GOD DAMN FUCK !!!

LOOK AT THIS !!!

FUCK FUCK FUCK !

program Project1;

{$APPTYPE CONSOLE}

uses
SysUtils;

var
i : longword;

begin

// [Hint] Project1.dpr(12): FOR or WHILE loop executes zero times - deleted
for i:=0 to 4294967295 do
begin
writeln('hi');
end;
end.

Ok... now I ll be calm and ask a simple question:

Why doesn't this work ?

SIGHhhhhhhhhhhhhhh :)

Like the C example was bad enough... but this totally sucks...

now the loop never gets executed ?!?!?!?!?!?!?!?!

Oh well.

Skybuck Flying wrote:
> var
> i : longword;
>
> begin
>
> // [Hint] Project1.dpr(12): FOR or WHILE loop executes zero times - deleted
> for i:=0 to 4294967295 do
> begin
> writeln('hi');
> end;
> end.
>
> Ok... now I ll be calm and ask a simple question:

Too late. Please calm down *before* you write your message. Expletives
are not appropriate here.

> Why doesn't this work ?

For the same reason the following will execute 4294967296 times:

var
i: LongWord;

List.Clear;
for i := 0 to List.Count - 1 do ;

You've encountered an edge case regarding loops and unsigned variables.
I consider it a bug. Please report it to Borland on Quality Central.
There may still be time to get it fixed before Delphi 9 ships.

The reason is so obvious, but nobody will be
prepared to bother to explain it to such a dickhead as yourself!

"Skybuck Flying" <nospam> wrote in message
news:qcp1
[..]

"Rob Kennedy" <me3> wrote in message
news:84u1
> Skybuck Flying wrote:
> > var
> > i : longword;
> >
> > begin
> >
> > // [Hint] Project1.dpr(12): FOR or WHILE loop executes zero times -
deleted
>
> Too late. Please calm down *before* you write your message. Expletives
> are not appropriate here.
>> For the same reason the following will execute 4294967296 times:
>
> var
> i: LongWord;
>
> List.Clear;
> for i := 0 to List.Count - 1 do ;
>
> You've encountered an edge case regarding loops and unsigned variables.
> I consider it a bug. Please report it to Borland on Quality Central.
> There may still be time to get it fixed before Delphi 9 ships.

Can you report it for me and say 'Credits to Skybuck for finding this bug'
lol :D

And credit yourself for understand or seeing that this is a bug :D hehehehe.

Anyway for now I solved it like this:

for i:=0 to 2147483647 do
begin

end;

etc...

31 bits.

For the current code that's enough since it uses a break to break anyway...

It will never reach 31 or 32 bits anyway ;)

But still.. a bug is a bug :D hehehe.

And one of these days I might need a loop like that for real ! ;)

Skybuck.

Skybuck Flying wrote:
> Can you report it for me and say 'Credits to Skybuck for finding this bug'
> lol :D

You can do it. It's not that hard. I have faith in you.

I may or may not report it myself, if I can reproduce it tonight. If I
report it, I assure you there will be no mention of you. So, do you want
the credit or not?

> And credit yourself for understand or seeing that this is a bug :D hehehehe.

You saw that it was a bug, too. Borland doesn't need a complete
explanation of *why* the bug occurs; that much will be apparent to
anyone who looks at it. All Borland needs is a demonstration of the bug,
which you already have.

> 31 bits.
>
> For the current code that's enough since it uses a break to break anyway...
>
> It will never reach 31 or 32 bits anyway ;)

You might consider using a "while" loop instead.

When you use a "for" loop, you tell the reader that there is at least
some case in which the loop will run all the way through. A "while" loop
gives you more expressive power about what the loop's termination
conditions are.

> But still.. a bug is a bug :D hehehe.
>
> And one of these days I might need a loop like that for real ! ;)

All the more incentive to report it to Borland. I'm not kidding. Please
report the bug.

Hmm, i really think your problem could be the fact that
Longword is more of an integer, thus you will only get the
- or + range of the count.


Skybuck Flying wrote:
[..]