I never noticed this before.

I was looking through the Win32 SDK Referance, when I came across this peculiar tidbit:

Quote:

---

GetMessage does not remove WM_PAINT messages from the queue. The messages remain in the queue until processed.
Note that the function return value can be TRUE, FALSE, or -1. Thus, you should avoid code like this:
while (GetMessage( lpMsg, hWnd, 0, 0)) ...
The possibility of a -1 return value means that such code can lead to fatal application errors.

---

Is it just me or does just about every single example of a windows programme use that line of code?

Hmmm...interesting. I use that line of code. Never had any problems with it. Has anybody?

So apparently we should be doing

while(GetMessage(&m, h, 0, 0) > 0)

I've also noticed that every example I've seen was the way described by shtarker.

Does this mean we should take care of all the other functions returning this so-called "boolean" value?

Why couldn't they just stick to 0 and 1??????

They realy should tell Petzold. As his MS press WIN32 bible uses it everywhere.

I use -1 and define it as INVALID. Not in the msg loop though.

Goes with having a menu in a child window. According to MS you can't.

From GetMenu() help in MSVS v6
"If the window is a child window, the return value is undefined."

In fact you can, but only in Win9x, 2000 and XP don't allow it and refuse to show the window at all.

Well most the code for my WinMain was appropriated (read copied) from the Win32 Bible, so only every single one of my programs has that loop in it.
But so far that part of my programme hasn't crashed on me, or atleast i don't think it has.

I am so lucky that I don't even use GetMessage heh :).

Thank'z to the fact that I am sort of "self-tought", I use 'PeekMessage', and a boolean variable that i'z only changed in the Message Procedure! :)

Maybe a good architecture 2 u'ze now that there i'z this new compication... BUT I have one comment to make about this...

A logical statement (such as 'if','while','for', ect.), I beleive, will take anything '<=0' as a false...

So I don't see why:

while(-1)

woulden't work...

~SPH :-p:):-p

>So I don't see why:

while(-1)

woulden't work... <

Because your beliefs are incorrect and you can't spell for ****.

PeekMesage()
does not differentiate WM_QUIT msg's,
does not allow the app to sleep when no msg's (and therefore uses more CPU resources resulting slower code).

If PeekMessage() causes such a code slow down then why does every real time game use this method for its main loop??? If that was the case then I dont think every single time ive looked at a 3D engine for Win32 it has used this method to do the message loop. Im not saying this is the best way im just pointing out that you should be careful about what you say b/c you dont know what he/she is developing.

Because a game is a constant stream of input or paint msg's. There are always mgs in to be returned and the last thing it wants is to be put to sleep by Windows.

Game loop checking for WM_QUIT msg's removing msg's as we go. The fact there is extra code in the main msg loop will cause some slow down as the if() and extra while() is evaluated but should not be noticeable.

Code:

---

while( !done )
{
      while( PeekMessage( &msg, 0, 0, 0, PM_REMOVE ) )
    {
        TranslateMessage( &msg );
        DispatchMessage( &msg );
        if(msg.message == WM_QUIT)
          done = 1;
    }
}

---

Games are not the only type of app. Some need to sit and wait to be used without using up your CPU constantly checking for new msg's. Or may have to run 24 /7 waking up as external events require the program.

(In my answer you will notice I put the slower code in braces as it was a secondary concern to me. If you are running games the last thing you need is _other_ app's slowing your PC down because the programmer used an inappropriate msg loop)

Well actually... thats nowere NEAR how I use it, and also I am so freaking insulted by the 2nd to last post up from here I feal I should not even post very much for my pure RAGE will get me kicked off of this board...

SPH ...

Whos post you talking about? Mine or novacains? B/C mine is true!

Sorensen is right on both counts.

FALSE == zero
TRUE != zero

therefore -1 == TRUE (in bool terms, not false)

and as to "self-tought", you should have consulted an expert or at least a spell checker.

Post how you use PeekMessage() then.

Quote:

---

Originally posted by minime6696
Well actually... thats nowere NEAR how I use it, and also I am so freaking insulted by the 2nd to last post up from here I feal I should not even post very much for my pure RAGE will get me kicked off of this board...

SPH ...

---

What are you complaining about now?

I think he is refering to Sorensen kindly pointing out the invalidity of his argument: "A logical statement (such as 'if','while','for', ect.), I beleive, will take anything '<=0' as a false... "

Quote:

---

Originally posted by shtarker
I think he is refering to Sorensen kindly pointing out the invalidity of his argument: "A logical statement (such as 'if','while','for', ect.), I beleive, will take anything '<=0' as a false... "

---

Well Sorenesen is correct.........and he isnt the first person to complain about haX0r talk on this board....

Anyway......if he doesnt agree....try

Code:

---

if(-1)cout << "Hello World";

---

If anything under 0 is a false then "Hello World" will not appear on screen.........which is not the case....

false == 0

true != false

Oh ok i think my professor went over that a few years ago lol but i dont remember. But that is a good thing to know.

Quote:

---

Originally posted by minime6696
I should not even post very much for my pure RAGE will get me kicked off of this board...

---

Is that Dean? Dean-o! How's the 4x4 doing, Buddy?

That most certainly is NOT Dean