Pd-dev April 2005

pd-dev@lists.iem.at

22 participants
53 discussions

paths in flext buildsys
by Frank Barknecht 16 Apr '05

16 Apr '05

Hi, I started to convert my externals to the flext buildsystem and ran into a problem with fluid. The sources have this layout: TOPDIR=/cvs/.../footils/fluid cpp source: TOPDIR/fluid/main.cpp package.txt: TOPDIR/package.txt package.txt: NAME=fluid~ SRCS=fluid/main.cpp LIBS=-lfluidsynth Now I run into an error while compiling because not all directories get auto-created: $ flext-build.sh pd gcc build make -f /usr/lib/flext/buildsys/gnumake-sub.mak PLATFORM=lnx RTSYS=pd COMPILER=gcc BUILDPATH=/usr/lib/flext/buildsys/ PKGINFO=package.txt BUILDCLASS=ext TARGETMODE=release TARGETTYPE=single _build_ make[1]: Entering directory `/home/freak/big/audio/pd/cvs-new/externals/footils/fluid' g++ -c -ffast-math -DNDEBUG -O3 -march=athlon -pthread -fPIC -DFLEXT_SYS=2 -I../../../pd/src -I/usr/include/flext fluid/main.cpp -o pd-linux/release-single/fluid/main.opp Assembler messages: FATAL: can't create pd-linux/release-single/fluid/main.opp: No such file or directory make[1]: *** [pd-linux/release-single/fluid/main.opp] Error 1 make[1]: Leaving directory `/home/freak/big/audio/pd/cvs-new/externals/footils/fluid' make: *** [build-release-single] Error 2 The directories "pd-linux/release-single/" get created automatically, hoever not the final one, "pd-linux/release-single/fluid/". Now what should I do: Create the full directory-tree for all build files? Or is there a way to tell the buildsys to create the directories itself as needed? Ciao -- Frank Barknecht _ ______footils.org__ _ __latest track: "scans" _ http://footils.org/cms/show/41

2 2

flext build system location
by Thomas Grill 13 Apr '05

13 Apr '05

Hi all, approaching a release of a new flext version including the build system i ask myself where the flext build system should be installed on unix systems. It contains of a hierarchy of fixed script and make files and of some user-editable configuration files. Is it a good idea to install that in /usr[/local]/lib/flext, or should this rather be in */share/flext or /usr/local/lib/pd/flext (which is not existing on OSX) ? Meanwhile the header files have an established position in /usr/local/include/flext and the libraries in /usr/local/lib. thanks, Thomas

2 1

patch to ease the compilation of gem2pdp
by IOhannes m zmoelnig 13 Apr '05

13 Apr '05

hi. i was having a hard time to compile gem2pdp (forgot to mention that in my last post) on linux my g++ (3.3.5 (Debian 1:3.3.5-8)) refuses to cast (t_int*) to (int*) i therefore had to change the type of m_packet0 in gem2pdp.h to "int". furthermore i find it very annoying that i have to set the path to my gem and pdp installation by hand in configure.ac since i prefer command-line flags i have added 3 of them (--with-gemdir=<gempath> --with-pdpdir=<pdppath> --with-pddir=<pdpath>) to the configure.ac, which allows me to overrule the hardcoded paths attached is a patch against the CVS. hope this works mfg.asdr IOhannes

1 0

Re: [PD] building all external on linux
by IOhannes m zmoelnig 13 Apr '05

13 Apr '05

hi gÃŒnter geiger wrote: > On Tue, 12 Apr 2005, IOhannes m zmoelnig wrote: > >>what makes structuring code as libraries vs. singular externals a "key >>to collaboration" ? > .... > > Therefore I think it would be better to have single externals in all > cases where there is no shared code. I understand that this point of > view is not shared by the other developers, I just based my decisions > 2 and a half years ago on that, and that is how I explained it. most of what you say about externals vs libraries is certainly true (see below). i wasn't arguing about the pros and cons of your build-system which certainly has its merits. what i didn't like is the attitude of saying more or less "by using libraries you hinder collaboration", implying that developers doing so are second class coders and a disgrace to the entire free software community. > > I think this enabled a very simple, robust and easy to maintain build > system. yes it's really nice (though somewhat "interesting") > > I think it just makes it easier. > + to avoid nameclashes that one again... i mean it doesn't really avoid nameclashes, it just shifts the decision away to some supervisor who decides which library's object will get included; and most often this decision will be based (note: i am wildly assuming things here) on bias (because the one who has written the external adds it to the built system), on time (when did the supervisor first noted that a certain object exists and it could be added to the built system; this does not necessary mean that it really was the first (publically available (e.g. sourceforge-cvs)) object with this name), on discrimination (objects that are (for whatever reasons) part of a library will not get included; if someone later on decides to write a object with the same name as an external it will get included) > + to have a common build system > + to make a list of existing externals > + to make the build system resistant to errors in single externals > + to come up with a common consistent set of externals > - make it hard to share code true > > wheras libraries > - make it hard to find out which externals they contain > - obscure nameclashes > - fail to build if one single external doesnt build true > - have to be loaded explicitly by the user i don't think that this is necessarily a bad thing. i rather prefer a mail client that does not enable every possible feature (with bells and whistles and everything) by default. there are objects in the current built system which have severe bugs that make pd a crashy experience. often i do not have the time, permission or knowledge to solve such bugs. > + make it possible to share code between different objects mfg.asd.r IOhannes

1 0

Re: [PD] building all external on linux
by Frank Barknecht 13 Apr '05

13 Apr '05

Hallo, IOhannes m zmoelnig hat gesagt: // IOhannes m zmoelnig wrote: > imo, it is bad by design to use a local (dead) copy of a header-file > when there is living one next door. I believe this is a leftover from the time, when there wasn't an "official" header in the CVS, because "devel" was HEAD. It should be changed, shouldn't it? And this topic maybe should go to pd-dev as well. Ciao -- Frank Barknecht _ ______footils.org__ _ __latest track: "scans" _ http://footils.org/cms/show/41

3 4

[ pure-data-Patches-1083495 ] s_loader.c: osx external loading
by SourceForge.net 13 Apr '05

13 Apr '05

Patches item #1083495, was opened at 2004-12-11 11:25 Message generated for change (Comment added) made by eighthave You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=478072&aid=1083495&group_… Category: puredata Group: bugfix Status: Open Resolution: None Priority: 5 Submitted By: Jamie Tittle (tigital) Assigned to: Nobody/Anonymous (nobody) Summary: s_loader.c: osx external loading Initial Comment: ...years ago, the NSLINKMODULE_OPTION_PRIVATE flag was introduced to NSLinkModule() in an effort to prevent pd from crashing when a library is loaded that has duplicate symbols to an already loaded library...sure, it's desirable to not crash, but this flag has the side effect of artificially partitioning the loaded symbols such that we can't access one library easily from another... ...so, this patch removes the PRIVATE flag and adds another RETURN_ON_ERROR flag, which works as advertised...I've also added a small bit of code to spit out some info about what the error was, and in which file... ...another point to consider is the linker flag -flat_namespace, and whether it is needed at all anymore (the default would be -twolevel_namespace)... ---------------------------------------------------------------------- >Comment By: Hans-Christoph Steiner (eighthave) Date: 2005-04-13 00:03 Message: Logged In: YES user_id=27104 Check out externals/build/darwin/makefile. Its been using -bundle -bundle_loader ../path/to/pd successfully for a while now. externals/OSCx is using -bundle_loader.. but has -flat_namespace, so maybe that should be changed as well. ---------------------------------------------------------------------- Comment By: Jamie Tittle (tigital) Date: 2005-04-12 13:56 Message: Logged In: YES user_id=634044 ...after further exploration with linking and loading other externals, I think we can leave -flat_namespace behind and just use something like the following: PREFIX = $(HOME)/pd-0.38-4/ LFLAGS = -bundle -bundle_loader $(PREFIX)/bin/pd ...notice that we don't need to say "-two_level_namespace", because it is default...without the -bundle_loader bit, you get all kinds of problems with undefined symbols... ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=478072&aid=1083495&group_…

1 0

New flext: pool crashes
by Frank Barknecht 13 Apr '05

13 Apr '05

Hi, I experience a strange crash with current [pool], checked out some minutes ago and flext checked out then as well. I'm using Miller's Pd 0.38 HEAD from CVS, that is, I'm not using the "devel_X" branch. _The crash_ If I create a simple patch: [set x y( | [pool] and if I then bang [set x y( *twice* Pd immediatly crashes. Under gdb, this is in the backtrace: pool 0.2.1pre - hierarchical storage object, (C)2002-2005 Thomas Grill Program received signal SIGSEGV, Segmentation fault. [Switching to Thread -1209710912 (LWP 13024)] 0xb7636105 in poolval::Set () from /usr/lib/pd/extra/pool.pd_linux (gdb) bt #0 0xb7636105 in poolval::Set () from /usr/lib/pd/extra/pool.pd_linux #1 0xb7636a3c in pooldir::SetVal () from /usr/lib/pd/extra/pool.pd_linux #2 0xb7630153 in pool::set () from /usr/lib/pd/extra/pool.pd_linux #3 0xb7633b15 in pool::flext_c_m_set () from /usr/lib/pd/extra/pool.pd_linux #4 0xb76bf518 in flext_base_single::TryMethTag () from /usr/lib/pd/extra/xsample.pd_linux #5 0xb76bf6d9 in flext_base_single::FindMeth () from /usr/lib/pd/extra/xsample.pd_linux #6 0xb76424de in flext_base_single::CbMethodHandler () from /usr/lib/pd/extra/pool.pd_linux #7 0xb76bfbcf in flext_base_single::cb_px_anything () from /usr/lib/pd/extra/xsample.pd_linux #8 0x0809b756 in pd_typedmess (x=0x82613d0, s=0x1, argc=2, argv=0x80f2b30) at m_class.c:749 #9 0x0809d037 in outlet_anything (x=0x8261f14, s=0x8203b88, argc=2, argv=0x80f2b30) at m_obj.c:380 #10 0x0809b756 in pd_typedmess (x=0x82614dc, s=0x1, argc=2, argv=0x80f2b30) at m_class.c:749 #11 0x0809ebc4 in binbuf_eval (x=0x8263024, target=0x82614dc, argc=1, argv=0xbfffd698) at m_binbuf.c:574 #12 0x08061dec in text_click (z=0x82614c0, glist=0x825ed30, xpix=235, ypix=136, shift=136713436, alt=-1073752424, dbl=0, doit=0) at g_text.c:300 #13 0x08079745 in canvas_doclick (x=0x825ed30, xpos=235, ypos=136, which=1, mod=0, doit=1) at g_editor.c:742 #14 0x0807ddeb in canvas_mousedown (x=0x8263024, xpos=-7.73854845e-06, ypos=-7.73854845e-06, which=-7.73854845e-06, mod=-7.73854845e-06) at g_editor.c:1172 #15 0x0809b88d in pd_typedmess (x=0x825ed30, s=0x1, argc=136703280, argv=0x436b0000) at m_class.c:728 #16 0x0809b756 in pd_typedmess (x=0x82148d0, s=0x1, argc=4, argv=0x80f2b08) at m_class.c:749 #17 0x0809ebc4 in binbuf_eval (x=0x8263024, target=0x82148d0, argc=0, argv=0x0) at m_binbuf.c:574 #18 0x080a6fc0 in socketreceiver_read (x=0x82137f8, fd=0) at s_inter.c:509 #19 0x080a8b5f in sys_domicrosleep (microsec=136720420, pollem=1) at s_inter.c:169 #20 0x080a879e in sys_pollgui () at s_inter.c:787 #21 0x080a3828 in m_scheduler () at m_sched.c:456 #22 0x080ad6db in main (argc=136720420, argv=0x8263024) at s_entry.c:27 (gdb) Does it ring a bell with someone? Ciao -- Frank Barknecht _ ______footils.org__ _ __latest track: "scans" _ http://footils.org/cms/show/41

2 5

[ pure-data-Patches-1083495 ] s_loader.c: osx external loading
by SourceForge.net 12 Apr '05

12 Apr '05

Patches item #1083495, was opened at 2004-12-11 11:25 Message generated for change (Comment added) made by tigital You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=478072&aid=1083495&group_… Category: puredata Group: bugfix Status: Open Resolution: None Priority: 5 Submitted By: Jamie Tittle (tigital) Assigned to: Nobody/Anonymous (nobody) Summary: s_loader.c: osx external loading Initial Comment: ...years ago, the NSLINKMODULE_OPTION_PRIVATE flag was introduced to NSLinkModule() in an effort to prevent pd from crashing when a library is loaded that has duplicate symbols to an already loaded library...sure, it's desirable to not crash, but this flag has the side effect of artificially partitioning the loaded symbols such that we can't access one library easily from another... ...so, this patch removes the PRIVATE flag and adds another RETURN_ON_ERROR flag, which works as advertised...I've also added a small bit of code to spit out some info about what the error was, and in which file... ...another point to consider is the linker flag -flat_namespace, and whether it is needed at all anymore (the default would be -twolevel_namespace)... ---------------------------------------------------------------------- >Comment By: Jamie Tittle (tigital) Date: 2005-04-12 12:56 Message: Logged In: YES user_id=634044 ...after further exploration with linking and loading other externals, I think we can leave -flat_namespace behind and just use something like the following: PREFIX = $(HOME)/pd-0.38-4/ LFLAGS = -bundle -bundle_loader $(PREFIX)/bin/pd ...notice that we don't need to say "-two_level_namespace", because it is default...without the -bundle_loader bit, you get all kinds of problems with undefined symbols... ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=478072&aid=1083495&group_…

1 0

name-battle round 2: SCons
by ix＠replic.net 12 Apr '05

12 Apr '05

so, these weird recursion errors SCons was driving me crazy over, are beacuse there are multiple C files with the same name.. for example maxlib/urn.c and cyclone/hammer/urn.c.. basically you cant do that, so if you want to try out the new externals build system, cd build/ and type 'scons' and then 'scons install' if it all went well, but it definitely wont unless you: rm src/urn.* rm src/match.* rm src/prepend.* rm src/sinh.c rm src/delta~.c rm src/reson~.c rm src/abs~.c rm src/svf~.c rm src/matrix~.c rm src/atan2~.c ..scripting the removal of these files in scons would be hackish, even by my standards... c

2 1

help with first gui-external
by Robert Gründler 10 Apr '05

10 Apr '05

hi list, i've started learning C a few weeks ago. At the moment, i'm trying to write a waveform editor for pd, and up till now, i'm at a point where i'd need some advice from more experienced developers concerning my code. The code is mainly based on B. Bogarts „popup“, the hslider- and the garray objects. Here are my questions: 1. I'd like to output the selection start/end points as milliseconds. How can i get the samplerate of a soundfile? Right now, i'm simply using pd's system-samplerate for samples-to-milliseconds conversion . 2. When i change the appearance of the mouse cursor over a tk-item using <Enter> and <Leave>, how can i get back to use the normal pd mouse cursors after leaving again? (it stays in the appearance i use in the <Leave> command) I think i have to make the <Leave> command use the current pd-mouse- state somehow. 3. What do i need to concern about when writing gui-objects for Graph- on-Parent use? At the present state, pd crashes when i try to open a subpatcher including my object as GOP (Memory Access Violation). Also i'd be very grateful if anyone could have a look at the code and tell me, where it can be optimized, cause i'm not quite sure if i didn't express some parts pretty complicated. source is here-> http://rcjackson.members.wekay.at/download/ Thanks for your help Robert here's a helpfile: #N canvas 130 89 1024 524 10; #X obj 710 239 soundfiler; #X obj 241 218 waver sepp; #X floatatom 241 409 5 0 0 0 - - -; #X floatatom 534 408 5 0 0 0 - - -; #X text 597 408 ms; #X msg 241 152 mode \$1; #X msg 154 63 1; #X msg 258 66 2; #X msg 301 66 3; #X floatatom 387 161 5 0 0 0 - - -; #X floatatom 534 166 5 0 0 0 - - -; #X obj 619 189 table foo; #X obj 618 213 table bar; #X msg 21 152 set foo; #X msg 102 154 set bar; #X obj 710 157 openpanel; #X obj 710 121 bng 15 250 50 0 empty empty empty 0 -6 0 8 -262144 -1 -1; #X msg 710 188 read -resize \$1 foo; #X msg 864 189 read -resize \$1 bar; #X obj 862 133 bng 15 250 50 0 empty empty empty 0 -6 0 8 -262144 -1 -1; #X obj 864 162 openpanel; #X text 330 67 zoom; #X text 252 45 select; #X text 114 13 move selection; #X text 112 27 (shift-drag changes loop size); #X text 362 141 selection start; #X text 498 143 selection end; #X text 125 409 selection start; #X text 431 408 selection end; #X text 283 407 ms; #X connect 1 0 2 0; #X connect 1 1 3 0; #X connect 5 0 1 0; #X connect 6 0 5 0; #X connect 7 0 5 0; #X connect 8 0 5 0; #X connect 9 0 1 1; #X connect 10 0 1 2; #X connect 13 0 1 0; #X connect 14 0 1 0; #X connect 15 0 17 0; #X connect 16 0 15 0; #X connect 17 0 0 0; #X connect 18 0 0 0; #X connect 19 0 20 0; #X connect 20 0 18 0;

2 1

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

Pd-dev April 2005