package provided via pip is outdated

[Axel1492]

Error when trying to run font-sdf.py

[i] Using GLFW (GL 4.3)
[i] Requesting "OpenSans-Regular.ttf" from remote server
[w] Data not available on remote server
[w] Falling back to default font
Traceback (most recent call last):
  File "C:\...\font-sdf.py", line 26, in <module>
    labels.append(text, regular, origin = (x,y,z), scale=scale, anchor_x="left")
  File "C:\Program Files\Python36\lib\site-packages\glumpy\graphics\collections\sdf_glyph_collection.py", line 76, in append
    V, I = self.bake(text, font, anchor_x, anchor_y)
  File "C:\Program Files\Python36\lib\site-packages\glumpy\graphics\collections\sdf_glyph_collection.py", line 128, in bake
    glyph = font[charcode]
  File "C:\Program Files\Python36\lib\site-packages\glumpy\graphics\text\sdf_font.py", line 75, in __getitem__
    self.load('%c' % charcode)
  File "C:\Program Files\Python36\lib\site-packages\glumpy\graphics\text\sdf_font.py", line 130, in load
    data,offset,advance = self.load_glyph(face, charcode)
  File "C:\Program Files\Python36\lib\site-packages\glumpy\graphics\text\sdf_font.py", line 102, in load_glyph
    hires_data = np.zeros( (hires_height,hires_width), np.double)
TypeError: 'float' object cannot be interpreted as an integer

[rougier]

It's probably because of hires_data = np.zeros( (hires_height,hires_width), np.double)
This should help: hires_data = np.zeros( (int(hires_height),int(hires_width)), np.double)

[Axel1492]

so I basically fixed it in that file:

        # Pad high resolution glyph with a blank border and normalize values
        # between 0 and 1
        hires_width  = int((1+2*self._padding)*width)
        hires_height = int((1+2*self._padding)*height)
        hires_data = np.zeros( (hires_height,hires_width), np.double)
        ox,oy = int(self._padding*width), int(self._padding*height)
        hires_data[oy:oy+height, ox:ox+width] = G/255.0

problem continues:

  File "C:\Program Files\Python36\lib\site-packages\glumpy\graphics\collections\base_collection.py", line 517, in _update
    texture = texture.reshape(shape[0],shape[1],4)
TypeError: 'float' object cannot be interpreted as an integer

solved:

            # shape[2] = float count is only used in vertex shader code
            texture = texture.reshape(int(shape[0]),int(shape[1]),4)
            self._uniforms_texture = texture.view(Texture2D)
            self._uniforms_texture.interpolation = gl.GL_NEAREST

After this the position (origin) in labels.append(text, regular, origin = (x,y,z), scale=scale, anchor_x="left") used to render the text in font-sdf.py has no effect on the rendering position...

[Axel1492]

so ...from what I have seen so far this seems to be a general issue with Python 3 and glumpy. As an example (int)/(int)=(float) in newer versions of python and not anymore (int)/(int)=(int) ....to maintain old behaviour all integer divisions have to be done with // operator (python>=2.2) or int().

[Axel1492]

OK ....I figured my problem out. I used pip install glumpy but that does give me an outdated version wich simply does not work. I recommend you update the package.

[rougier]

Thanks for the report . There is a bunch of PR waiting to be merge and after that I can make a new release. I leave this issue opened until I make the release.