Should dmd have given me a warning at least?

[WhatMeWorry via Digitalmars-d-learn]

On Friday, 12 September 2014 at 22:53:35 UTC, Ali Çehreli wrote:
> On 09/12/2014 03:44 PM, WhatMeWorry wrote:
>>
>> // the following two lines compile cleanly but when executed, 
>> I get
>> // D:\Projects\Derelict>02_SimpleOpenGL_3_3_program.exe
>> // object.Error: Access Violation
>> // ----------------
>>
>> string glShadingLangVer =
>> to!string(glGetString(GL_SHADING_LANGUAGE_VERSION));
>
> According to the documentation, you are supposed to make sure 
> that glGetString does not return 0:
>
>   https://www.opengl.org/sdk/docs/man2/xhtml/glGetString.xml
>
> (See Notes and Errors there.)
>
>> writeln("glShadingLangVer is ", glShadingLangVer);
>>
>>
>>
>>
>> glGetString has the following signature:
>>
>> const GLubyte* glGetString(GLenum name);
>>
>> I presume the const is causing the problem.
>
> No, const would have caused a compilation error. Access 
> Violation is a runtime error caused by invalid memory access 
> location like 0. ;)
>
> Ali

Isn't this a contradiction.  The documentation says "glGetString
returns a pointer to a static string..."  But further on down it
then says "If an error is generated, glGetString returns 0."
This is the numeric value zero, right?

So how can glGetString() return both a string and a zero?

Btw, your book is excellent, Ali.