www.digitalmars.com         C & C++   DMDScript  

digitalmars.D.learn - An annoying compile issue

reply WhatMeWorry <kheaser gmail.com> writes:
I'm porting over some C++/glm/openGL code.  I've used gl3n for a 
while now (successfully) to port over glm code, but I've got this 
pebble in my shoe:

glm::mat4 model;
model = glm::scale(model, glm::vec3(size, 1.0f));   // size is a 
vec2

So my D code consists of:

mat4 model;
model = model.scale(vec3(size, 1.0f));  // size is a vec2


Error: function gl3n.linalg.Matrix!(float, 4, 4).Matrix.scale 
(float x, float y, float z) is not callable using argument types 
(Vector!(float, 3))


First, isn't Vector!(float,3) a template which creates a tuple 
consisting of float, float, float; so aren't the argument types 
identical?

I can workaround the issue by simply
model = model.scale(size.x, size.y, 1.0f);

but that seems like cheating :)

Is there an elegant D fix. Not to imply that there is any other 
kind.
Oct 25 2015
next sibling parent Rikki Cattermole <alphaglosined gmail.com> writes:
On 26/10/15 4:44 PM, WhatMeWorry wrote:

 I'm porting over some C++/glm/openGL code.  I've used gl3n for a while
 now (successfully) to port over glm code, but I've got this pebble in my
 shoe:

 glm::mat4 model;
 model = glm::scale(model, glm::vec3(size, 1.0f));   // size is a vec2

 So my D code consists of:

 mat4 model;
 model = model.scale(vec3(size, 1.0f));  // size is a vec2


 Error: function gl3n.linalg.Matrix!(float, 4, 4).Matrix.scale (float x,
 float y, float z) is not callable using argument types (Vector!(float, 3))


 First, isn't Vector!(float,3) a template which creates a tuple
 consisting of float, float, float; so aren't the argument types identical?


It is not a tuple.
vec2 != vec3. They are different types altogether.
Oct 25 2015
prev sibling parent reply Adam D. Ruppe <destructionator gmail.com> writes:
On Monday, 26 October 2015 at 03:44:31 UTC, WhatMeWorry wrote:

 First, isn't Vector!(float,3) a template which creates a tuple 
 consisting of float, float, float; so aren't the argument types 
 identical?


I'm not familiar with the gl3n library so I'm not sure what their 
Vector is, but most D templates are not created from tuples and 
don't work that way.

Looking at the source, it indeed seems to be based on a static 
array rather than a tuple:
https://github.com/Dav1dde/gl3n/blob/master/gl3n/linalg.d#L49

---
  vt[dimension] vector; /// Holds all coordinates, length conforms 
dimension.
---

Which means there's no simple, built-in way to getting a tuple 
out of that. You could write one though, and recently there was a 
post about maybe putting that in Phobos, but it isn't there now.


 model = model.scale(size.x, size.y, 1.0f);


meh that's prolly what I'd do... unless you wanted to modify the 
library and add some expansion code to that vector type.
Oct 25 2015
parent WhatMeWorry <kheaser gmail.com> writes:
On Monday, 26 October 2015 at 03:53:21 UTC, Adam D. Ruppe wrote:

 On Monday, 26 October 2015 at 03:44:31 UTC, WhatMeWorry wrote:

 First, isn't Vector!(float,3) a template which creates a tuple 
 consisting of float, float, float; so aren't the argument 
 types identical?


 I'm not familiar with the gl3n library so I'm not sure what 
 their Vector is, but most D templates are not created from 
 tuples and don't work that way.

 Looking at the source, it indeed seems to be based on a static 
 array rather than a tuple:
 https://github.com/Dav1dde/gl3n/blob/master/gl3n/linalg.d#L49

 ---
  vt[dimension] vector; /// Holds all coordinates, length 
 conforms dimension.
 ---

 Which means there's no simple, built-in way to getting a tuple 
 out of that. You could write one though, and recently there was 
 a post about maybe putting that in Phobos, but it isn't there 
 now.


 model = model.scale(size.x, size.y, 1.0f);


 meh that's prolly what I'd do... unless you wanted to modify 
 the library and add some expansion code to that vector type.


Ok. If it's good enough for ADR, it's good enough for me.  Thanks 
all.
Oct 26 2015