Re: [PD] symbolarray
Hallo, Jonathan Wilkes hat gesagt: // Jonathan Wilkes wrote: > Does [drawpolygon] affect creation time with that many elements? Most likely not: None of the symbols is drawn. The only thing that's drawn is a static polygon to let you click somewhere to look at the data structure in the builtin editor. I found that my [drawpolygon] object even is specified incorrectly: I forgot to add the line width. :) You could even remove the polygon without affecting functionality. [symbolarray] may illustrate how useful (and speedy) Pd's data structures can be without graphical interface. Ciao -- Frank ___ Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
Re: [PD] My first external: [fsm] finite state machine for pd
Hi Bryan, thanks for the feedback. Bridging between fsm and gfsm is a great idea! I thought about dot export and expressions too, but why should i reinvent the wheel? :) What i'm actually planning to add, is a pd-patch exporter, to export state machines as patches, to be even more portable. (it happens regularly, that downloaded patches refuse to work, because of dependencies, so i think it's in many cases a good idea, to avoid the usage of externals). But first of all, i have to get fsm stable. Currently it crashes, when i delete a state (didn't happen in earlier versions :( )... need to figure out, how to connect externals to a debugger & memory watcher... Have a nice sunday, lsw~ Am 19.04.2009, 11:46 Uhr, schrieb Bryan Jurish : moin lsw, pretty cool, especially since it has next to no dependencies (unlike [gfsm], which needs glib and of course libgfsm ;-) Any chance of adding some AT&T-style I/O routines for compatibility? That way, users could easily switch back and forth between [gfsm_automaton] and [fsm] representations by bouncing the data over the filesystem (ugly but portable), so you could get visualization with dot, regex compilation, markov chain generation, weighted transitions, etc. etc. ... marmosets, Bryan On 2009-04-15 23:59:32, lsw appears to have written: Dear list, i have been absent from pd-list for a few years, but i'd like to change that. :) During the last days i wrote my first external: fsm - finite state machine for pd I wrote this with algorithmic sequencing and sequence recognition in mind. You can find builds for win32, osx and linx, as well as the sources and help-patch with examples at http://floppy35.de/pd/ I didn't test the windows and linux builds (but osx-version seems to work fine), so please tell me, if it works or if it destroys your machine and deletes your mp3-collection. ;) Unfortunately i saw a bit late that moocow had a similar idea earlier... however, my external seems to be a bit different (low-level approach), so it might be still useful for some of you. It's written in plain C and consists of one single c-file, so it should be easy to port and compile for different platforms. Feedback is appreciated. All the best, lsw~ -- ~ ___ Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
Re: [PD] gem list operations on the gpu
Sebastian Saa wrote: Hello, is it possible to run gemlist operations on the gpu? im experimenting with recursive structures in gem , i cannot have to much recursions because my computer become slow, so i was thinking if its posible to make those calculations on the gpu, is this possible? hi sebastian, it depends on what exactly you're trying to do. it is possible to calculate a recursive structure once and then upload it to the GPU using a displaylist. basically you are creating a geometry that you can access like a model. have a look at the example in examples/GEM/09.OpenGL/02.displaylist.pd But this will not help you, if the recursion needs to be recalculated for every frame. for this you could use vertex shaders (again, depending on what you're trying to achieve). with the upcoming GEM release and if your gfx card supports it, it should also be possible to use geometry shaders as well. the other options are pdlua (with luagl) or writing your own GEM object... marius. ___ Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
Re: [PD] My first external: [fsm] finite state machine for pd
moin lsw, pretty cool, especially since it has next to no dependencies (unlike [gfsm], which needs glib and of course libgfsm ;-) Any chance of adding some AT&T-style I/O routines for compatibility? That way, users could easily switch back and forth between [gfsm_automaton] and [fsm] representations by bouncing the data over the filesystem (ugly but portable), so you could get visualization with dot, regex compilation, markov chain generation, weighted transitions, etc. etc. ... marmosets, Bryan On 2009-04-15 23:59:32, lsw appears to have written: > Dear list, > > i have been absent from pd-list for a few years, but i'd like to change > that. :) > > During the last days i wrote my first external: fsm - finite state > machine for pd > I wrote this with algorithmic sequencing and sequence recognition in mind. > You can find builds for win32, osx and linx, as well as the sources and > help-patch with > examples at http://floppy35.de/pd/ > I didn't test the windows and linux builds (but osx-version seems to > work fine), > so please tell me, if it works or if it destroys your machine and > deletes your > mp3-collection. ;) > > Unfortunately i saw a bit late that moocow had a similar idea earlier... > however, my > external seems to be a bit different (low-level approach), so it might > be still useful > for some of you. It's written in plain C and consists of one single > c-file, so it should > be easy to port and compile for different platforms. > > Feedback is appreciated. > > All the best, > lsw~ -- Bryan Jurish "There is *always* one more bug." jur...@ling.uni-potsdam.de -Lubarsky's Law of Cybernetic Entomology ___ Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
