Hi!
I'm testing the possibilities of GLBasic.
I just downloaded / installed / upgraded GLBasic to GLBasic IDE, Version: 5.235.
I copy / paste the first sample.
PRINT "HELLO WORLD!" ,100,100
SHOWSCREEN
MOUSEWAIT
END
I compile / run the app and this show's on screen then message text.
The problem happens when I touch the screen with the mouse [function MOUSEWAIT jump to END for exit the app].
The computer freezes 4-5 secs before exit on this hardware:
HP Pavillion dv4000. Cpu: Intel Centrino 1.6 Ghz. 1 Gb RAM.
Video Card: Intel Mobile 915GM/GMS / 910GML 128 Mb.
and...
The computer freezes 18 secs before exit on this hardware:
UMPC Asus R2H. Cpu: Intel Celeron-M 900 Mhz. 1 Gb RAM.
Video Card: Intel Mobile 915GM/GMS / 910GML 128 Mb. (It's the same video card).
I tried to compile for CE devices... and in my HP6815 the application exists instantaneously.
I found other post on this Forum with the same problem.
I upgraded the drivers of my Video Card to lastest version.
Please, ¿any solution?.
TIA, Hark0
This is really bad. I have no idea, why this is so.
Please try this:
PRINT "Hello World", 0,0
SHOWSCREEN
MOUSEWAIT
foo()
FUNCTION foo:
LOCAL start = GETTIMERALL()
INLINE
extern void __ShutDownGLBasic();
__ShutDownGLBasic();
ENDINLINE
DEBUG "shutdown: " + (GETTIMERALL()-start) + "\n"
ENDFUNCTION
Then enable the debug mode, and run the program. After you press the mouse button, the program crashes horribly. But in the editor's output, you should finr "shutdown 132.222" or something. Please post this line here.
Ok, give me 5 minutes.
Hi!
I have passed this test on 5 computers:
PC 1: - NVIDIA GeForce4 MX 4000 - Intel Pentium 4 - 3Ghz - 512 Mb RAM - Windows XP
--------------------------------- Start Output
*** Finished *** Elapsed: 1.9 sec. Time: 09:55 Build: 1 succeeded.
Start debug session.
Injection started
--------------------------------- End of Output
PC 2: - NVIDIA GeForce4 8600 GS - Intel Core 2 Quad - 2.66Ghz - 4Gb RAM - Windows Vista
--------------------------------- Start Output
*** Finished *** Elapsed: 1.6 sec. Time: 10:00 Build: 1 succeeded.
Start debug session.
Injection started
shutdown: 30.98026425
--------------------------------- End of Output
PC 3: - SiS M760GX - Mobile AMD Sempron 1.58 Ghz - 1Gb RAM - Windows XP
--------------------------------- Start Output
*** Finished *** Elapsed: 5.4 sec. Time: 10:14 Build: 1 succeeded.
Start debug session.
Injection started
--------------------------------- End of Output
PC 4: - Intel Mobile 915GM/GMS / 910GML 128 Mb. - HP Pavillion dv4000 - Cpu: Intel Centrino 1.6 Ghz. 1 Gb RAM - Windows XP
--------------------------------- Start Output
*** Finished *** Elapsed: 2.6 sec. Time: 10:20 Build: 1 succeeded.
Start debug session.
Injection started
--------------------------------- End of Output
PC 5: (UMPC) - Intel Mobile 915GM/GMS / 910GML 128 Mb - UMPC Asus R2H. Cpu: Intel Celeron-M 900 Mhz. 1 Gb RAM - Windows XP
--------------------------------- Start Output
*** Finished *** Elapsed: 9.0 sec. Time: 10:45 Build: 1 succeeded.
Start debug session.
Injection started
--------------------------------- End of Output
Comments about this test:
PC 1:
Shows the Msgbox: Please update your GFX adapter's driver so it can render OpenGL by HARDWARE.
The app start and show "Hello world". Close at click mouse. NO shutdown text debug message (!)
PC 2:
The app start and show "Hello world". Close at click mouse. All done. ;-)
PC 3:
The app start and show "Hello world". Close at click mouse. NO shutdown text debug message (!)
PC 4:
The app start BUT NOT show "Hello world". The app window appear "freezed". Close at click mouse. NO shutdown text debug message (!)
PC 5:
The app start SHOW "LOADING". The app window appear "freezed". Close at click mouse. NO shutdown text debug message (!)
:noggin:
Please help.
TIA, Hark0
Another test on the UMPC computer:
I started a project on DEV-C++ 4.9.9.2..... with OpenGL.
I loaded the default template for OpenGL.
When I run the app, this shows a triangle rotating in multicolor.
Esc return to OS. Zero problem.
(With this test only I was trying to reject failures due to OpenGL).
TIA, Hark0
much month ago i have the same problem.
i think its the initialisation of opengl window and closing by ending the programm.
as i wrapped the xors3d, schranz0r code a routine that the opengl window doesn´t opening at start programm.
only the xors3d dx9 window opening.
with that the games doesn´t hang for seconds.
perhaps gernot can do in next update that the opengl window opening and closing intern routine is a little bit better. (if thats the problem!)
sorry for my bad english...
Hmm i hope you can german ;)
also ich denke das hat was mit den treibern zu tun. ich habe intel gma treiber (original ungeupdated) und habe auch das das fenster noch so 3 sekunden steht. (MacMini). Lade ich aber das gleich system (ist BootCamp installation) per vmware fusion dann hängts nicht. Grafiktreiber = anderer.
Mfg
Quote from: trucidare on 2008-Sep-23
Hmm i hope you can german ;)
also ich denke das hat was mit den treibern zu tun. ich habe intel gma treiber (original ungeupdated) und habe auch das das fenster noch so 3 sekunden steht. (MacMini). Lade ich aber das gleich system (ist BootCamp installation) per vmware fusion dann hängts nicht. Grafiktreiber = anderer.
Mfg
I used a web translator....
PASTING:
also I think this what has to do with the drivers. I have Intel gma driver (original ungeupdated) and the window also has still so 3 seconds stands. (MacMini). If I load, however, the immediately system (boat camp is not an installation) by vmware fusion then hängts. Graphic driver = more other.
hmmm
Quote from: Heiko on 2008-Sep-23
much month ago i have the same problem.
i think its the initialisation of opengl window and closing by ending the programm.
as i wrapped the xors3d, schranz0r code a routine that the opengl window doesn´t opening at start programm.
only the xors3d dx9 window opening.
with that the games doesn´t hang for seconds.
perhaps gernot can do in next update that the opengl window opening and closing intern routine is a little bit better. (if thats the problem!)
sorry for my bad english...
Hi Heiko... Can you send-me a sample code (zip) or else?
TIA, Hark0
Then you have to ask Schranz0r because the code what i have is based on one xors3d routine to set HWND.
Perhaps Gernot have an inline idea for a function special for you to close all when app ends.
PC1: Please update to the latest driver from NVidia.
PC 3,4,5: I don't get it. It's insane. I'll try to reproduce that on a virtual machine.
Is there any software that's installed on all these 3 pcs?
Did you try running the program again? Maybe a 2nd time helps?
On PC2 you get 30 seconds to free the memory. Totally crazy.
I would send you a debug file that helps me finding this bug, if you want?
Quote from: Kitty Hello on 2008-Sep-23
PC1: Please update to the latest driver from NVidia.
Ok, no problem... it's a old PC on office.
PC 3,4,5: I don't get it. It's insane. I'll try to reproduce that on a virtual machine.
Is there any software that's installed on all these 3 pcs?
Did you try running the program again? Maybe a 2nd time helps?
I tried the test with many configurations (such as kill unnused win32 apps). Same result. :(
On PC2 you get 30 seconds to free the memory. Totally crazy.
I would send you a debug file that helps me finding this bug, if you want?
You get my email at my profile info. :good:
Thanks for all.
=D
Might be worth checking to make sure there are no system devices in a red or yellow state and are all functioning correctly.
It would also be worth finding out whether the problem occurs in full screen and/or in a window.
Texture clamping can also cause strange problems, so it might be worth looking at that.
When I was programming using BlitzMax for a UMPC (in my case an Asus machine), restoring back to the desktop was slow too - you've got a combination of slow processor, slow storage device, not particularly fast memory and not much of it anyway.
Would be worth upgrading your copy of GLBasic too.
Quote from: MrTAToad on 2008-Sep-24
...
When I was programming using BlitzMax for a UMPC (in my case an Asus machine), restoring back to the desktop was slow too - you've got a combination of slow processor, slow storage device, not particularly fast memory and not much of it anyway.
...
I used Blitzmax, Pelles C, Dev-C++ on my Asus R2H (umpc) and I no have any problem (on Dev-C++ using SDL, the OpenGL not fails). Sincerelly the machine is not the "Most speedest Hardware", but the problem of my initial post are strange, due the popularity of the video card use (Intel 915GM).
Regards, Hark0
I'm tracing it already. It almost solved, but there's still the problem, that the textures don't show now.
Hi:
First of all please excuseme my very bad english (I'm Spanish).
I make test and I discover some things.
The tests was make with VAIO TX3HP (with INTEL 945GM).
See this code:
---------------------------
PRINT "HELLO",0,0
SHOWSCREEN
MOUSEWAIT
END
---------------------------
If you test it with (intel graphic chipset), and you use 640x480 o 800x600, the text draw well (no problem).
But if you use 1024x768 (in window mode) then the text not show.
If you use 1024x768 at full screen, then YES show correctly.
The problem is the coordinates in some modes takes -20 pixels, then the text NOT show.
But if you use PRINT "HELLO",0,25 then you see the HELLO at 0,0 position.
This happen only in some cases (with 1024x768 in window mode, maybe other resolutions).
And Me too have the same problem with sleep when the program exit (with window graphics take 4 o 5 seconds, and with full screen take 10 seconds or more).
I dont know if you understand me, if not i make for you some test. :good:
Kind Regards.
Ouch!
So, make your game 640x480, then do this:
DRAWLINE 0,0, 639,479, RGB(255,255,255)
SHOWSCREEN
MOUSEWAIT
and then post a screenshot here, please.
I have some news for you. =D
I discovert the problem is use the same or more resolution than your screen resolution.
Example:
If your desktop resolution is 1024x768x32 then test this:
SETSCREEN 1024,768,0
PRINT "HELLO BOY",0,0
SHOWSCREEN
WAITMOUSE
END
Test yourshelf change the SETSCREEN X,Y with your Desktop resolution or more. :good:
Definitely a driver bug.
Quote from: Kitty Hello on 2008-Sep-25
Definitely a driver bug.
Maybe, but I get the error in different computers (with different graphics board).
If you test in your own computer, you get the same error. ;-)
For those users who have the problem that I was mentioning to the beginning of the thread about Pause/freeze at exit app, a "temporal" solution that I have found is the installation of VirtualBox software.
VirtualBox don't use the Intel video driver, but shows the messagebox about "upgrade your hardware for support OpenGL".
I think is better to do 1 click in the msgbox for that to wait 4 seconds.
=D
Regards
The wait at program end is fixed in version 6.
We also found out, that when your desktop is 1024x768, and you try 1024x768 windowed, mode, the visible area is _not_ 1024x768, since you can't create a window bigger than the desktop size in Windows. Thus, you have to subtract the size for the window caption area, and thus loose the upper few pixels.
Good work Gernot :good:
QuoteGernot, do you know how to write a portable function that determines the size of the client area within a window? I could probably come up with some Win32 api code for that, but this wouldn't be cross platform.
No, sorry.