GOSUB RETURN a bug of parser?

[TempodiBasic]

Hi guys and gals

following the code of Dav in another thread I have fallen into this feature / bug of the parser that comes out in runtime.

please copy paste and run this code to get the same experience

Code: QB64: [Select]

1. SCREEN 0
2. COLOR 12, 1
3. CLS
4. PRINT "I'm before the GOSUB"
5. GOSUB 1
6. GOSUB 3
7. GOSUB 2
8. GOSUB 3
9. COLOR 15, 3
10. PRINT "I'm in the main after all Gosubs"
11. PRINT "After me it will be triggered a runtime error for three times"
12.  
13. ' END REM it seems that END is a must  if there is a GOSUB before and after
14. '         the last line of code of the main program
15.  
16.  
17. 1:
18. DIM c AS STRING
19. PRINT "I'm into the GOSUB 1"
20. RETURN
21.  
22. 2:
23. c = "I'm a string"
24. RETURN
25.  
26. 3:
27. COLOR 14, 4
28. PRINT c; " it is a global string"
29. COLOR 12, 1
30. RETURN

So with the goal of distinguish between a starting point of a GOSUB and a call of a GOSUB it must be the use of the END to point out the end of the main program.
waiting feedbacks

[Unseen Machine]

Well, the code runs line after line, as they are GOSUB's and not actual SUB's/FUNCTION's they get run after your main code block. Hence the return command has no idea of where it should return to, so yes you'll need to use END or SYSTEM afterwards or simply add a SkipGosubs line

Code: QB64: [Select]

1. SCREEN 0
2. COLOR 12, 1
3. CLS
4. PRINT "I'm before the GOSUB"
5. GOSUB 1
6. GOSUB 3
7. GOSUB 2
8. GOSUB 3
9. COLOR 15, 3
10. PRINT "I'm in the main after all Gosubs"
11. PRINT "After me it will be triggered a runtime error for three times"
12.  
13. ' END REM it seems that END is a must  if there is a GOSUB before and after
14. '         the last line of code of the main program
15.  
16. GOTO SkipGosubs
17.  
18. 1:
19. DIM c AS STRING
20. PRINT "I'm into the GOSUB 1"
21. RETURN
22.  
23. 2:
24. c = "I'm a string"
25. RETURN
26.  
27. 3:
28. COLOR 14, 4
29. PRINT c; " it is a global string"
30. COLOR 12, 1
31. RETURN
32.  
33. SkipGosubs:
34.  

Unseen

[FellippeHeitor]

Yeah, not a bug in the parser. A bug in your code ;-)

Use END or SYSTEM.

[TempodiBasic]

Thanks
So my deprecated lazyness  to type is the real bug of this code! :-)

I thank you also for the alternative to END to solve the issue so your main message to me is control the flow of your program! and you can do this control using canonical ways like END  SYSTEM or alternative ways like the jump of GOSUB blocks.
I have found the deprecated STOP keyword that in an interpreted BASIC was useful for debug but here in QB64 is a duplicated of END.
It is not possible to write a modern code block (SUB/FUNCTION) before the Block of GOSUB!