loop every x seconds

[Shawn Anderson]

I have a simple program that loads data from an external file and displays it in a listview.

I want it to refresh the information every 10 seconds by calling a function called loadInfo() every 10 seconds after the single dialog window loads.

How do I do that? I tried putting it in the callback function but that doesn't work.

This seems like it should be simple to do.

[Michael Mattias]

Use SetTimer() and process WM_TIMER notification message.

Example here:
GUI + Worker Thread + Abort Demo 11-24-07

You may as well run the "get data from file" in a separate thread of execution, but its not necessary. That demo includes on-screen clock, which is basically the "proof" that the screen is getting refreshed.

MCM

[Dave Biggs]

What about TimeSetEvent

[Michael Mattias]

>What about TimeSetEvent

What about a Waitable Timer? Waitable Timer Object Demo June 2005

Lots of ways to do it. For that matter, LOOP on TIME$ can be made to work, but that would not only be ugly, it would earn negative panache points.

BUT..... why are we checking file contents on a timer in the first place? Seems to me we should be looking for a change in the file.

http://www.powerbasic.com/support/pbforums/showthread.php?t=23000

MCM

[Shawn Anderson]

Thanks everybody,

SetTimer() and WM_TIMER worked perfectly.

I couldn't get TimeSetEvent to work, it caused a GPF.

[Michael Mattias]

Well, since SetTimer()/WM_TIMER is working this may not be that important but...

>TimeSetEvent to work, it caused a GPF.

No, TimeSetEvent did not cause the GPF. The cause would be "programmer error."

It sounds like you are fairly new to programming with PB for Windows; and if you are a refugee from "Visual Anything" you probably have limited exposure to direct calls to the Windows' API.

However, the WinAPI is bug-free except AFAIK for GetPrivateProfileSection() on Windows/9x. (It's fine after that). (The error? it has an 'off by one' error in the return value).

FWIW, TimeSetEvent is the long way tp do what you want to do. SetTimer()/WM_TIMER is a much better fit.


MCM

[Shawn Anderson]

No, TimeSetEvent did not cause the GPF. The cause would be "programmer error."

Ha! yes clearly. I certainly didn't mean to imply that the function itself was faulty. Maybe I should of said "My program causes a GPF when I try to use timeSetEvent".

It sounds like you are fairly new to programming with PB for Windows; and if you are a refugee from "Visual Anything" you probably have limited exposure to direct calls to the Windows' API.

At least with PBWIN. I do a lot with PBCC (probably 90% of what I do) for CGI but callbacks and API functions are relatively rare for me. I get the gist of it but just don't do it enough to have it down smoothly. Usually a point in the right direction and I can figure it out. Since I've started using PB, about 9 years ago, I've done maybe 10 PBWIN programs total but hundreds of PBCC, mostly CGI.

[Mike Doty]

'Michael's code compiles fine in PBWin8. Trying to figure out why not with PBWIN9?

http://www.powerbasic.com/support/pbforums/showthread.php?t=35845

The cookie demo seems a bit hard to follow. It might be nice to pull out the required code.

'Syntax PBWin8 THREAD CREATE ProcName (param) [SUSPEND] TO hThread  'compiles fine PBWIN8
'Syntax PBWin9 THREAD CREATE FuncName (param) [StackSize,] [SUSPEND] TO hThread  'TO missing PBWIN9
THREAD CREATE WorkerThreadFunction (CBHNDL) SUSPEND TO hThread

[Michael Mattias]

>Trying to figure out why not with PBWIN9?
Referenced code:

Compiler: PowerBASIC Inc PB/Windows version 7.02

Haven't tried 9x myself, but I have to believe you still get a message from the compiler if it finds something it doesn't like. Message not shown.

The cookie demo?

Author: Michael Mattias, Racine WI April, 2001  using PB/DLL v 6.0

I code better now. Deal with it.
-- OR --
Do a 'find' here on FindFirstChangeNotification. I'm pretty sure others have created applications using these synchronization objects.

Generic comments re "old demos" not compiling with new version(s) of compiler:
This whine's time has come

[Mike Doty]

It was shown above (scrolls off screen)

'Syntax PBWin9 THREAD CREATE FuncName (param) [StackSize,] [SUSPEND] TO hThread 'TO missing PBWIN9

Good morning, Michael.
Here is another whine.
Why is the width of these code windows so small? Could the BBS software increase the width a bit, please.

[Michael Mattias]

FWIW, on the cookie demo (which really is not that different from some of my other demos), this one leaps out at me...

IF hFind <> %Invalid_HANDLE_VALUE THEN

In the current compilers and version of Win32API.INC, this will ALWAYS be true, since hFind is declared as a LONG and %INVALID_HANDLE_VALUE is defined as &hFFFFFFFF??? (a DWORD)... 6x would have found equality when hFind = -1&, but 7x+ would not.

I'll bet there's all kinds of LONG/DWORD mismatches in that code; and who knows what other things I would find were I inclined to do so.

But... maybe some ambitious person will update that to 8x or 9x and post it as a "reply" to the original. (What a great idea for one of those people who ask for project ideas!)

Then again, this may just be one of those cases where Mr. Zale has a very large smile on his face: Everyone has become 'DDT-Dependent' and a simple little SDK-style program sends them scurrying back to mama.

MCM

[Michael Mattias]

'Syntax PBWin9 THREAD CREATE FuncName (param) [StackSize,] [SUSPEND] TO hThread 'TO missing PBWIN9

Oh, "TO missing" is the error message?

Well my eyes are not what they used be but I sure as hell can see "TO hThread" in

THREAD CREATE WorkerThreadFunction (CBHNDL) SUSPEND TO hThread

Call me crazy, but this sounds like a compiler bug just begging to be properly reported.

Try putting a comma before SUSPEND and see what happens.


MCM

[Mike Doty]

My eyes went out last night trying to figure it out.

[Cliff Nichols]

MCM

In the current compilers and version of Win32API.INC, this will ALWAYS be true, since hFind is declared as a LONG and %INVALID_HANDLE_VALUE is defined as &hFFFFFFFF??? (a DWORD)... 6x would have found equality when hFind = -1&, but 7x+ would not.

Did you not already do one??? (or close to it) that did something along the lines of (If outside of range then cycle back to in range)
I KNOOOOOWWWWWW it has to be hidden in your libraries somewhere (and yes I know when not used often its hard to find)...I saw it somewhere in the forums, and everytime I look to find it...danged if I can.

Unfortunately the only hint of a clue I have is it had a line of if some value add &HFFFF to it....but the idea is clear as mud so I may not be describing correctly

[Michael Mattias]

I mostly use now

#INCLUDE "WIN32API.INC"
%INVALID_HANDLE_VALUE_LONG = -1& 

...
  LOCAL hFind AS LONG
 .....
  IF hFind <> %INVALID_HANDLE_VALUE_LONG THEN

Then there's always the MACRO Tom Hanlin came up with:

MACRO eq32 (a,b) = BITS???(a) = BITS???(b) 

... 
    IF Eq32 (hFind, %INVALID_HANDLE_VALUE)  THEN  ' 32bit compare of LONG with DWORD

MCM