tool is still a work in progress. Just
because it doesn't work right now doesn't mean it never will. Does not the
term Beta mean anything to you?

--
Jonathan Allen


"Dave Haskell" <NOhaskellsSPAM@pacbell.net> wrote in message
news:3ac37f6d$1@news.devx.com...
> Jonathan,
>
> > You problem seems to be that you don't understand strongly typed
operators.
> > No problem, it is easy to learn.
> <snip>
>
> Uh, they were talking about what the upgrade wizard does. Does it do what
you
> stated?
>
>

[Joe \Nuke Me Xemu\ Foster]

"Jonathan Allen" <greywolf@cts.com> wrote in message <news:3ac384c2@news.devx.com>...

> > Oh, and about Boolean logic -- let's see what Beta 2 brings.
>
> I don't see anything wrong with the Boolean logic. You just have to be aware
> of what data type you are using. Like I told Dan, don't use a BitAnd on a
> Boolean or a And on an Integer and you will be fine.

Could the And operator be overloaded? If the compiler can tell that both
operands to And, Or, Xor, Eqn, etc., are Boolean, perhaps it could put
in short-circuiting automagically if there's a nearby Option Lazy.

At this rate, we'll soon have as many tweaks as MSVC++.

--
Joe Foster <mailto:jfoster@ricochet.net> Greed = God? <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!

[Dave Haskell]

Jonathan,

> > Uh, they were talking about what the upgrade wizard does. Does it do what
> you
> > stated?
>
> It might. Don't forget, the upgrade tool is still a work in progress. Just
> because it doesn't work right now doesn't mean it never will.

Yep, we'll see the final version eventually.

> Does not the
> term Beta mean anything to you?

***? You were trying to "teach" Dan, and I was simply pointing out that what
you were "teaching" was not on the topic that lead Dan to write what he did.

[Jonathan Allen]

> Could the And operator be overloaded?

Only if we dropped short-circuiting. Otherwise its behavior will vary based
on the operands, which would be too inconsistent.

--
Jonathan Allen


"Joe "Nuke Me Xemu" Foster" <joe@bftsi0.UUCP> wrote in message
news:3ac38fd4@news.devx.com...
> "Jonathan Allen" <greywolf@cts.com> wrote in message
<news:3ac384c2@news.devx.com>...
>
> > > Oh, and about Boolean logic -- let's see what Beta 2 brings.
> >
> > I don't see anything wrong with the Boolean logic. You just have to be
aware
> > of what data type you are using. Like I told Dan, don't use a BitAnd on
a
> > Boolean or a And on an Integer and you will be fine.
>
> Could the And operator be overloaded? If the compiler can tell that both
> operands to And, Or, Xor, Eqn, etc., are Boolean, perhaps it could put
> in short-circuiting automagically if there's a nearby Option Lazy.
>
> At this rate, we'll soon have as many tweaks as MSVC++.
>
> --
> Joe Foster <mailto:jfoster@ricochet.net> Greed = God?
<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!
>
>

[Dan Barclay]

On Thu, 29 Mar 2001 09:38:51 -0800, "Sjoerd Verweij"
<nospam.sjoerd@sjoerd.org> wrote:

>> I think you may want to pay closer attention to my position on this
>> stuff. I'd like to get into dotnet. The issue is that MS thinks VB
>> is a playtoy to be changed at a whim.
>
>The reasoning seems to be:

Nope, the reasoning is on a higher level than that. None of the
individual issues is insurmountable.

The issue is this: For someone invested in an application (or
considering it now), the concept of investing in the code for a term
approaching the usefulness of the application is a very high priority.

MS has changed VB before, with the response that "we get it now" and
"yes, VB is intended for development of real applications". Now,
they've changed it again and for no good technical reason. There are,
quite literally, billions of lines of existing MS Basic code that must
now be abandoned. Some reasonable percentage of that was intended for
(and appropriate for) movement into "the next big thing".

>
>- We're losing DF, arrays, -1, lots of stuff that just has to be changed due
>to interoperability considerations. Sure, why interoperate... we could
>become FoxPro! Uh-uh... VB is in NET, and the pain will have to be borne.
>
>- Hey, while we're at it... we're breaking almost all existing code anyway,
>so why not clean up some other things?

Well, we *weren't* breaking almost all existing code. As I've said
before, some things need to change in C/C++ as well. Do you think
letting a changefest get out of hand over there would work? Nope,
because MS itself (including the dotnet crew) depend on that code.

>Taken by themselves, some changes are gratuitous. But since the jump is huge
>anyway because of other issues, why not improve consistency? For example,
>are you really that attached to Set

The only change that is needed for "the jump" as you call it is loss
of DF and some will argue that point. If you look around I think
you'll find me explaining that change rather than trashing it. The
rest of them are gratuitous and there is no need for it. As for
consistency, consistency with the *existing* language is far more
important. Just think of the confusion of those trying to maintain
existing applications and provide dotnet extensions of (or
replacements for) them.

Again, none of the changes (individually) cannot be worked around with
the possible exception of logical expressions. However, there are
only two primary problems:

Language Stability
Trust

>Oh, and about Boolean logic -- let's see what Beta 2 brings.

Yes, we have hopes they will fix it but they haven't committed yet.

Dan
Language Stability is a *feature* I wish VB had!
(#6)

[Dan Barclay]

On Thu, 29 Mar 2001 10:04:32 -0800, "Jonathan Allen"
<greywolf@cts.com> wrote:

>You problem seems to be that you don't understand strongly typed operators.

I understand strongly typed operators very well. The issue is that
you'll have to go through the code and change it.

The remark was made that just changing to bitwise would take care of
the problem. My point is that it will not, and it won't. Simple
enough?

Dan
Language Stability is a *feature* I wish VB had!
(#6)

[Sjoerd Verweij]

> I don't see anything wrong with the Boolean logic. You just have to be
aware
> of what data type you are using. Like I told Dan, don't use a BitAnd on a
> Boolean or a And on an Integer and you will be fine.

I like the real operators a whole lot better -- I was just talking about the
migration wizard.

[Joe \Nuke Me Xemu\ Foster]

"Jonathan Allen" <greywolf@cts.com> wrote in message <news:3ac3b099@news.devx.com>...

> > Could the And operator be overloaded?
>
> Only if we dropped short-circuiting. Otherwise its behavior will vary based
> on the operands, which would be too inconsistent.

Call it an optional optimization then. If .NOT languages must differ only
in the use of curly braces, why bother having multiple .NOT languages?

--
Joe Foster <mailto:jfoster@ricochet.net> Sacrament R2.45 <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!

[Dan Barclay]

On Thu, 29 Mar 2001 13:55:48 -0800, "Jonathan Allen"
<greywolf@cts.com> wrote:

>> Could the And operator be overloaded?
>
>Only if we dropped short-circuiting. Otherwise its behavior will vary based
>on the operands, which would be too inconsistent.

You have no clue what you're talking about. Short circuiting is
determined *after* logical functionality is determined.

Existing DOS apps short circuit and handle And (under the hood) as
straight logical decisions or as bitwise operators depending on
context. This is done only as an optimization and is not a guaranteed
behavior but only because they chose not to define it as a language
behavior. See the compiler examples in:

http://www.mvps.org/vb/index2.html?tips/truth.htm

Dan
Language Stability is a *feature* I wish VB had!
(#6)

[Jonathan Allen]

If A() And B() Then

If A and B return Booleans, B is not always executed.
If they return Integers, then B is always executed.

This is what I meant by inconsistent behavior.

--
Jonathan Allen


"Dan Barclay" <Dan@MVPs.org> wrote in message
news8h7cts56gckll5se91m1klm5ddt01pqpi@4ax.com...
> On Thu, 29 Mar 2001 13:55:48 -0800, "Jonathan Allen"
> <greywolf@cts.com> wrote:
>
> >> Could the And operator be overloaded?
> >
> >Only if we dropped short-circuiting. Otherwise its behavior will vary
based
> >on the operands, which would be too inconsistent.
>
> You have no clue what you're talking about. Short circuiting is
> determined *after* logical functionality is determined.
>
> Existing DOS apps short circuit and handle And (under the hood) as
> straight logical decisions or as bitwise operators depending on
> context. This is done only as an optimization and is not a guaranteed
> behavior but only because they chose not to define it as a language
> behavior. See the compiler examples in:
>
> http://www.mvps.org/vb/index2.html?tips/truth.htm
>
> Dan
> Language Stability is a *feature* I wish VB had!
> (#6)

If I had to use curly braces, I'd rather switch to Delphi ...

Jens

I still don't get all this whining about True being changed.
True couold have the value "LittleGreenMenFromOuterSpace" for all I care. I
was always taught not to rely on intrinsic values of constants. Ins't that
the reason they are defined ?

Jens

[Jon Ogden]

"Zane Thomas" <zane@mabry.com> wrote in message
news:3ae65071.312501750@news.devx.com...
> On Wed, 28 Mar 2001 20:49:45 GMT, kylix_is@hotmail.com (Mike Mitchell)
> wrote:
>
> >My seven-year-old TV does it in less than that.
>
> Again, So F*cking What? That stupid 'argument' has been rebutted a number
> of times already and yet - as with other stupid 'arguments' you make - you
> simply persist. Quit wasting everyone's time, you're not nearly as cute
> as you think you are.
>
The one thing we did learn was that his TV is as old as the computers he
programs on. He is an embarrassment. I suspect that he's driven many of the
..NOT group into silence because he kept agreeing with them.