Thanks everyone it now works natacha

cgc@humboldtblvd.com wrote:

Quoting zmoelnig@iem.at:

...

[pix_coordinate] is quite likely to not work properly with non-power-of-2 textures.

It does work. Use 0..height and 0..width instead of the normalized 0..1 values. The only limitation lies in the lack of repeat mode for the mapping - clamp to edge is the only one supported so far.

oh sorry. of course you are right with this. but in my opinion this is not really "work properly" (although it is openGL-standard) the discussion might turn out similar to the "float" vs. "int" talk @ colours.

now: what i really like about normalized values is, that they are abstracted and independent (aka: "normalized") .

short discussion of making [pix_coordinate] use normalized coordinates:

• you don't have to know/care whether your pixImage is power-of-two or

not (same point as in super-dooper-everything [pix_texture] vs. [pix_texture]/[pix_texture2])

• if your image is non-power-of-2 you don't have to know the dimensions.

• not openGL-comformant
• work to do.

i don't really understand why non-power-of-2 textures became such problems or rather: why implementations turned out to do it that complicated and if it was for quality's sake, why they didn't do it twice: once with normalized and once with not-normalized tex-coordinates (this is how openGL handles is anyway: provide umpteen functions for the same functionality in case someone wanted it to do it this or that way)

mfg.as.dr IOhannes

Thomas I am having some difficulties with the python external on os x. i cannot seem to get it to compile, i tried the CVS i had and the binary that is available but it tosses and error regarding the python location.

i have python2.2, and i was going to retry the CVS but the cvs server is down??

Pat

thomas i am having an error while trying to compile xsample too.

In file included from source/main.cpp:11: source/main.h:23:2: #error You need at least flext version 0.4.5 In file included from source/main.cpp:11: source/main.h: In static member function static void xsample::arrscale(int, const t_sample*, t_sample*, float, float)': source/main.h:156: no methodflext::ScaleSamples' source/main.h: In static member function static void xsample::arrmul(int, const t_sample*, t_sample*, float)': source/main.h:159: no methodflext::MulSamples' source/main.cpp: In function void xsample_main()': source/main.cpp:17: no methodflext::post' source/main.cpp:18: no method flext::post' source/main.cpp:19: no methodflext::post' source/main.cpp:20: no method flext::post' source/main.cpp:21: no methodflext::post' source/main.cpp:25: no method `flext::post' make: *** [pd-darwin/main.o] Error 1

On Mon, 5 Jan 2004, B. Bogart wrote to natacha Roussel:

I see your associated with concordia. Is there any interest from the the design department to do a PD workshop? (I have done a number in the past and am in Montreal often)

Hi. I have been away from home for the last two weeks. I'm pretty sure that people from the Pd-mtl users group would be interested in such a workshop eventually. Last month I organised the first meeting of the users group in my (big) bedroom. ;-) I will call a second meeting shortly. I don't know when you'd like a workshop or when you would be in Montréal.

I have not heard of much interest in Pd in Concordia arts/design yet (from what I've heard/seen/done, it's mostly MAX and jMax), but there are a few people getting involved in Pd stuff at the UdeM compsci dept (I was invited to a talk about their own video plugin...).

Mathieu Bouchard http://artengine.ca/matju