As the title implies, this is an update to the Pages Jiggle.qtz from the OpenCL Developer Examples for Quartz Composer that allows it to function properly.
Actually, I just posted a completely correctly, fixed version of the Mesh Bend filter, so I should have waited on posting this, though it works fine with this example.
The problem with the mesh filters is that an input get assigned (like "attenuation" in Mesh Noise), which isn't allowed, so you make a new float, and assign that.
This is actually more faithful to the full functionality of the original Mesh Bend filter, and provides 100% similarity, with code correctly altered to fit OpenCL standard.
It would need to be placed in the folder to load the Coil.dae.
to bad the ipad magic is a private quartz core CAFilter. i mean the page curl is very cool on both ipad and qc. just tracking the touches angles and points etc.. to do the flip is complicated but you cant stop there you have to be able to zoom in and out plus scroll and track, pan, rotate etc to be able to read the pdf. i can read pdf on my ipad like this. i mean personally, i think its against the rules to submit an app using private frameworks. pretty sure this qc example is fair to use.
I had times of thinking about rolling back to the GM on a partition even, but things are just moving forward, and it doesn't make sense. Ultimately, if you want to make sure anything will be able to run successfully for everyone, you have to be able to test it on a recent (or most recent) build.
I would venture to guess that these changes to the mesh filters are totally backwards compatible, but not the other way around, as it's a matter of the OpenCL compiler having been corrected to catch errors that it didn't before.
Congratulations on getting a working model for Pages Jiggle back in the public sphere.
& thank you.
cool what did you do to make it work ?
A little iPad magic...
Actually, I just posted a completely correctly, fixed version of the Mesh Bend filter, so I should have waited on posting this, though it works fine with this example.
The problem with the mesh filters is that an input get assigned (like "attenuation" in Mesh Noise), which isn't allowed, so you make a new float, and assign that.
This is actually more faithful to the full functionality of the original Mesh Bend filter, and provides 100% similarity, with code correctly altered to fit OpenCL standard.
It would need to be placed in the folder to load the Coil.dae.
to bad the ipad magic is a private quartz core CAFilter. i mean the page curl is very cool on both ipad and qc. just tracking the touches angles and points etc.. to do the flip is complicated but you cant stop there you have to be able to zoom in and out plus scroll and track, pan, rotate etc to be able to read the pdf. i can read pdf on my ipad like this. i mean personally, i think its against the rules to submit an app using private frameworks. pretty sure this qc example is fair to use.
"You sunk my Software Update!" ;)
sometimes i think about going back to 10.6.1
I was getting impatient! ;)
I had times of thinking about rolling back to the GM on a partition even, but things are just moving forward, and it doesn't make sense. Ultimately, if you want to make sure anything will be able to run successfully for everyone, you have to be able to test it on a recent (or most recent) build.
I would venture to guess that these changes to the mesh filters are totally backwards compatible, but not the other way around, as it's a matter of the OpenCL compiler having been corrected to catch errors that it didn't before.