a mysmatched error in Opengl mode

[TempodiBasic]

Hi friends

I post here this experience to share results and thinking...

please copy this code and paste it into QB64 IDE

Code: QB64: [Select]

1. SCREEN _NEWIMAGE(800, 600, 32)
2.  
3.  
4.  
5. END
6.  
7. SUB gl ()
8.    _glviewport(0,0,_width, _height)
9. END SUB

as you can see the IDE gives back this message of error
Syntax error on line 8 (or if you put the cursor on line 8 of the code : Syntax error on current line).
I think that this kind of error message can be lead to misunderstandment...
the real error is NOT the line 8 of code, but the line 7 in which the name of the SUB is wrong... just if you want type the sub of Opengl that is _GL() and not GL()....
BUT in this case why IDE doesn't give a more specific error message like keyword out of _GL() sub or You cannot place this here
Why is this correction  useful? For the same reason that the IDE advices the user if DATA instructions are in the wrong place.
Thanks to read

[SMcNeill]

Nothing wrong with your SUB gl.  The only problem is you don’t have a sub _gl

SCREEN _NEWIMAGE(800, 600, 32)
 
 
 
END
 
SUB gl ()
   _GLVIEWPORT(0,0,_WIDTH, _HEIGHT)
END SUB

SUB _GL
   gl
END SUB

I’m thinking the above should work without any issues, unless there’s perhaps a naming conflict.

[Dimster]

Some of these error codes are a treat to decipher. In this case "No sub _gl found" would more enlightning. I would guess that it may be a huge task to try and describe every possible error condition arising from a particular line of code but it does feel like, the more the language expands the less insightful the error code message.

[Unseen Machine]

Some of these error codes are a treat to decipher. In this case "No sub _gl found" would more enlightning. I would guess that it may be a huge task to try and describe every possible error condition arising from a particular line of code but it does feel like, the more the language expands the less insightful the error code message.

QB64's implementation of GL is a hack....the fact that we have to use a seperate sub to use GL calls makes coding GL projects quite tedious especially as we cant actually control when the GL sub is called only handle if it gets to do anything with an AllowGL variable. Also annoying is the fact that all variables that get used by the _GL sub need to be Shared or Stactic and as you've found out you cant overload the GL function!

the real error is NOT the line 8 of code, but the line 7

I've noticed this too, the error line always seems to be one line off.

If you're really interested in OpenGl and want to use it for more than anything basic i'd suggest coding in C++. If not i've attached the basic setup i use for all my GL programs and included a basic sample that renders a coloured triangle.

Code: QB64: [Select]

1. SCREEN _NEWIMAGE(400, 300, 32)
2.  
3. _DISPLAY
4. _DISPLAYORDER _GLRENDER
5.  
6. _FPS 30 '    // Set maximium frames per second
7.  
8. DIM SHARED CamX!, CamY!, CamZ!, CamXRot%, CamYRot%, CamZRot%
9.  
10. '// Initial camera settings
11. CamZ! = -10
12.  
13. DO
14.   _LIMIT 30
15.  
16.   _DISPLAY
17. LOOP
18.  
19.  
20.  
21. SUB _GL
22.  
23.   _glClearColor 0, 0, 0, 1 '// Set background (clear) color to black
24.   _glClearDepth 1 '                 // Set clear depth value to farthest
25.   _glEnable _GL_DEPTH_TEST '         // Enables depth buffer for hidden surface removal
26.  
27.   '// Setup a perspective view
28.   _glMatrixMode _GL_PROJECTION
29.   _glLoadIdentity
30.   _gluPerspective 60, 400 / 300, .1, 400
31.   _glMatrixMode _GL_MODELVIEW
32.   _glLoadIdentity
33.  
34.   _glClear _GL_COLOR_BUFFER_BIT OR _GL_DEPTH_BUFFER_BIT '// Clear screen and depth buffers
35.   _glLoadIdentity
36.  
37.   '// Set the view to the cameras position
38.   _glRotatef CamXRot%, 1, 0, 0
39.   _glRotatef CamYRot%, 0, 1, 0
40.   _glRotatef CamZRot%, 0, 0, 1
41.   _glTranslatef CamX!, CamY!, CamZ!
42.  
43.   '// Draw the triangle
44.   _glBegin _GL_TRIANGLES
45.   _glColor3f 1, 0, 0
46.   _glVertex3f -1, -1, 0
47.   _glColor3f 0, 1, 0
48.   _glVertex3f 0, 1, 0
49.   _glColor3f 0, 0, 1
50.   _glVertex3f 1, -1, 0
51.   _glEnd
52.  
53. END SUB
54.  
55.  
56.  

Code: QB64: [Select]

1. '// GDK_GL_2 Base Code - By John Onyon a.k.a Unseen Machine \\
2. '/////////////////////////////////////////////////////////////////////////////////
3.  
4. REM $INCLUDE:'GDK_GL\GDK_GL.bi' '// Include a reference to the GDK_GL.bi library
5. '/////////////////////////////////////////////////////////////////////////////////
6.  
7. DIM SHARED Mouse(1) AS MouseState, Kb(1) AS KeyBoardState '// For Input capture
8. DIM SHARED CamVec AS VECTOR_GL, RadHelp#, Init_GL, Allow_GL AS _BYTE '// For Camera, Movement and GL initialisation
9.  
10. '// Initial Setup \\
11. SCREEN _NEWIMAGE(800, 600, 32) '// Create a 32 bit screen
12. _DISPLAY '// Turn off auto display
13. _DISPLAYORDER _GLRENDER , _SOFTWARE '// Set the display to render GL stuff then Software stuff
14. _FPS 30 '    // Set maximum frames per second
15.  
16. '// Define a variable to assist in first person camera movement caluclations
17. RadHelp# = (4 * ATN(1)) / 180
18.  
19. '// Now everything is set up we can allow SUB _GL, first we initialise it. \\
20. Init_GL = True
21. Allow_GL = True
22.  
23.  
24. '// Main program loop
25. DO
26.  
27.  _LIMIT 30 '// Limit the loop to a maximum of 30 loops per second \\
28.  
29.  
30.  '// Get the state off input devices \\
31.  GDK_Mouse_GetState Mouse(0) '// Get the mouse state
32.  GDK_Keyboard_GetState Kb(0) '// Get keyboard state
33.  
34.  '// Do game logic here \\
35.  
36.  
37.  '// Copy the mousestate/keyboardstate for comparisons next loop \\
38.  Mouse(1) = Mouse(0)
39.  Kb(1) = Kb(0)
40.  
41.  _DISPLAY
42.  
43. LOOP
44.  
45. '// End of main program loop
46.  
47.  
48. SUB _GL
49.  
50.  IF Allow_GL THEN
51.  
52.   _glClearColor 0, 0, 0, 1 '// Set background (clear) color to black
53.   _glClearDepth 1 '         // Set clear depth value to farthest
54.   _glEnable _GL_DEPTH_TEST '// Enables depth buffer for hidden surface removal
55.   _glEnable _GL_TEXTURE_2D '// Allow Texturing
56.  
57.   '// Setup a perspective view
58.   _glMatrixMode _GL_PROJECTION
59.   _glLoadIdentity '// Reset the stack
60.   _gluPerspective 80, 800 / 600, 1, 400 '// Field of view, Aspect ratio, near depth, far depth
61.   _glMatrixMode _GL_MODELVIEW
62.   _glLoadIdentity
63.  
64.   IF Init_GL = True THEN '// If GL is in initialisation mode (Loading stuff such as model and textures)
65.  
66.    '// Load everything here
67.  
68.  
69.  
70.    '// Change the GL mode setting
71.    Init_GL = False
72.  
73.   ELSE '// GL is in
74.  
75.    '// Clear the depth buffer and reset the view matrix (effectively CLS)
76.    GDK_GL_CLS
77.  
78.    '// Set enviroment to camera coordinates/Rotation
79.    GDK_GL_VECTOR_APPLY_RT CamVec
80.  
81.   END IF
82.  
83.  END IF
84. END SUB
85.  
86. '// library inclusions
87. REM $INCLUDE:'GDK_GL\GDK_GL.BM'
88. REM $INCLUDE:'UnseenGDK.bm'
89.  
90.  

Happy coding.

Unseen

[TempodiBasic]

Hi Unseen Machine

thanks for your feedbacks and suggestions coming out from your wide experience...
I must leave QB64 for C++ for a good OpenGl program! It is hard to choose.
For now I download you code posted here .

Surely I can affirm that my little experience in Opengl (I have used Opengl only now from within QB64) let me think OpenGl like a Batch language little flexible and much robust.
But maybe it appears so because in this manner is implemented into QB64...I'll discover this on going on