Re: [PD] Changing array curves with mouse interaction
Thanks for the feedback on this 'problem'. I forgot to mention that the question arouse when a student of mine used a table of size 3. The proportional clicking area then got comparable very little if used a bigger table. Did not realise that before. /Björn Eriksson Den 18 apr 2013 18.51, Miller Puckette m...@ucsd.edu skrev: I'm not sure but I believe that's because of rounding to the nearest pixel. I don't believe Tcl/TK does any anti-aliasing, even if the underlying graphical system does. cheers Miller On Thu, Apr 18, 2013 at 12:30:55PM -0400, Billy Stiltner wrote: anyone figured out why sometimes ... ___ Pd-list@iem.at mailing list UNSUBSCRIBE and account-management - http://lists.puredata.info/listinfo/pd-list
Re: [PD] Sensors GPIO Raspberry Pi Pd
Well that looks a total mess... I did sudo apt-get install i2c-dev before all this, as well as all the things in http://www.instructables.com/id/Raspberry-Pi-I2C-Python/#step1 except I didn't use any python. The segfaults might be because you are asking for more than 32 bytes in a read, or something wrong in the code, the listing below looks like it goes wrong right at the beginning. Martin On 2013-04-19 17:37, Julian Brooks wrote: As I'm new to all this C stuff I just had a look inside the 'hello' file and there's a few bits in there which may be of interest: ^@D6T_checkPEC says 0x%02X (which I think is ok from looking at the code?) then it all goes wrong ^@^@^@Unable to create socket (%d) ^@^@^@ %s ^@^@^@127.0.0.1 http://127.0.0.1^@^@^@Unable to make address from %s ^@%s: initialized:%d ^@/dev/i2c-0^@^@Failed to open the bus. (%d) ^@^@^@Failed to acquire bus access and/or talk to slave. (%d) ^@^@^@^@Write failed (%d) ^@^@Read failed (%d) ^@^@^@%d ^@0x%02X ^@^@^@d6t8l^@^@^@ %d^@ ^@^@^@sendto error (%d) BTW - I've removed libi2c-dev again and will leave it that way for now. Going to stop now but will be back tomorrow. Cheers, Julian On 19 April 2013 22:07, Julian Brooks jbee...@gmail.com mailto:jbee...@gmail.com wrote: Edited your original file and all I changed was the ip address to 127.0.0.1 and still seg fault (double checked /etc/hosts too). Also tried reinstalling libi2c-dev and then running 'hello' - same. On 19 April 2013 21:54, Julian Brooks jbee...@gmail.com mailto:jbee...@gmail.com wrote: hmmm - both files: ./hello-2 Segmentation fault Try again... On 19 April 2013 21:51, Julian Brooks jbee...@gmail.com mailto:jbee...@gmail.com wrote: Ok -found I had to remove 'libi2c-dev'. Then builds. More soon... On 19 April 2013 21:28, Julian Brooks jbee...@gmail.com mailto:jbee...@gmail.com wrote: Hey Martin, When I try to compile hello.c I get: gcc -o hello hello.c In file included from hello.c:8:0: /usr/include/linux/i2c-dev.h:38:8: error: redefinition of 'struct i2c_msg' /usr/include/linux/i2c.h:67:8: note: originally defined here /usr/include/linux/i2c-dev.h:90:7: error: redefinition of 'union i2c_smbus_data' /usr/include/linux/i2c.h:125:7: note: originally defined here Dunno if this is at all relevant but maybe this is a good time to say I have a rev1 RPi so I'm on i2c 0 not 1. The Pi install is very up to date though, including recent runs of hexxeh's 'rpi-update' tool so all the system's fresh. I did attempt to change the number of bytes to be read which perhaps would explain why the .h file's are complaining but I don't understand it for your version which is 'as-is'? In fact why don't I attach the notes I've made of the changes to the c file you sent. Was vaguely hoping I might be able to say 'ta da' but have fallen at the 1st fence:( bugger. Julian On 19 April 2013 19:51, Julian Brooks jbee...@gmail.com mailto:jbee...@gmail.com wrote: Hi Martin, Meant to add re setting baud rate: I've been making use of the gpio utility that comes with wiringPi https://projects.drogon.net/raspberry-pi/wiringpi/the-gpio-utility/ Very handy for getting a quick visualisation of the current state of all the pins and also easy-access to setting the baud rate too (amongst other stuff). Julian On 19 April 2013 14:36, Martin Peach martin.pe...@sympatico.ca mailto:martin.pe...@sympatico.ca wrote: Hi Julian, Yes I've been messing with coding it in c on the pi and sending the data to a [netreceive] in a Pd patch on another machine. I'm attaching the source code for the pi part and the Pd patch. The code can be compiled on the pi with gcc -o hello hello.c You need to set the IP address of the receiving machine in the code (I have 192.168.2.15, it could be 127.0.0.1 for the same machine). I tried changing the baud rate with sudo modprobe -r i2c_bcm2708 sudo modprobe i2c)bcn2708
Re: [PD] flext install (for another amateur)
Hi Chris, i have now looked into the Pd-extended 0.43.4 package. The layout of files has been changed in a way that it is incompatible to the flext build system (and also to the autoconf version). I will have to see how i can fix this. What you can do for the time being is download Miller's vanilla version and compile your external object using that. You can then use the binary with pd-extended as well. all the best, gr~~~ -- Thomas Grill http://g.org Am 17.04.2013 um 13:40 schrieb Chris Jack: Hi Thomas, Thanks a lot for your help. Seeing as I have not actively installed gcc from any other source I suppose it must be the one from the xcode/developer sdk. Btw, my sdk is 'MacOSX10.6.sdk' - the one appropriate for osx 10.6.8, afaik! My Pd (extended 0.43.4) installation does actually contain the g_canvas.h file in /Applications/Pd-extended.app/Contents/Resources/ include/pdextended I have the config-mac-gcc...etc file pointing PDPATH to / Applications/Pd-extended.app/Contents/Resources ...need I point a different variable to Resources/include/pdextended in order to see the header files? Cheers, On 16 April 2013 15:42, Thomas Grill g...@g.org wrote: Hi Chris, first of all, do you use the gcc coming with xcode or from some other source (fink, macports)? It seems that your platform SDK is somewhat strange, judging from the autoconfig build as it doesn't contain BlockMoveData and BlockZero. As for the buildsys build, it seems that your pd distribution (pd- extended?) doesn't contain the g_canvas.h file, which is a header file private to pd, but necessary for flext to work. Some pd distros do contain it, while others don't. Btw., there's a mailing list specialized to flext matters. all the best, Thomas -- Thomas Grill http://g.org Am 16.04.2013 um 15:59 schrieb Chris Jack: Hey folks, I have been trying in vain to install flext (cvs download) on my 10.6.8 snow leopard Mac. I've tried it every which way - both using the buildsys and autoconfig - and just cannot get the bugger to build. I have refreshed before each try with a new copy of flext via cvs. Using autoconfig I seem to get this error during build: flutil.cpp: In static member function ‘static void flext_single::CopyMem(void*, const void*, int)’: flutil.cpp:33: error: ‘BlockMoveData’ was not declared in this scope flutil.cpp: In static member function ‘static void flext_single::ZeroMem(void*, int)’: flutil.cpp:44: error: ‘BlockZero’ was not declared in this scope make[1]: *** [libflext_pd_s_a-flutil.o] Error 1 make: *** [install-recursive] Error 1 And using buildsys these errors: Chris-Jacks-Computer-2:flext chrisjack$ bash build.sh pd gcc make -f ./buildsys/gnumake-sub.mak PLATFORM=mac RTSYS=pd COMPILER=gcc BUILDPATH=./buildsys/ PKGINFO=package.txt BUILDCLASS=flext USRCONFIG=config.txt USRMAKE=build/gnumake-mac- gcc.inc TARGETMODE=release TARGETTYPE=single _build_ buildsys/mac/gnumake-gcc-targets.inc:22: warning: overriding commands for target `pd-darwin/release-single' buildsys/mac/gnumake-gcc-targets.inc:18: warning: ignoring old commands for target `pd-darwin/release-single' mkdir -p pd-darwin/release-single mkdir -p pd-darwin/release-single/ g++ -c -DNDEBUG -DFLEXT_EXPORTS -DFLEXT_SYS=2 -DPD -I/Applications/ Pd-extended.app/Contents/Resources//src source/flext.h -o pd-darwin/ release-single/flext.h.gch touch source/flbase.cpp mkdir -p ./ g++ -c -ffast-math -Os -ftree-vectorize -isysroot /Developer/SDKs/ MacOSX10.6.sdk -arch i386 -mmmx -msse -msse2 -msse3 - mtune=prescott -DNDEBUG -DFLEXT_EXPORTS -DFLEXT_SYS=2 -DPD -I pd- darwin/release-single -I/Applications/Pd-extended.app/Contents/ Resources//src source/flbase.cpp -o pd-darwin/release-single/ flbase.opp_i386 source/flbase.cpp:29:22: error: g_canvas.h: No such file or directory source/flbase.cpp: In static member function ‘static bool flext_obj_single::GetParamSym(t_atom, const t_symbol*, _glist*)’: source/flbase.cpp:130: error: ‘canvas_realizedollar’ was not declared in this scope source/flbase.cpp:133: error: ‘canvas_realizedollar’ was not declared in this scope make[1]: *** [pd-darwin/release-single/flbase.opp_i386] Error 1 make: *** [build-release-single] Error 2 I am pointing the variables in the txt files to the correct locations, of that I am sure. I have also begun with a fresh install of Pd-extended 0.43.4 - could this be it? Too new-a-build? Need I use Pd vanilla to get flext to play ball? I'd really rather not as I already have a bunch of stuff invested in Pd-extended and I want to avoid having to compile countless externals just to run the same projects in vanilla (though this would be a very long- winded option). Your helpings are highly appreciated. Cheers, CJ. -- Chris Jack art/research http://itsmrjack.com/ discus jockeying/sound art http://soundcloud.com/itsmrjack/ ___ Pd-list@iem.at mailing list
[PD] [PD-announce] pd 0.44-3 released
Hi All - I've released 0.44-3 (source, and compiled for Windows and Mac) - I don't have my Pi handy so will be updating the pi-compiled version later. The main update is to fix the gain of the hip~ filter, which I still had wrong. I was threatening to try also to fix largefile support for 32-bit linux installations but that ended up leading down a chain of changes that I think is too complicated to be able to do safely in a bugfix release... so I'll try to get that resolved for 0.45. Also for 0.45 (and probably my main project right now) is to expand the new text object into something that will combine the functionalities of qlist, textfile, and coll, and integrate texts into data structures. cheers Miller ___ Pd-announce mailing list pd-annou...@iem.at http://lists.puredata.info/listinfo/pd-announce ___ Pd-list@iem.at mailing list UNSUBSCRIBE and account-management - http://lists.puredata.info/listinfo/pd-list
Re: [PD] flext install (for another amateur)
Hey Thomas, Thanks for getting back to me on this. To be honest I thought I was being a bit stupid so am actually a little relieved that it wasn't *just* me - though of course it is unfortunate that it's not currently working with 0.43.4. I have actually had success compiling it for Pd vanilla 0.44 ...though it refused to put the files in Contents/Resources/extra - as I had instructed it to. However I was able to use it to build the GApop external successfully on first go. Thanks for your time, Chris. On 20 April 2013 19:57, Thomas Grill g...@g.org wrote: Hi Chris, i have now looked into the Pd-extended 0.43.4 package. The layout of files has been changed in a way that it is incompatible to the flext build system (and also to the autoconf version). I will have to see how i can fix this. What you can do for the time being is download Miller's vanilla version and compile your external object using that. You can then use the binary with pd-extended as well. all the best, gr~~~ -- Thomas Grill http://g.org Am 17.04.2013 um 13:40 schrieb Chris Jack: Hi Thomas, Thanks a lot for your help. Seeing as I have not actively installed gcc from any other source I suppose it must be the one from the xcode/developer sdk. Btw, my sdk is 'MacOSX10.6.sdk' - the one appropriate for osx 10.6.8, afaik! My Pd (extended 0.43.4) installation does actually contain the g_canvas.h file in /Applications/Pd-extended.app/Contents/Resources/include/pdextended I have the config-mac-gcc...etc file pointing PDPATH to /Applications/Pd-extended.app/Contents/Resources ...need I point a different variable to Resources/include/pdextended in order to see the header files? Cheers, On 16 April 2013 15:42, Thomas Grill g...@g.org wrote: Hi Chris, first of all, do you use the gcc coming with xcode or from some other source (fink, macports)? It seems that your platform SDK is somewhat strange, judging from the autoconfig build as it doesn't contain BlockMoveData and BlockZero. As for the buildsys build, it seems that your pd distribution (pd-extended?) doesn't contain the g_canvas.h file, which is a header file private to pd, but necessary for flext to work. Some pd distros do contain it, while others don't. Btw., there's a mailing list specialized to flext matters. all the best, Thomas -- Thomas Grill http://g.org Am 16.04.2013 um 15:59 schrieb Chris Jack: Hey folks, I have been trying in vain to install flext (cvs download) on my 10.6.8 snow leopard Mac. I've tried it every which way - both using the buildsys and autoconfig - and just cannot get the bugger to build. I have refreshed before each try with a new copy of flext via cvs. Using autoconfig I seem to get this error during build: flutil.cpp: In static member function ‘static void flext_single::CopyMem(void*, const void*, int)’: flutil.cpp:33: error: ‘BlockMoveData’ was not declared in this scope flutil.cpp: In static member function ‘static void flext_single::ZeroMem(void*, int)’: flutil.cpp:44: error: ‘BlockZero’ was not declared in this scope make[1]: *** [libflext_pd_s_a-flutil.o] Error 1 make: *** [install-recursive] Error 1 And using buildsys these errors: Chris-Jacks-Computer-2:flext chrisjack$ bash build.sh pd gcc make -f ./buildsys/gnumake-sub.mak PLATFORM=mac RTSYS=pd COMPILER=gcc BUILDPATH=./buildsys/ PKGINFO=package.txt BUILDCLASS=flext USRCONFIG=config.txt USRMAKE=build/gnumake-mac-gcc.inc TARGETMODE=release TARGETTYPE=single _build_ buildsys/mac/gnumake-gcc-targets.inc:22: warning: overriding commands for target `pd-darwin/release-single' buildsys/mac/gnumake-gcc-targets.inc:18: warning: ignoring old commands for target `pd-darwin/release-single' mkdir -p pd-darwin/release-single mkdir -p pd-darwin/release-single/ g++ -c -DNDEBUG -DFLEXT_EXPORTS -DFLEXT_SYS=2 -DPD -I/Applications/Pd-extended.app/Contents/Resources//src source/flext.h -o pd-darwin/release-single/flext.h.gch touch source/flbase.cpp mkdir -p ./ g++ -c -ffast-math -Os -ftree-vectorize -isysroot /Developer/SDKs/MacOSX10.6.sdk -arch i386 -mmmx -msse -msse2 -msse3 -mtune=prescott -DNDEBUG -DFLEXT_EXPORTS -DFLEXT_SYS=2 -DPD -I pd-darwin/release-single -I/Applications/Pd-extended.app/Contents/Resources//src source/flbase.cpp -o pd-darwin/release-single/flbase.opp_i386 source/flbase.cpp:29:22: error: g_canvas.h: No such file or directory source/flbase.cpp: In static member function ‘static bool flext_obj_single::GetParamSym(t_atom, const t_symbol*, _glist*)’: source/flbase.cpp:130: error: ‘canvas_realizedollar’ was not declared in this scope source/flbase.cpp:133: error: ‘canvas_realizedollar’ was not declared in this scope make[1]: *** [pd-darwin/release-single/flbase.opp_i386] Error 1 make: *** [build-release-single] Error 2 I am pointing the variables in the txt files to the correct locations, of that I am sure. I have also begun with a fresh install of Pd-extended 0.43.4
Re: [PD] Sensors GPIO Raspberry Pi Pd
Oh sorry, it segfaults if you don't pass an argument at startup. (1 if it's already initialized, 0 if not) The line begining if (argc 0) should say if (argc 1). Martin On 2013-04-20 13:30, Martin Peach wrote: Well that looks a total mess... I did sudo apt-get install i2c-dev before all this, as well as all the things in http://www.instructables.com/id/Raspberry-Pi-I2C-Python/#step1 except I didn't use any python. The segfaults might be because you are asking for more than 32 bytes in a read, or something wrong in the code, the listing below looks like it goes wrong right at the beginning. Martin On 2013-04-19 17:37, Julian Brooks wrote: As I'm new to all this C stuff I just had a look inside the 'hello' file and there's a few bits in there which may be of interest: ^@D6T_checkPEC says 0x%02X (which I think is ok from looking at the code?) then it all goes wrong ^@^@^@Unable to create socket (%d) ^@^@^@ %s ^@^@^@127.0.0.1 http://127.0.0.1^@^@^@Unable to make address from %s ^@%s: initialized:%d ^@/dev/i2c-0^@^@Failed to open the bus. (%d) ^@^@^@Failed to acquire bus access and/or talk to slave. (%d) ^@^@^@^@Write failed (%d) ^@^@Read failed (%d) ^@^@^@%d ^@0x%02X ^@^@^@d6t8l^@^@^@ %d^@ ^@^@^@sendto error (%d) BTW - I've removed libi2c-dev again and will leave it that way for now. Going to stop now but will be back tomorrow. Cheers, Julian On 19 April 2013 22:07, Julian Brooks jbee...@gmail.com mailto:jbee...@gmail.com wrote: Edited your original file and all I changed was the ip address to 127.0.0.1 and still seg fault (double checked /etc/hosts too). Also tried reinstalling libi2c-dev and then running 'hello' - same. On 19 April 2013 21:54, Julian Brooks jbee...@gmail.com mailto:jbee...@gmail.com wrote: hmmm - both files: ./hello-2 Segmentation fault Try again... On 19 April 2013 21:51, Julian Brooks jbee...@gmail.com mailto:jbee...@gmail.com wrote: Ok -found I had to remove 'libi2c-dev'. Then builds. More soon... On 19 April 2013 21:28, Julian Brooks jbee...@gmail.com mailto:jbee...@gmail.com wrote: Hey Martin, When I try to compile hello.c I get: gcc -o hello hello.c In file included from hello.c:8:0: /usr/include/linux/i2c-dev.h:38:8: error: redefinition of 'struct i2c_msg' /usr/include/linux/i2c.h:67:8: note: originally defined here /usr/include/linux/i2c-dev.h:90:7: error: redefinition of 'union i2c_smbus_data' /usr/include/linux/i2c.h:125:7: note: originally defined here Dunno if this is at all relevant but maybe this is a good time to say I have a rev1 RPi so I'm on i2c 0 not 1. The Pi install is very up to date though, including recent runs of hexxeh's 'rpi-update' tool so all the system's fresh. I did attempt to change the number of bytes to be read which perhaps would explain why the .h file's are complaining but I don't understand it for your version which is 'as-is'? In fact why don't I attach the notes I've made of the changes to the c file you sent. Was vaguely hoping I might be able to say 'ta da' but have fallen at the 1st fence:( bugger. Julian On 19 April 2013 19:51, Julian Brooks jbee...@gmail.com mailto:jbee...@gmail.com wrote: Hi Martin, Meant to add re setting baud rate: I've been making use of the gpio utility that comes with wiringPi https://projects.drogon.net/raspberry-pi/wiringpi/the-gpio-utility/ Very handy for getting a quick visualisation of the current state of all the pins and also easy-access to setting the baud rate too (amongst other stuff). Julian On 19 April 2013 14:36, Martin Peach martin.pe...@sympatico.ca mailto:martin.pe...@sympatico.ca wrote: Hi Julian, Yes I've been messing with coding it in c on the pi and sending the data to a [netreceive] in a Pd patch on another machine. I'm attaching the source code for the pi part and the Pd patch. The code can be compiled on the pi with gcc -o hello hello.c You need to set the IP address of the receiving machine in the code (I have 192.168.2.15, it could be 127.0.0.1
Re: [PD] Sensors GPIO Raspberry Pi Pd
So I tested the 4X4 sensor and it actually works! Here is the code for a reader that sends to [netreceive 3], and a receiver patch. You need to set the IP to that of the machine running Pd, and maybe other settings before compiling, as with the previous version. Martin On 2013-04-20 17:17, Martin Peach wrote: Oh sorry, it segfaults if you don't pass an argument at startup. (1 if it's already initialized, 0 if not) The line begining if (argc 0) should say if (argc 1). Martin On 2013-04-20 13:30, Martin Peach wrote: Well that looks a total mess... I did sudo apt-get install i2c-dev before all this, as well as all the things in http://www.instructables.com/id/Raspberry-Pi-I2C-Python/#step1 except I didn't use any python. The segfaults might be because you are asking for more than 32 bytes in a read, or something wrong in the code, the listing below looks like it goes wrong right at the beginning. Martin On 2013-04-19 17:37, Julian Brooks wrote: As I'm new to all this C stuff I just had a look inside the 'hello' file and there's a few bits in there which may be of interest: ^@D6T_checkPEC says 0x%02X (which I think is ok from looking at the code?) then it all goes wrong ^@^@^@Unable to create socket (%d) ^@^@^@ %s ^@^@^@127.0.0.1 http://127.0.0.1^@^@^@Unable to make address from %s ^@%s: initialized:%d ^@/dev/i2c-0^@^@Failed to open the bus. (%d) ^@^@^@Failed to acquire bus access and/or talk to slave. (%d) ^@^@^@^@Write failed (%d) ^@^@Read failed (%d) ^@^@^@%d ^@0x%02X ^@^@^@d6t8l^@^@^@ %d^@ ^@^@^@sendto error (%d) BTW - I've removed libi2c-dev again and will leave it that way for now. Going to stop now but will be back tomorrow. Cheers, Julian On 19 April 2013 22:07, Julian Brooks jbee...@gmail.com mailto:jbee...@gmail.com wrote: Edited your original file and all I changed was the ip address to 127.0.0.1 and still seg fault (double checked /etc/hosts too). Also tried reinstalling libi2c-dev and then running 'hello' - same. On 19 April 2013 21:54, Julian Brooks jbee...@gmail.com mailto:jbee...@gmail.com wrote: hmmm - both files: ./hello-2 Segmentation fault Try again... On 19 April 2013 21:51, Julian Brooks jbee...@gmail.com mailto:jbee...@gmail.com wrote: Ok -found I had to remove 'libi2c-dev'. Then builds. More soon... On 19 April 2013 21:28, Julian Brooks jbee...@gmail.com mailto:jbee...@gmail.com wrote: Hey Martin, When I try to compile hello.c I get: gcc -o hello hello.c In file included from hello.c:8:0: /usr/include/linux/i2c-dev.h:38:8: error: redefinition of 'struct i2c_msg' /usr/include/linux/i2c.h:67:8: note: originally defined here /usr/include/linux/i2c-dev.h:90:7: error: redefinition of 'union i2c_smbus_data' /usr/include/linux/i2c.h:125:7: note: originally defined here Dunno if this is at all relevant but maybe this is a good time to say I have a rev1 RPi so I'm on i2c 0 not 1. The Pi install is very up to date though, including recent runs of hexxeh's 'rpi-update' tool so all the system's fresh. I did attempt to change the number of bytes to be read which perhaps would explain why the .h file's are complaining but I don't understand it for your version which is 'as-is'? In fact why don't I attach the notes I've made of the changes to the c file you sent. Was vaguely hoping I might be able to say 'ta da' but have fallen at the 1st fence:( bugger. Julian On 19 April 2013 19:51, Julian Brooks jbee...@gmail.com mailto:jbee...@gmail.com wrote: Hi Martin, Meant to add re setting baud rate: I've been making use of the gpio utility that comes with wiringPi https://projects.drogon.net/raspberry-pi/wiringpi/the-gpio-utility/ Very handy for getting a quick visualisation of the current state of all the pins and also easy-access to setting the baud rate too (amongst other stuff). Julian On 19 April 2013 14:36, Martin Peach martin.pe...@sympatico.ca mailto:martin.pe...@sympatico.ca wrote: Hi Julian, Yes I've been messing with coding it in c on the pi and sending the data to a [netreceive] in a Pd patch on another machine. I'm attaching the source code for the
Re: [PD] Sensors GPIO Raspberry Pi Pd
On 2013-04-20 21:09, Julian Brooks wrote: Oh and btw Still don't know why I can't compile the .c files on the pi with libi2c-dev installed but I can't. Presuming the compiling is working for you Martin? Yes it works for me. I don't have the same /usr/include/linux/i2c-dev.h as you so no redefinition errors, not sure which package(s) install that file. Martin ___ Pd-list@iem.at mailing list UNSUBSCRIBE and account-management - http://lists.puredata.info/listinfo/pd-list
