shearZ3D

[TBurnettIII]

Shouldn't the glm::shearZ3D transform be:

m[2][0] = s;
m[2][1] = t;

[plasmacel]

I have checked what are you talking about, and you are right.
GLM should follow the GLSL convention and handle it's matrices in column-major fashion.
According to that, not only shearZ3D, but all the shear functions (and possibly other transformations) are wrong, because it looks like they are handled in a row-major fashion, which is the transposed.

http://www.cs.colorado.edu/~mcbryan/5229.03/mail/55.htm
http://run.usc.edu/cs480-s13/lec05-viewing/05-viewing-6up.pdf

[TBurnettIII]

Thanks for verifying that issue. Has a bug report been submitted then?

Thomas.

From: Thom de Villa [mailto:notifications@github.com]
Sent: Sunday, October 18, 2015 4:25 AM
To: g-truc/glm glm@noreply.github.com
Cc: TBurnettIII tburnett@rattansoftware.com
Subject: Re: [glm] shearZ3D (#403)

GLM should follow the GLSL convention and store it's matrices in column-major order.
According to that, not only shearZ3D, but all the shear functions are wrong.

references
http://www.cs.colorado.edu/~mcbryan/5229.03/mail/55.htm http://www.cs.colorado.edu/%7Emcbryan/5229.03/mail/55.htm
http://run.usc.edu/cs480-s13/lec05-viewing/05-viewing-6up.pdf

—
Reply to this email directly or view it on GitHub #403 (comment) . https://github.com/notifications/beacon/ANlv6KyzJR1kZjD0bVSgYx2SB01yMlvNks5o81zkgaJpZM4F8RN5.gif

[Groovounet]

This issue is fixed in GLM 0.9.8 branch for GLM 0.9.8.3 release.

Thanks for reporting,
Christophe