Pd-dev

pd-dev@lists.iem.at

2 participants
9620 discussions

Find external and install it in an custom folder does not work anymore in Pd 052.1on Debian ...
by Winfried Ritsch 12 Jan '22

12 Jan '22

First, good new year and thanks for 052.1 ! ... In previous version I got a file dialog and then I could store it e.g. in my "libs" folder within my project, which I also advised to my students sharing my projects, since they operate on different OS... and this is an easy task for most users. Now I get an error (see below) [1]. Since I do not want to make an Pd folder, since I work on many different (studio) computers and do not want to pollute the home directory with any Application folder like "Pd" [2] on every account ... So is this behavior intentional or an error ... so I will do an issue report. If intentional add an option to save the external into an chosen folder, since this is really important for distribution of Pd applications. mfg winfried [1] Error response: > couldn't open "/home/winfried/.local/lib/pd/extra/dekentmp.0": no such file > or directory couldn't open "/home/winfried/.local/lib/pd/extra/dekentmp.0": > no such file or directory> > while executing > > "open $downloadfilename w" > > (procedure "::deken::utilities::download_file" line 4) > invoked from within > > "::deken::utilities::download_file $URL $fullpkgfile > "::deken::download_progress""> > (procedure "::deken::clicked_link" line 16) > invoked from within > > "::deken::clicked_link > {http://puredata.info/Members/musil/software/iemgui/1.21-1/iemgui[v1.21-1]( > Darwin-amd64-32)(Darwin-i386-32)(Linux-amd64-32)(Wind..."> > (command bound to event) [2] Also I do not like the recent behaviour to be forced to choose default folders for externals installation, since also on headless computers when setup automatically, it cannot answer an file dialog on first startup, I do not know who had this idea or missed the discussion, to ask for one on first start of Pd and not as an option to add it later...

2 2

Re: [PD-dev] Find external and install it in an custom folder does not work anymore in Pd 052.1on Debian ...
by Dan Wilcox 07 Jan '22

07 Jan '22

> On Jan 6, 2022, at 12:00 PM, pd-dev-request(a)lists.iem.at wrote: > > Message: 2 > Date: Wed, 05 Jan 2022 20:19:07 +0100 > From: Winfried Ritsch <ritsch(a)iem.at <mailto:ritsch@iem.at>> > To: pd-dev(a)lists.iem.at <mailto:pd-dev@lists.iem.at> > Subject: Re: [PD-dev] Find external and install it in an custom folder > does not work anymore in Pd 052.1on Debian ... > Message-ID: <1975418.ThMadnPRZM@ras> > Content-Type: text/plain; charset="iso-8859-1" > > Thank you for the response. > > Am Mittwoch, 5. J?nner 2022, 13:07:38 CET schrieb Dan Wilcox: >>> On Jan 5, 2022, at 12:00 PM, pd-dev-request(a)lists.iem.at <mailto:pd-dev-request@lists.iem.at> wrote: >>> >>> Message: 1 >>> Date: Tue, 04 Jan 2022 20:58:02 +0100 >>> From: Winfried Ritsch <ritsch(a)iem.at <mailto:ritsch@iem.at> <mailto:ritsch@iem.at <mailto:ritsch@iem.at>>> >> >>> [1] Error response: >> That perhaps looks like a problem with deken itself. >> > It happens if Pd Document path is disabled and system path cleared. > > Not really a problem, but previous behaviour starting a file dialog did set the > external path, so we have to change this settings before using find > externals... anyway an error shouldnt happen. Sounds like a bug yes. Please open an issue, if there isn't one already. >>> [2] Also I do not like the recent behaviour to be forced to choose default >>> folders for externals installation, since also on headless computers when >>> setup automatically, it cannot answer an file dialog on first startup, I >>> do not know who had this idea or missed the discussion, to ask for one on >>> first start of Pd and not as an option to add it later... >> > Thanks, and one correction to above, not headless but mouseless and > keyboardless, on headless it does not ask for default folder. Ok good. I would be very remiss if Pd hung on a GUI plugin when running -nogui... >> A very long discussion went on, part of which ended up as a proposal: >> >> https://github.com/pure-data/pure-data/pull/152 <https://github.com/pure-data/pure-data/pull/152> >> <https://github.com/pure-data/pure-data/pull/152 <https://github.com/pure-data/pure-data/pull/152>> >> > Thank you for the hint, now I understand... I am too late for respones ;-). > > ... anyway I thought Pd is gentleware: respectful to environments, not > wolvesware which tend to leave their marks everywhere where they can ... > but maybe it's only us old hacker generation, the younger generation got used > to it. I'm not going to rehash the whole thing but I tried to find a compromise for both, even if it's not perfect. >> There is a key in the GUI settings for the Documents path which can be set >> to "DISABLED" to avoid triggering the initial dialog. This is what is set >> when you select No. >> > ok, so I can do on remote command: > $ mkdir -p ~/.config/Pd/org.puredata.pd.pd-gui > $ echo DISABLED > .config/Pd/org.puredata.pd.pd-gui/docspath.conf I don't know the details of the Linux conf handling, but that looks correct. For macOS, this would be the equivalent: $ defaults write org.puredata.pd.pd-gui docspath "DISABLED" >> Also, the documents directory handling is implemented as a plugin, so it >> could be deleted or renamed to avoid it entirely. >> > I will take a look in this, but better dont touch the pd distributions > packages.. Maybe we need to add a plugin override in general (-noplugin "docsdir") or a docspath override specifically, ala -nodocsdir? I tried to make sure if doesn't ask if the home directory is not available or not writable (ala readonly setups), but maybe there can be another hacker-friendly check? Logic is in the plugin init proc: https://github.com/pure-data/pure-data/blob/master/tcl/pd_docsdir.tcl#L40 <https://github.com/pure-data/pure-data/blob/master/tcl/pd_docsdir.tcl#L40> My original thinking was that if you didn't want the docspath handling, you'd just press No once and never see it again, however if you are working with more than a couple normal desktop installs, this can of course not be the case. > mfg winfried >> -------- >> Dan Wilcox >> @danomatika <http://twitter.com/danomatika <http://twitter.com/danomatika>> >> danomatika.com <http://danomatika.com/> <http://danomatika.com/ <http://danomatika.com/>> >> robotcowboy.com <http://robotcowboy.com/> <http://robotcowboy.com/ <http://robotcowboy.com/>> > > > -- > - ao.Univ.Prof. DI Winfried Ritsch > - ritsch(a)iem.at <mailto:ritsch@iem.at> - http://iem.at/ritsch <http://iem.at/ritsch> > - Institut fuer Elektronische Musik und Akustik > - University of Music and Dramatic Art Graz > - Tel. ++43-316-389-3510 (3170) Fax ++43-316-389-3171 -------- Dan Wilcox @danomatika <http://twitter.com/danomatika> danomatika.com <http://danomatika.com/> robotcowboy.com <http://robotcowboy.com/>

3 2

Re: [PD-dev] Find external and install it in an custom folder does not work anymore in Pd 052.1on Debian ...
by Dan Wilcox 05 Jan '22

05 Jan '22

> On Jan 5, 2022, at 12:00 PM, pd-dev-request(a)lists.iem.at wrote: > > Message: 1 > Date: Tue, 04 Jan 2022 20:58:02 +0100 > From: Winfried Ritsch <ritsch(a)iem.at <mailto:ritsch@iem.at>> > > > [1] Error response: That perhaps looks like a problem with deken itself. > [2] Also I do not like the recent behaviour to be forced to choose default > folders for externals installation, since also on headless computers when > setup automatically, it cannot answer an file dialog on first startup, I do not > know who had this idea or missed the discussion, to ask for one on first start > of Pd and not as an option to add it later... A very long discussion went on, part of which ended up as a proposal: https://github.com/pure-data/pure-data/pull/152 <https://github.com/pure-data/pure-data/pull/152> There is a key in the GUI settings for the Documents path which can be set to "DISABLED" to avoid triggering the initial dialog. This is what is set when you select No. Also, the documents directory handling is implemented as a plugin, so it could be deleted or renamed to avoid it entirely. -------- Dan Wilcox @danomatika <http://twitter.com/danomatika> danomatika.com <http://danomatika.com/> robotcowboy.com <http://robotcowboy.com/>

2 1

Re: [PD-dev] [PD] [PD-announce] Pd version 0.52-0test3 released
by IOhannes m zmoelnig 15 Dec '21

15 Dec '21

On 12/12/21 21:05, Miller Puckette wrote: > Nope - Wish 8.6.12 seems to depend on newer Mac features than are > on OSX 10.7. > > But I'm OK with continuing to compile for 10.7 on my old machine and > make it available to people like me :) i think i found the problem why Tcl/Tk-8.6.12 cannot be compiled on OSX-10.7 https://trac.macports.org/ticket/63868 i'll do some more tests (in the vein of whether we can create an amd64/arm64 universal binary that runs with OSX-10.7 through macOS-12 (well macOS-10.15 as this is the last version i can currently test myself) and submit a PR if all goes well. i think it would be fabulous if we could use Tcl/Tk>=8.6.12 on all platforms. gfmasdr IOhannes

1 0

Re: [PD-dev] [PD] [PD-announce] Pd version 0.52-0test3 released
by IOhannes m zmölnig 15 Dec '21

15 Dec '21

Am 12. Dezember 2021 21:05:14 MEZ schrieb Miller Puckette via Pd-list <pd-list(a)lists.iem.at>: > > >Nope - Wish 8.6.12 seems to depend on newer Mac features than are >on OSX 10.7. > >But I'm OK with continuing to compile for 10.7 on my old machine and >make it available to people like me :) > So how do we proceed from here? Currently my CI builds are unable to create a working Wish.app themselves, so they fetch the universal builds from miller's site. I don't really like this much, as I think the upload was mainly done for testing whether it works at all, rather than for deployment. So: should these binaries be included in mac/stuff/? Or not (which i'm fine with as well, see my post scriptum. In any case I'd like to know) (taking this to PD-dev) mfg.sfg.jfd IOhannes PS: I can't believe I'm writing this - as I'm typically in the forefront of opposing any binary assets in git repositories. probably I should just upload the packages to some CI related "asset store".

1 0

fixed the Pd_documentation link in http://puredata.info/docs/manuals/pd/pd-documentation
by Antoine Rousseau 09 Dec '21

09 Dec '21

FYI, I've fixed the link to Miller's site that appears at the top of http://puredata.info/docs/manuals/pd/pd-documentation ; it was using a "www." prefix that didn't work. I've set it to " http://msp.ucsd.edu/Pd_documentation/index.htm", which is how my browser finally resolves it. BTW, this page (on puredata.info) is outdated. I'm not sure of the correct way to update it to the latest (github) version. Also, in the latest (github) version (and the Miller's site version) of index.htm, the link is "http://msp.ucsd.edu/software.html", but that seems wrong (as this points to the root of Miller's "software" page, not the "latest version of this [documentation] page". Antoine

2 2

Question about right hand generic inlet creation
by Iain Duncan 08 Dec '21

08 Dec '21

Hi folks, I'm not sure if I'm understanding the externals tutorial properly. It seems to me that it is saying that one cannot make a generic inlet past the first one, as in something that passes any incoming message to the same handler as an array of atoms. Is that the case or am I totally misunderstanding that? And if it's not the case, if someone could tell me how to make the equivalent of class_addanything happen for other inlets than the first, that would be lovely. thanks! iain

2 3

how to cast printed pointers? (commit c019784b83 throwing warnings)
by Miller Puckette 03 Dec '21

03 Dec '21

To Pd dev: After merging commit c019784b83 (get rid of a 3 more bogus (unsigned long) casts) I'm now getting compiler warnings like this: g_editor.c:1884:73: warning: format ‘%lx’ expects argument of type ‘long unsigned int’, but argument 4 has type ‘struct _glist *’ [-Wformat=] 1884 | snprintf(cbuf, MAXPDSTRING - 2, "pdtk_canvas_setparents .x%lx", c); | ~~^ ~ | | | | | struct _glist * | long unsigned int .. I presume there's a good reason to change the code (what if "long" is the wrong size?) but I'd like to figure out a way to do it that doesn't throw warnings. Alternatively, if it doesn't cause bugs to just cast pointers to long for the purpose of making tags, let's just revert to the old (unsigned long) cast. cheers Miller

4 8

Re: [PD-dev] why no more "sys_get_audio_params" in s_stuff.h? What to do now?
by Dan Wilcox 24 Oct '21

24 Oct '21

I think Miller's answer is: > in the meantime objects that use s_stuff.h definitions will > probably have to track changes in the internal APIs. Which, I believe, indicates the externals need to be updated. You could make your own local version and only use it if the Pd version is >= 0.52. As to how it should be done and where is the code? I dunno. I'd suggest looking at old versions of the Pd sources to see what the old function did, then see if there is a way to replicate it in whatever manner works with Miller's updates. The functions in s_stuff.h have always been subject to change. > On Oct 23, 2021, at 12:00 PM, pd-dev-request(a)lists.iem.at wrote: > > By the way, the [audiosettings] externals from extended and available in > deken in the mediasettings library has the same issue and fails to create. > > what do we do now? -------- Dan Wilcox @danomatika <http://twitter.com/danomatika> danomatika.com <http://danomatika.com/> robotcowboy.com <http://robotcowboy.com/>

2 1

why no more "sys_get_audio_params" in s_stuff.h? What to do now?
by Alexandre Torres Porres 22 Oct '21

22 Oct '21

Hi, next release (or current state on gitub) breaks one of my externals - [sr~] - because it uses 'sys_get_audio_params' from s_stuff.h and it got removed from the code. One idea of this object is that it can change Pd's running sample rate from within the patch. By the way, the [audiosettings] externals from extended and available in deken in the mediasettings library has the same issue and fails to create. what do we do now? thanks cheers

2 1

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