Pd-dev December 2012

pd-dev@lists.iem.at

14 participants
173 discussions

Re: [PD-dev] [PD] [PD-announce] pd 0.43-3 released
by IOhannes m zmoelnig 09 Dec '12

09 Dec '12

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 2012-07-03 23:11, Miller Puckette wrote: > Hi all, > > Pd version 0.43-3 is available on > http://crca.ucsd.edu/~msp/software.htm cool. > > I'm ready to start hacking on 0.44. The most urgent thing seems to > be for me to go back and work on the audio system (and perhaps > tweak MIDI a bit more too). i have a project lying on my hardisk for >1 year (iirc, shortly before the 0.32.0 release), that is related to that: separating the audio system(s) from Pd-core. background: currently Pd supports a number of audio-backends. while these backends are mainly separated into several source files (e.g. s_audio_alsa.c for ALSA), there is a lot of theoretically backend-agnostic code (s_audio, s_midi, s_main) cluttered with "#ifdef USEAPI_ALSA" and the like. this makes the current code badly readable and thus maintainable, it also makes it hard to add new audio systems as backends (and the more you add, the more complicated the codebase gets). more background: currently all audio backends are linked to the main "Pd" binary. if your version of Pd is compiled for alsa, jack, OSS and portaudio (the default on debian), you have quite a number of dependencies. (as maintainer of the "puredata" debian package i prefer to keep necessary dependencies at a minimum). furthermore, Pd supports outdated/deprecated backends like OSS (both audio and midi). in practice i haven't used OSS-midi for a number of years (and i doubt whether it is actually still supported by the mainline kernels), and i used OSS-audio only seldomly in very specific setups. from this i follow, that the majority of people have to fight their way through a lot of dead options that are more confusing than helpful, and which should probably be removed in most cases. suggestion: so what i ended up doing, was separate the audio-backends from the pd-core code a bit more, in a similar way to how Pd's object registration works. the means that people could write new audio-backends and load them on-demand, just like "externals". it also means that Pd's non-audiosystem codebase has only a minimum of "#ifdef USEAPI_ALSA" (there's only one, when it comes to loading the "internal" audio-system at startup). all this also applies to MIDI. if there's interest, i could cleanup the patches (and make them work again with 0.43.3) and submit them for review. if there's confusion, i could try to try again and explain what i want. fgmasdr IOhannes -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.12 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iEYEARECAAYFAk/z9ewACgkQkX2Xpv6ydvRvMwCgv2NPmX2Hx9wZSubBHA3mKpRd IrkAoJbWuNJF1Gu2X7Vg1T+f1NHUdBbi =fPn6 -----END PGP SIGNATURE-----

5 12

[ pure-data-Patches-3591846 ] portaudio fixes
by SourceForge.net 08 Dec '12

08 Dec '12

Patches item #3591846, was opened at 2012-12-02 08:56 Message generated for change (Comment added) made by zmoelnig You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=478072&aid=3591846&group_… Please note that this message will contain a full copy of the comment thread, including the initial issue submission, for this request, not just the latest update. Category: puredata Group: bugfix >Status: Closed >Resolution: Accepted Priority: 9 Private: No Submitted By: IOhannes m zmölnig (zmoelnig) Assigned to: Miller Puckette (millerpuckette) Summary: portaudio fixes Initial Comment: this patches fixes the current portaudio implementation. since it is too big to be added to the patch-tracker (i imported portaudio-20121129), you can download it from http://iem.at/~zmoelnig/0001-updated-to-portaudio-20121129-snapshot.patch.b… ---------------------------------------------------------------------- >Comment By: IOhannes m zmölnig (zmoelnig) Date: 2012-12-08 07:54 Message: a different patchset (based on the 20121031 pa-snapshot has made it into Pd-0.44) ---------------------------------------------------------------------- Comment By: Miller Puckette (millerpuckette) Date: 2012-12-03 10:46 Message: ... in case this makes it easier, here's the snapshot I last used: http://crca.ucsd.edu/~msp/tmp/pa_snapshot_20121031.tgz if you try diff -r --brief between this and Pd's copy in src/portaudoi you can see how I pruned the portaudio tree to include only stuff relevant to Pd. Oddly, I had to supply one extra file though, pd/portaudio/src/os/mac_osx/pa_mac_hostapis.c ---------------------------------------------------------------------- Comment By: Miller Puckette (millerpuckette) Date: 2012-12-03 10:05 Message: I tried this and got: [msp@fuzz pd]$ git apply ~/bis/var/download/0001-updated-to-portaudio-20121129-snapshot.patch /home/msp/bis/var/download/0001-updated-to-portaudio-20121129-snapshot.patch:287: trailing whitespace. /home/msp/bis/var/download/0001-updated-to-portaudio-20121129-snapshot.patch:300: trailing whitespace. /home/msp/bis/var/download/0001-updated-to-portaudio-20121129-snapshot.patch:336: trailing whitespace. * The text above constitutes the entire PortAudio license; however, /home/msp/bis/var/download/0001-updated-to-portaudio-20121129-snapshot.patch:341: trailing whitespace. * they can be incorporated into the canonical version. It is also /home/msp/bis/var/download/0001-updated-to-portaudio-20121129-snapshot.patch:342: trailing whitespace. * requested that these non-binding requests be included along with the warning: squelched 32389 whitespace errors warning: 32394 lines add whitespace errors. ... is there any way you can supply a patch that doens't add trailing white space to 1/2 ot the Pd source? thanks M ---------------------------------------------------------------------- Comment By: IOhannes m zmölnig (zmoelnig) Date: 2012-12-02 09:07 Message: raising priority, since the current state of affairs breaks all builds from current git. what i did: - after the import of the portaudio-20121101 snapshot, some files where missing to smoothly integrate with Pd's autotools system. - so I re-imported portaudio with the missing files (and used the 20121126 snapshot, since that was the one i found on the portaudio website) - I tried to keep the import to a minimum (but imported all files an ordinary make would want to run a compilation) - i added the new portaudio build-system to pd/configure.ac (so pa's configure will get called automatically by Pd's configure) - i also adjusted the link-path of libportaudio in pd/src/Makefile.am to it's new location ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=478072&aid=3591846&group_…

1 0

[ pure-data-Bugs-3593741 ] 8 channel Ogg Vorbis file does not play at correct speed
by SourceForge.net 07 Dec '12

07 Dec '12

Bugs item #3593741, was opened at 2012-12-07 09:59 Message generated for change (Tracker Item Submitted) made by danielhjames You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=478070&aid=3593741&group_… Please note that this message will contain a full copy of the comment thread, including the initial issue submission, for this request, not just the latest update. Category: pd-extended Group: v0.43 Status: Open Resolution: None Priority: 5 Private: No Submitted By: Daniel James (danielhjames) Assigned to: Hans-Christoph Steiner (eighthave) Summary: 8 channel Ogg Vorbis file does not play at correct speed Initial Comment: Using 0.43.4-extended-20121114 on Ubuntu Lucid, I can play a two channel Ogg Vorbis file or an eight channel WAV file but I cannot play an eight channel Ogg Vorbis file (made with Audacity, plays correctly in VLC) without the result sounding very slowed down. Please see attached screenshot of a demo patch. Please et me know if you would like sample audio files. ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=478070&aid=3593741&group_…

1 0

building for Windows on GNU/Linux
by Hans-Christoph Steiner 05 Dec '12

05 Dec '12

Hey Miller, I see that you are using MSVC via Wine on GNU/Linux. I don't know how well supported that arrangement is, for a widely supported version of that arrangement, you should try running the MinGW builds on GNU/Linux. The makefile.mingw should work fine for that just set compiler (i.e. make CC=/path/to/mingw-gcc). Here's lots of docs for doing this on Fedora: https://fedoraproject.org/wiki/MinGW Most distros include a mingw-gcc cross-compiler, so they just need to 'yum install', 'apt-get install', etc. .hc

2 2

[ pure-data-Patches-3578746 ] [bugfix]: non-2^n-BS bug
by SourceForge.net 05 Dec '12

05 Dec '12

Patches item #3578746, was opened at 2012-10-20 10:47 Message generated for change (Comment added) made by nobody You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=478072&aid=3578746&group_… Please note that this message will contain a full copy of the comment thread, including the initial issue submission, for this request, not just the latest update. Category: puredata Group: None Status: Open Resolution: None Priority: 5 Private: No Submitted By: Nobody/Anonymous (nobody) Assigned to: Miller Puckette (millerpuckette) Summary: [bugfix]: non-2^n-BS bug Initial Comment: I'm really not a 'spammer'! ...nor a hypochondriac... I've tested several versions of pd (0.42.5 - 0.43.3 (ext/van)). None of them is working with non-2^n-BSes! On the mailing-list ( http://www.mail-archive.com/pd-dev@iem.at/msg11344.html ), Claude Heiland-Allen states that it is already implemented.. (Thanks Claude) But I can't confirm that. I've tried the patch he posted there, - without success: The blocksize is always rounded to the next bigger power of 2, it seems! I don't know what he's on, but I'm on windows. Maybe that makes the difference.. (Is this a hardware- or Os-dependent feature..??) Please take a look at the attached pd-patch. It's my attempt to work out the problem.. If someone could give me a resonable answer, I promise I'll never come back again ;) ...for this issue... PS: It feels a little annoying if one does not get a real response or not a response at all! (Eg. look at the feature requests. The last one on the 1st page is like 2 years old, so you can't say there are too many to respond to each one..). This especially is true if the response is longer than the acual answer like "Yes", "No" or "I don't know". That would have been a big help for me already. Although it's not making the bug disappear automatically... Anything like if it is a useful/good idea is something I'd like to know, when posting something, for exapmle.. But no, I won't be scared away so easily.. ---------------------------------------------------------------------- Comment By: Nobody/Anonymous (nobody) Date: 2012-12-05 01:17 Message: Enjoyed looking through this, very good stuff, regards . <a href="http://social.webinova.us/jcow/blogs/viewstory/21641" title="Most Current In Addition To Modern Maxi Garments On The Net">Most Current In Addition To Modern Maxi Garments On The Net</a> ---------------------------------------------------------------------- Comment By: IOhannes m zmölnig (zmoelnig) Date: 2012-10-30 00:50 Message: feel free to report oddities as bugs - and please report each oddity as a separate bug (in the bug-tracker rather than follow-ups on this ticket) als make sure you report issues and don't ask questions like "how does this work?" :-) (and if you want to discuss the tracker/forum/mailinglist/IRC thing, please join the mailinglist and/or the IRC...!) ---------------------------------------------------------------------- Comment By: Nobody/Anonymous (nobody) Date: 2012-10-29 23:13 Message: ...oh now it's here... Just in case you're working on BS-things anyways I only wanted to mention that I have some more BS-related oddities. If you are interested, I can do a search and gather them. If you are talking about the hurleur-forum, I always thought that was not the right place for the devs?!? ..I haven't been there for a while ...since I have no problems with the installation... ;) I never wanted it to be only a question.. and it finally turned out to be a real bug. ..after the third try to convince you... ...I do what I can to help. ---------------------------------------------------------------------- Comment By: IOhannes m zmölnig (zmoelnig) Date: 2012-10-25 07:43 Message: @op: if you don't feel comfortable with the mailinglists, why not use another support-channel (IRC, forum?) the sf-trackers are here to track bugs and issues, not questions. ---------------------------------------------------------------------- Comment By: IOhannes m zmölnig (zmoelnig) Date: 2012-10-25 07:38 Message: indeed i can confirm the bug (as discussed on the mailinglist). i'll attach a (simplistic?) fix that appears to make it work with current git (and this time i tried hard to avoid introducing trailing whitespaces) ---------------------------------------------------------------------- Comment By: Nobody/Anonymous (nobody) Date: 2012-10-24 06:38 Message: I didn't know about that. I find the lists to be a little confusing.. There are so many, with all the archives. And sometimes some posts (from sf) only occur at the archive if I'm not mistaken..??? But I'll take a closer look... Anyways, did you (or someone else) take a look at the file?? Oh and just let me take this chance to thank you very very much for this brilliant piece of software. This is far more than programming. If it wasn't that good, it wouldn't have take me that much time... ..I mean I wouldn't have spent it... ---------------------------------------------------------------------- Comment By: Miller Puckette (millerpuckette) Date: 2012-10-20 21:38 Message: Sorry - SF is a pain for me - can you follow up on the mailing list? That way more people will be able to find the long thread that will doubtless ensue. ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=478072&aid=3578746&group_…

1 0

[ pure-data-Patches-2947822 ] space character in gui labels causes disaster
by SourceForge.net 04 Dec '12

04 Dec '12

Patches item #2947822, was opened at 2010-02-08 05:12 Message generated for change (Comment added) made by eighthave You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=478072&aid=2947822&group_… Please note that this message will contain a full copy of the comment thread, including the initial issue submission, for this request, not just the latest update. Category: puredata Group: None Status: Open Resolution: Fixed Priority: 5 Private: No Submitted By: Matteo Sisti Sette (sistisette) Assigned to: Miller Puckette (millerpuckette) Summary: space character in gui labels causes disaster Initial Comment: Steps to reproduce: 1 - Open the attached patch Note the size of the radio and toggle. Note also they have "send" symbols set to "xxx" and "yyy" respectively 2 - Click on the big bang at the top This composes a symbol containing space characters and sends it as a label to both the radio and the toggle. This seems to work: the label is actually set with spaces, and the radio and toggle still work. However, if you right-click on any of them and try to select "properties", the properties dialog won't open 3 - Now save the patch and close it 4 - Open the patch again. The radio and toggle properties have been completely "reset": their size have been reset to default, and also the "number" property of the radio. They have lost their "send" symbol too: you can see the outlet and if you use them, the "r" objects won't receive anything. However, if you open the properties dialog of the radio or toggle, you'll see that the "send" property is still there (not the same can be said for the size and number); if you now do Apply or Ok, the send symbol will be restored. So, spaces in the label of a gui object cause weirdnesses and potential disaster. Note that data is lost (the properties of the gui objects) if the patch is saved after assigning such a label, and i is lost silently with no error message at any moment. There are ather "forbidden" characters ithat don't behave properly n labels, such as the open square bracket "[" (as reported in another bug report). However the space character is more disatrous as it causes data loss. Object should either accept ALL characters for a label, or properly detect and refuse forbidden characters, which should be documented, avoiding unexpected consequences. ---------------------------------------------------------------------- >Comment By: Hans-Christoph Steiner (eighthave) Date: 2012-12-04 11:31 Message: Patch #0001 should be applied regardless. Its a simple fix that also address having characters like " [ ] in labels. As for patch #0002 , the iemguis also replace spaces with _ when you enter the name in the Properies panel. Symbols can have spaces, and you can set the label with a [label symbol( message, so as IOhannes pointed out, the missing piece of the puzzle is saving properly. Since patch #0002 makes Pd write out the file format that is already supported, I think it would be good to include it. Then strnescape() will also have to change when we implement proper escaping support throughout Pd. Unless you have plans to implement that in 0.44, I think its definitely worth including incremental solutions like patch #0002 ---------------------------------------------------------------------- Comment By: Miller Puckette (millerpuckette) Date: 2012-12-03 19:27 Message: Note the much simpler way the "number box" finesses this problem. I'm scared of adding code to m_binbuf.c to escape spaces (what will happen when we later allow backslashes into the mix?). But without the binbuf patch the fix to the IEM guis doesn't work - so I'm tempted for now just to do as in the number box - it's not incompatible to "do it right" at some ater date when we can thonk the whole thing through (it's all over the Pd code and all needs to be figured out together). ---------------------------------------------------------------------- Comment By: IOhannes m zmölnig (zmoelnig) Date: 2012-11-13 16:45 Message: added another patch (0002_*) that implements escaping spaces and tabs when saving the files to disk. i guess both patches should be applied for full fun. ---------------------------------------------------------------------- Comment By: IOhannes m zmölnig (zmoelnig) Date: 2012-11-13 15:36 Message: the bug.pd file does not correctly show the labels with blanks after reloading but it doesn't throw any warnings) however, the bugfix-patch doesn't work for me. nevertheless the problem is known and is not in the pd->gui communication, but rather in the way the label is stored on disk (and then interpreted on reload): afaict, blanks (and other special chars) really must be escaped with backslash (e.g. "A\ A\ A") when saved to disk ---------------------------------------------------------------------- Comment By: Hans-Christoph Steiner (eighthave) Date: 2012-11-13 12:52 Message: the values were not quoted, so just quoting them with {} fixes it. The patch is attached. ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=478072&aid=2947822&group_…

1 0

adding a built-in [change] object to the built-in GUI objects
by Hans-Christoph Steiner 04 Dec '12

04 Dec '12

Lots of patches use the built-in GUI objects for displays, and often a fast stream of events is hooked straight up to the GUI object, causing the GUI object to send many pointless updates, like draw commands when the number hasn't changed, or multiple draw commands per screen refresh cycle. I propose to only send the GUI update commands when the relevant value has changed. I think this should apply to both the main element, like the slider knob, and the label for that GUI object, since that's often used as a display. The code change is pretty simple, but I was wondering if people thought there could be any problems caused by this Here is the needed change, for example, for the hslider knob: index b352fb9..88681fc 100644 --- a/src/g_all_guis.h +++ b/src/g_all_guis.h @@ -185,6 +185,7 @@ typedef struct _hslider t_iemgui x_gui; int x_pos; int x_val; + int x_prev_val; int x_center; int x_thick; int x_lin0_log1; index 470771a..e1a3c83 100644 --- a/src/g_hslider.c +++ b/src/g_hslider.c @@ -33,7 +33,7 @@ static t_class *hslider_class; static void hslider_draw_update(t_gobj *client, t_glist *glist) { t_hslider *x = (t_hslider *)client; - if (glist_isvisible(glist)) + if (glist_isvisible(glist) && x->x_val != x->x_prev_val) { int r = text_xpix(&x->x_gui.x_obj, glist) + (x->x_val + 50)/100; int ypos=text_ypix(&x->x_gui.x_obj, glist); @@ -57,6 +57,7 @@ static void hslider_draw_update(t_gobj *client, t_glist *glist) x->x_thick = 0; } } + x->x_prev_val = x->x_val; } }

4 7

[ pure-data-Patches-3591224 ] make error message text match previously existing version of
by SourceForge.net 04 Dec '12

04 Dec '12

Patches item #3591224, was opened at 2012-11-29 16:57 Message generated for change (Comment added) made by millerpuckette You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=478072&aid=3591224&group_… Please note that this message will contain a full copy of the comment thread, including the initial issue submission, for this request, not just the latest update. Category: puredata Group: bugfix >Status: Pending >Resolution: Accepted Priority: 7 Private: No Submitted By: Hans-Christoph Steiner (eighthave) Assigned to: Miller Puckette (millerpuckette) Summary: make error message text match previously existing version of Initial Comment: pd_connnect::pd_readsocket now also can print a "MISSING CLOSE-BRACE" error, this patch just makes the text of that error message the exact same as the one that existed before so that it only needs to be translated once ---------------------------------------------------------------------- >Comment By: Miller Puckette (millerpuckette) Date: 2012-12-03 20:57 Message: accepted for 0.44 - oddly it was out of date lready so applied by hand ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=478072&aid=3591224&group_…

1 0

[ pure-data-Patches-3587384 ] Replacing deprecated functions
by SourceForge.net 04 Dec '12

04 Dec '12

Patches item #3587384, was opened at 2012-11-14 21:06 Message generated for change (Comment added) made by millerpuckette You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=478072&aid=3587384&group_… Please note that this message will contain a full copy of the comment thread, including the initial issue submission, for this request, not just the latest update. Category: puredata Group: bugfix >Status: Pending >Resolution: Accepted Priority: 5 Private: No Submitted By: Nobody/Anonymous (nobody) Assigned to: Miller Puckette (millerpuckette) Summary: Replacing deprecated functions Initial Comment: vexp_fun.c currently uses the old BSD-style functions drem and finite. These functions have been removed from XCode 4.4 and later, so that the expr~ family of externals fails to compile now. This patch replaces drem with remainder and finite with isfinite, which should work everywhere. ---------------------------------------------------------------------- >Comment By: Miller Puckette (millerpuckette) Date: 2012-12-03 20:32 Message: accepted for 0.44 ---------------------------------------------------------------------- Comment By: Hans-Christoph Steiner (eighthave) Date: 2012-11-20 12:34 Message: Makes sense to me. remainder() is the C99 replacement for drem(), and isfinite() is the C99 replacement for finite(). C99 marks drem() and finite() as deprecated. ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=478072&aid=3587384&group_…

1 0

[ pure-data-Patches-3587383 ] before loading a GUI plugin, check its not already loaded
by SourceForge.net 04 Dec '12

04 Dec '12

Patches item #3587383, was opened at 2012-11-14 21:05 Message generated for change (Comment added) made by millerpuckette You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=478072&aid=3587383&group_… Please note that this message will contain a full copy of the comment thread, including the initial issue submission, for this request, not just the latest update. Category: puredata Group: bugfix >Status: Pending >Resolution: Accepted Priority: 7 Private: No Submitted By: Hans-Christoph Steiner (eighthave) Assigned to: Miller Puckette (millerpuckette) Summary: before loading a GUI plugin, check its not already loaded Initial Comment: only load GUI plugins with the same name once, to prevent mayhem The actual file name of the plugin is stored in a list. Whenever a plugin is being loaded, first pd-gui checks whether that file name is listed as loaded. If its loaded already, pd-gui prints a message then ignores it ---------------------------------------------------------------------- >Comment By: Miller Puckette (millerpuckette) Date: 2012-12-03 20:24 Message: Mayhem you say? we needs more of that, what? OK for 0.44 ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=478072&aid=3587383&group_…

1 0

Jump to page:

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 December 2012