[Development] Qt5 build issue for Harmattan

[Sean Harmer]

On Saturday 25 August 2012 12:45:10 Laszlo Papp wrote:
> > We cannot really fix this within Qt as each implementation is free to
> > 
> >> define
> >> GLchar to whatever they like within the confines of the spec. Usually
> >> it's
> >> just a typedef to char but that is not necessarily always the case.
> > 
> > I would personally appreciate a solution upstream, if possible. I know,
> > Harmattan is not a tier 1 platform, and will probably never be, but it is
> > a
> > regression for the build either way. We have had working builds before.
> 
> The code already has an entry like this:
> 
> # ifndef GLdouble
> typedef GLfloat GLdouble;
> # endif

Yes and that is on my list of things to see if we can remove. I left it 
present in the patch you referenced to play it safe.

> I can submit a patch to gerrit like this:
> 
> # ifndef GLchar
> typedef GLbyte GLchar;
> # endif

I am now looking at trying something similar to this. 

I am curious though, if GLchar is not present, how does that header declare 
the function glBindAttribLocation() which here has the declaration:

GL_APICALL void         GL_APIENTRY glBindAttribLocation (GLuint program, 
GLuint index, const GLchar* name);

> Would that be alright? This would fix the regression for systems not
> matching the khronos implementation. I do not personally find GLchar in the
> specifications.

Table 2.2 on page 12 of the spec lists GLchar as minimum bit width of 8 bits. 
See http://www.khronos.org/registry/gles/specs/2.0/es_full_spec_2.0.25.pdf

Sean
--
Dr Sean Harmer | sean.harmer at kdab.com | Senior Software Engineer
Klarälvdalens Datakonsult AB, a KDAB Group company
Tel. Sweden (HQ) +46-563-540090, USA +1-866-777-KDAB(5322)
KDAB - Qt Experts - Platform-independent software solutions