Pd-dev January 2006

pd-dev@lists.iem.at

27 participants
100 discussions

unit tests!
by Mathieu Bouchard 06 Jan '06

06 Jan '06

I'm starting yet another project, and this one is about unit-testing. Here are some of my ideas. It is in the CVS in /abstractions/pureunity/ ------------------8<--------cut-here--------8<------------------ PureUnity Copyright 2006 by Mathieu Bouchard <matju à artengine point ca> $Id: README,v 1.2 2005/12/29 18:47:56 matju Exp $ +-+-+--+---+-----+--------+-------------+---------------------+ GOALS 1. To provide a unit-test framework, which also provide benchmarking features, all made in Pd for use in Pd. 2. To provide tests for functionality in internals, externals, abstractions, etc., in a modularized way, in a DRY/OAOO fashion, thus abstracting out common features so that many objects share the same test patch for the features that they have in common. +-+-+--+---+-----+--------+-------------+---------------------+ TEST PROTOCOL new: create common (reusable) fixtures. inlet 0: bang: run all available tests in that class. individual tests don't have to be available through individual methods but may. If they do, the names of the methods must match those given in the test results. each test should build its own non-reusable fixtures and reinitialize common fixtures, not assuming that the previous tests have left the common fixtures in a normal state. outlet 0: test results. a sequence of lists like: list $name $passed? $accuracy $elapsed for example: list where: $name is a symbol $passed? is either 0 for failure or 1 for success $accuracy is a float proportional to relative error on math (if not applicable, use 0) $elapsed is a float, the time elapsed in milliseconds or it is the symbol "-" if not measured. +-+-+--+---+-----+--------+-------------+---------------------+ SEVERITIES (in decreasing order) * crash: Segmentation Fault, Bus Error, Illegal Instruction, Infinite Loop, etc. You can't deal with those errors at the level of the tests. Maybe there should be a way to tell a test object to skip certain tests, by name, in order to be able to perform as many tests as possible while waiting for a fix. It could become possible to rescue from some of those crashes if Pd supported exceptions (stack-unwinding). * corruption: this may cause future crashes and failures on innocent objects/features. I have no solution for this except to be careful. * post(),error(),pd_error(): Gets printed in the console. The problem is that those can't be handled by the test objects, so someone has to read them and interpret them. Also they prevent test objects to ensure that error conditions produce error messages. * pd_error2(): I wish this would exist. It would be sort of like pd_error() but it would produce a pd message instead, whose selector would be an error code, designed to be both localizable and [route]able. By default, that message would be sent to the console, but there would be an internal class designed to catch those messages. (If stack-unwinding were possible, it would be disabled by default on pd_error2 and could be enabled explicitly by-selector). * failure: a test object reports a problem through outlet 0. _ _ __ ___ _____ ________ _____________ _____________________ ... | Mathieu Bouchard - tél:+1.514.383.3801 - http://artengine.ca/matju | Freelance Digital Arts Engineer, Montréal QC Canada

13 47

DT_TEXTREL warning
by Jamie Bullock 06 Jan '06

06 Jan '06

Hi, I have just noticed that I am getting the following warning when linking PD external object libraries: ld: warning: creating a DT_TEXTREL in object. I think this is since ugrading to gcc 3.4.4. ld version 2.15.92.0.2. I can get rid of it by including -fPIC in the compiler flags. I don't seem to experience any problems with or without -fPIC, other than the warning. Is this something I should be concerned about? Jamie

2 1

writing loaders for the new loader system
by Hans-Christoph Steiner 06 Jan '06

06 Jan '06

Now that Thomas' loader patch is included in pd core, I'd like to try my hand at writing a loader. Its quite simple, as far as I can tell, to write an alternate loader that loads compiled C objects, but I don't see how to do it for other languages, like Java. What's the plan for having other languages access the Pd API? Also, it seems to me that ultimately everything should be loaded using loader functions on the loader queue, including traditional libs, dlls, and abstractions. That would make it quite clean, then there would also be the possibility of reordering the loaders. .hc ________________________________________________________________________ ____ "Looking at things from a more basic level, you can come up with a more direct solution... It may sound small in theory, but it in practice, it can change entire economies." - Amy Smith

1 0

[ pure-data-Bugs-1348910 ] crash when printing bang after [t s]
by SourceForge.net 05 Jan '06

05 Jan '06

Bugs item #1348910, was opened at 2005-11-05 10:49 Message generated for change (Settings changed) made by zmoelnig You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=478070&aid=1348910&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: Closed >Resolution: Fixed Priority: 5 Submitted By: eerne (eerne) Assigned to: Nobody/Anonymous (nobody) Summary: crash when printing bang after [t s] Initial Comment: pd-0.39-1 (linux, osx) sending a bang through a trigger [t s] then [print] crashes pd. [bng] | [t s] | [print] ---------------------------------------------------------------------- Comment By: Miller Puckette (millerpuckette) Date: 2005-11-06 02:47 Message: Logged In: YES user_id=313747 Thanks... updated my version so that "t s" sends out the symbol "symbol". Miller ---------------------------------------------------------------------- Comment By: Mathieu Bouchard (matju) Date: 2005-11-05 18:03 Message: Logged In: YES user_id=801174 fixed in devel_0_39 ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=478070&aid=1348910&group_…

1 0

[ pure-data-Patches-1358937 ] class loader for non alpha-numeric classnames
by SourceForge.net 05 Jan '06

05 Jan '06

Patches item #1358937, was opened at 2005-11-17 13:52 Message generated for change (Settings changed) made by zmoelnig You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=478072&aid=1358937&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-dev Group: wishlist >Status: Closed >Resolution: Accepted Priority: 5 Submitted By: IOhannes m zmÃ¶lnig (zmoelnig) Assigned to: Miller Puckette (millerpuckette) Summary: class loader for non alpha-numeric classnames Initial Comment: object-names containing non-alphanumerics like [||~] can (sometimes) be not represented on filesystems (e.g. "|" is a forbidden character) and more often they cannot be used to construct a valid setup-function ("||~_setup" or "||_tilde_setup" are really bad). the way the "~" is handled, is non-generic and thus sub-optimal. as a solution me and hcs proposed an encoding into alphanumeric-values, using a hexadecimal representation of all characters but [0-9A-Za-z_] (e.g. "+" is ascii 43 and is thus represented by "0x2b" (hex-value all lowercase and prepended with "0x") e.g. if we have a new class "mtx_||", pd first attempts to find a file called "mtx_||.dll". if it succeeds, it will try to call the "mtx_||_setup()" function. if that fails we suggest to try and call a function "setup_mtx_0x7c0x7c()" (the keyword setup is now at the beginning of the function-name, in order to prevent the names starting with numbers and in order to distinguish between the normal setup-methods). if no "mtx_||.dll" can be found, pd should then search for a file "mtx_0x7c0x7c.dll" (guaranteed to be representable on any filesystem); search this file for the 2 setup-functions. if all fails try to find "mtx_||.pd" and then "mtx_0x7c0x7c.pd" ---------------------------------------------------------------------- Comment By: Miller Puckette (millerpuckette) Date: 2005-12-31 21:24 Message: Logged In: YES user_id=313747 It's ugly, but I think it's really needed anyway. It would also be good someday to fix abstractions to do this too. Anyway, I took it for upcoming version 0.40 ---------------------------------------------------------------------- Comment By: IOhannes m zmÃ¶lnig (zmoelnig) Date: 2005-11-17 13:54 Message: Logged In: YES user_id=564396 attached is a patch against s_loader.c, s_class.c and m_stuff.h that achieves this functionality ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=478072&aid=1358937&group_…

1 0

[ pure-data-Patches-1353230 ] class loader hook
by SourceForge.net 05 Jan '06

05 Jan '06

Patches item #1353230, was opened at 2005-11-10 16:40 Message generated for change (Settings changed) made by zmoelnig You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=478072&aid=1353230&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: wishlist >Status: Closed >Resolution: Accepted Priority: 5 Submitted By: Thomas Grill (xovo) Assigned to: Nobody/Anonymous (nobody) Summary: class loader hook Initial Comment: Hi all, i'd like to propose a loader hook functionality, so that external logic can be used if the loading of a PD external binary object has failed. This opens up two possibilites: 1) it allows to use external objects written in languages other than C (above all scripting languages) which cannot provide an exposed class_setup function to be used in the same way as traditional C-coded externals. 2) it allows to create external object functionality more dynamically, since the exact name need not be known (and fixed into binary) beforehand. The attached patch is for s_loader.c. The sys_loader function would still have to be exposed in m_pd.h gr~~~ ---------------------------------------------------------------------- Comment By: Miller Puckette (millerpuckette) Date: 2005-12-31 21:54 Message: Logged In: YES user_id=313747 Had to fool with this and the "altname" patch to get them to work together. Should be working now (but I have no non-C externs to test this on...) ---------------------------------------------------------------------- Comment By: Thomas Grill (xovo) Date: 2005-11-23 12:20 Message: Logged In: YES user_id=350252 small usage example: (a "cil" external has to be loaded in PD to invoke the cil_setup function) static int loader(char *dirname,char *classname) { // TODO: check the classname, if it's in our scope // return 0 if it isn't // here we could load some binaries, like in sys_load_lib_internal // or we could just create new pd classes, like in typical xxx_setup functions // ok, class has been loaded return 1; } void cil_setup() { // install our class loader ::sys_loader(loader); } ---------------------------------------------------------------------- Comment By: Thomas Grill (xovo) Date: 2005-11-23 12:12 Message: Logged In: YES user_id=350252 here's a "diff -uw" patch from the pd root directory involving both m_pd.h and s_loader.c ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=478072&aid=1353230&group_…

1 0

[ pure-data-Patches-1368484 ] [poly] memory leak
by SourceForge.net 05 Jan '06

05 Jan '06

Patches item #1368484, was opened at 2005-11-28 19:35 Message generated for change (Settings changed) made by zmoelnig You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=478072&aid=1368484&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: 5 Submitted By: Mathieu Bouchard (matju) Assigned to: Nobody/Anonymous (nobody) Summary: [poly] memory leak Initial Comment: fixed [poly] memory leak removed unused vars ---------------------------------------------------------------------- Comment By: Miller Puckette (millerpuckette) Date: 2005-12-31 20:50 Message: Logged In: YES user_id=313747 Took for 0.40 ---------------------------------------------------------------------- Comment By: Thomas Grill (xovo) Date: 2005-11-30 14:46 Message: Logged In: YES user_id=350252 here's the missing patch: Index: x_midi.c =================================================================== RCS file: /cvsroot/pure-data/pd/src/x_midi.c,v retrieving revision 1.2.8.3 diff -u -r1.2.8.3 x_midi.c --- x_midi.c 30 Nov 2005 13:36:58 -0000 1.2.8.3 +++ x_midi.c 30 Nov 2005 13:44:24 -0000 @@ -1187,7 +1187,7 @@ static void poly_setup(void) { poly_class = class_new(gensym("poly"), - (t_newmethod)poly_new, (t_method)poly_clear, + (t_newmethod)poly_new, (t_method)poly_free, sizeof(t_poly), 0, A_DEFFLOAT, A_DEFFLOAT, 0); class_addfloat(poly_class, poly_float); class_addmethod(poly_class, (t_method)poly_stop, gensym("stop"), 0); ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=478072&aid=1368484&group_…

1 0

[ pure-data-Patches-1378927 ] OSX/pd 0.39: fix for loading externals using static libs
by SourceForge.net 05 Jan '06

05 Jan '06

Patches item #1378927, was opened at 2005-12-12 19:37 Message generated for change (Settings changed) made by zmoelnig You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=478072&aid=1378927&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: None Priority: 7 Submitted By: Thomas Grill (xovo) Assigned to: Nobody/Anonymous (nobody) Summary: OSX/pd 0.39: fix for loading externals using static libs Initial Comment: Hi all, unlike with pd 0.38, externals that share equal symbol names (like when they have been compiled using the same static lib) won't get loaded, but rather bail out with a "linker error". This can be easily fixed by including the NSLINKMODULE_OPTION_PRIVATE option in s_loader.c ret = NSLinkModule( image, filename, NSLINKMODULE_OPTION_BINDNOW | NSLINKMODULE_OPTION_RETURN_ON_ERROR | NSLINKMODULE_OPTION_PRIVATE); Is there a reason this has been taken out in pd 0.39? best greetings, Thomas ---------------------------------------------------------------------- Comment By: Miller Puckette (millerpuckette) Date: 2005-12-31 06:31 Message: Logged In: YES user_id=313747 I ended up adding "DL_OPEN" to the configure.in special case code for OSX. As soon as I can get a correct configure script (tried real hard already but can't get autoconf to work as advertised yet) this can be automatic. ---------------------------------------------------------------------- Comment By: Thomas Grill (xovo) Date: 2005-12-13 17:55 Message: Logged In: YES user_id=350252 here's the patch ---------------------------------------------------------------------- Comment By: Jamie Tittle (tigital) Date: 2005-12-12 20:40 Message: Logged In: YES user_id=634044 ...this was kinda covered in the bugtracker last year: "[ 1082938 ] sys_load_lib() on OSX too limited?" https://sourceforge.net/tracker/index.php?func=detail&aid=1082938&group_id=… ...I'd like to suggest we ditch the NSLinkModule() stuff on OSX, as it's been deprecated since 10.3.x: the replacement is (drumroll please), dlopen()! As a matter of fact, hans and I have been using a patch to do just this in the most recent development releases: it's currently available in /packages/patches/darwin/darwin_dlopen.patch, which I would attach, but it's not letting me atm... ...also, I strongly suggest that we make another change in the building of externals, from bundles to dynamic libs: -dynamiclib -mmacosx-version-min=10.3 -undefined dynamic_lookup (or suppress) ...this works for everything I've tried, but admittedly I haven't done it with all externals or flext... ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=478072&aid=1378927&group_…

1 0

[ pure-data-Patches-1285113 ] [print] cuts output, include this in help patch
by SourceForge.net 05 Jan '06

05 Jan '06

Patches item #1285113, was opened at 2005-09-08 18:56 Message generated for change (Settings changed) made by zmoelnig You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=478072&aid=1285113&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: documentation >Status: Closed >Resolution: Accepted Priority: 5 Submitted By: Piotr Majdak (petibub) Assigned to: Miller Puckette (millerpuckette) Summary: [print] cuts output, include this in help patch Initial Comment: [print] cuts the output if the message is a list with two or more symbols. The second and following symbols are printed up to the length of 78 characters and an asterisk is added. This behaviour should be mentioned in the help patch print-help.pd, see the attachment as an example. ---------------------------------------------------------------------- Comment By: Miller Puckette (millerpuckette) Date: 2005-12-31 22:02 Message: Logged In: YES user_id=313747 took for 0.40. ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=478072&aid=1285113&group_…

1 0

[ pure-data-Patches-1244757 ] {open, save}panel with settable path
by SourceForge.net 05 Jan '06

05 Jan '06

Patches item #1244757, was opened at 2005-07-25 22:26 Message generated for change (Settings changed) made by zmoelnig You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=478072&aid=1244757&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: wishlist >Status: Closed >Resolution: Accepted Priority: 5 Submitted By: IOhannes m zmÃ¶lnig (zmoelnig) Assigned to: Miller Puckette (millerpuckette) Summary: {open,save}panel with settable path Initial Comment: this patch (against the current 25/07/2005) HEAD of the CVS (pd-0.39test4) allows the current path of [openpanel]/[savepanel] to be set. once the path is set, it will not change anymore "automatically"; however, this patch gives even more wird tcl/tk output about "unknown commands" than the unpatched version does. there is clearly some bug in the u_main.tk ---------------------------------------------------------------------- Comment By: Miller Puckette (millerpuckette) Date: 2005-12-31 22:30 Message: Logged In: YES user_id=313747 I changed this somewhat (so that you can just send "symbol" to openpanel/savepanel)... it works OK except that, once you've used the feature, the savepanel wants to go back to the directory you specified afterward (not the one where you eventually opened the file!)... yuck... ---------------------------------------------------------------------- Comment By: IOhannes m zmÃ¶lnig (zmoelnig) Date: 2005-09-30 15:07 Message: Logged In: YES user_id=564396 with pd-0.39.0 the "unknown commands" are fixed. however, i still would enjoy the openpanel/savepanel with a settable path. ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=478072&aid=1244757&group_…

1 0

← Newer
1
2
3
4
5
6
7
8
9
10
Older →

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 January 2006