Hi all,
Pd version 0.45-0 is available on http://crca.ucsd.edu/~msp/software.htm or via git from sourceforge: git clone git://git.code.sf.net/p/pure-data/pure-data cd pure-data git checkout -b 0.45
As always I'm sure there will be problems here and there - you're welcome to report them on the Pd mailing list (pd-list@iem.at) which is always the fastest way to get me to see them.
cheers Miller
Pd-announce mailing list Pd-announce@iem.at http://lists.puredata.info/listinfo/pd-announce
Objects/messages/comments have settable box widths.
hi there, please, how can I use this new feature? I didn't get it. cheers
2013/8/23 Miller Puckette msp@ucsd.edu
Hi all,
Pd version 0.45-0 is available on http://crca.ucsd.edu/~msp/software.htm or via git from sourceforge: git clone git://git.code.sf.net/p/pure-data/pure-data cd pure-data git checkout -b 0.45
As always I'm sure there will be problems here and there - you're welcome to report them on the Pd mailing list (pd-list@iem.at) which is always the fastest way to get me to see them.
cheers Miller
Pd-announce mailing list Pd-announce@iem.at http://lists.puredata.info/listinfo/pd-announce
Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
'hover' over the right-hand side of any object/message/comment in edit mode and the cursor should change to a 'grow' cursor (double arrow) - then drag to the left or right.
enjoy... Miller On Fri, Aug 23, 2013 at 06:42:47PM -0300, Alexandre Torres Porres wrote:
Objects/messages/comments have settable box widths.
hi there, please, how can I use this new feature? I didn't get it. cheers
2013/8/23 Miller Puckette msp@ucsd.edu
Hi all,
Pd version 0.45-0 is available on http://crca.ucsd.edu/~msp/software.htm or via git from sourceforge: git clone git://git.code.sf.net/p/pure-data/pure-data cd pure-data git checkout -b 0.45
As always I'm sure there will be problems here and there - you're welcome to report them on the Pd mailing list (pd-list@iem.at) which is always the fastest way to get me to see them.
cheers Miller
Pd-announce mailing list Pd-announce@iem.at http://lists.puredata.info/listinfo/pd-announce
Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
anyone else having GUI trouble seeing sliders and number boxes move/change values?
i'm on mac os 10.7.5
cheers
2013/8/23 Miller Puckette msp@ucsd.edu
Hi all,
Pd version 0.45-0 is available on http://crca.ucsd.edu/~msp/software.htm or via git from sourceforge: git clone git://git.code.sf.net/p/pure-data/pure-data cd pure-data git checkout -b 0.45
As always I'm sure there will be problems here and there - you're welcome to report them on the Pd mailing list (pd-list@iem.at) which is always the fastest way to get me to see them.
cheers Miller
Pd-announce mailing list Pd-announce@iem.at http://lists.puredata.info/listinfo/pd-announce
Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
anyone else having GUI trouble seeing sliders and number boxes move/change values?
seem to work ok for me... but i only did a quick test.
10.8.4
just currious, is all the work jonathon wilkes been doing on the pd gui stuff (search, preferences, etc) on OSX expected to be a part of Pd Vanilla?
m
On Mon, Aug 26, 2013 at 1:36 AM, Alexandre Torres Porres porres@gmail.com wrote:
anyone else having GUI trouble seeing sliders and number boxes move/change values?
i'm on mac os 10.7.5
cheers
2013/8/23 Miller Puckette msp@ucsd.edu
Hi all,
Pd version 0.45-0 is available on http://crca.ucsd.edu/~msp/software.htm or via git from sourceforge: git clone git://git.code.sf.net/p/pure-data/pure-data cd pure-data git checkout -b 0.45
As always I'm sure there will be problems here and there - you're welcome to report them on the Pd mailing list (pd-list@iem.at) which is always the fastest way to get me to see them.
cheers Miller
Pd-announce mailing list Pd-announce@iem.at http://lists.puredata.info/listinfo/pd-announce
Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
On 26/08/13 22:03, me.grimm wrote:
just currious, is all the work jonathon wilkes been doing on the pd gui stuff (search, preferences, etc) on OSX expected to be a part of Pd Vanilla?
that is up to Miller, if it fits his plans or not, otherwise it's in extended, or imported manually where that is possible ... which is the same for all developments.
I haven't looked at it yet (altought it sounds like a Good Thing) and anyway I would want to talk with Jonathan first and not just assume it's cool to glom a bunch of his code. There's a downside - once something's sitting in vanilla it's much harder to change than it is out there in extended.
cheers Miller
On Tue, Aug 27, 2013 at 11:24:03AM +0800, Simon Wise wrote:
On 26/08/13 22:03, me.grimm wrote:
just currious, is all the work jonathon wilkes been doing on the pd gui stuff (search, preferences, etc) on OSX expected to be a part of Pd Vanilla?
that is up to Miller, if it fits his plans or not, otherwise it's in extended, or imported manually where that is possible ... which is the same for all developments.
Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
On 08/26/2013 11:24 PM, Simon Wise wrote:
On 26/08/13 22:03, me.grimm wrote:
just currious, is all the work jonathon wilkes been doing on the pd gui stuff (search, preferences, etc) on OSX expected to be a part of Pd Vanilla?
that is up to Miller, if it fits his plans or not, otherwise it's in extended, or imported manually where that is possible ... which is the same for all developments.
I don't think I've posted a diff of it yet, but I can do that tomorrow if there's interest.
A few points:
it's a very small change in c)
sys_gui/sys_vgui calls and replaces them with tcl variables
For the last point, I suppose I could clean it up a bit by putting macros in m_pd.h for all the tcl variables.
-Jonathan
Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
On 08/26/2013 11:24 PM, Simon Wise wrote:
On 26/08/13 22:03, me.grimm wrote:
just currious, is all the work jonathon wilkes been doing on the pd gui stuff (search, preferences, etc) on OSX expected to be a part of Pd Vanilla?
that is up to Miller, if it fits his plans or not, otherwise it's in extended, or imported manually where that is possible ... which is the same for all developments.
Here's a patch for 0.45. I had to change the canvas line item that resizes the comments into a rectangle, but it behaves exactly the same.
As far as I can tell it works as it should. Unless Miller changed something major, the audio and midi dialog prefs should get saved with clicking the "Apply" button. GUI prefs don't yet get saved. I'll add that at some point, but for now it's not to hard to select from the list.
There's still a "Preferences" cascade in the "File" menu that needs to be removed. That's fairly simple to do, but the prefs for startup libs are in there, and since I didn't add those to the prefs dialog I didn't want to remove the corresponding code with my patch.
As always, rice and beans are welcome: https://jwilkes.nfshost.com/donations.php
Best, Jonathan
Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
On 08/28/13 00:47, Jonathan Wilkes wrote:
+#X obj 407 2 pddp/pddplink http://puredata.info/dev/pddp -text pddp
ähm, has pddp made it into Pd-vanilla already?
it might be a good idea to submit multiple patches, each implementing a given functionality/feature, so they can be reviewed committed separately. in case of troubles this also makes it easier to track down the culprit.
gfmsdar IOhannes
On 08/28/2013 03:48 AM, IOhannes m zmölnig wrote:
On 08/28/13 00:47, Jonathan Wilkes wrote:
+#X obj 407 2 pddp/pddplink http://puredata.info/dev/pddp -text pddp
ähm, has pddp made it into Pd-vanilla already?
Oops, those objects are a mistake. Here's a revised patch where I've removed all the PDDP template from the help patches.
it might be a good idea to submit multiple patches, each implementing a given functionality/feature, so they can be reviewed committed separately. in case of troubles this also makes it easier to track down the culprit.
If someone wants feature x but not feature y I'll be happy to break it out for them. If not it's a waste of time-- this is a single dialog window and there's very little chance that three panes which don't affect each other would make problems more difficult to track down.
-Jonathan
gfmsdar IOhannes
Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
There is one little annoyance: if you select a GUI object, right click fro the properties and then go to change on of the fields in the dialog, you can't hit backspace, because this deletes the object and closes the properties dialog. Somehow the backspace takes effect on the canvas instead of the numbers of say changing the values of a slider.
but generally: amazing, thank you miller. :)
m.
Am 23.08.2013 um 21:00 schrieb Miller Puckette msp@ucsd.edu:
Hi all,
Pd version 0.45-0 is available on http://crca.ucsd.edu/~msp/software.htm or via git from sourceforge: git clone git://git.code.sf.net/p/pure-data/pure-data cd pure-data git checkout -b 0.45
As always I'm sure there will be problems here and there - you're welcome to report them on the Pd mailing list (pd-list@iem.at) which is always the fastest way to get me to see them.
cheers Miller
Pd-announce mailing list Pd-announce@iem.at http://lists.puredata.info/listinfo/pd-announce
Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
On 08/27/13 12:30, Max wrote:
There is one little annoyance: if you select a GUI object, right click fro the properties and then go to change on of the fields in the dialog, you can't hit backspace, because this deletes the object and closes the properties dialog. Somehow the backspace takes effect on the canvas instead of the numbers of say changing the values of a slider.
not on my pd-0.45.0. this is on debian with xfce4, focus-follows-mouse policy.
fgmasdr IOhannes
Am 27.08.2013 um 18:57 schrieb IOhannes m zmölnig zmoelnig@iem.at:
On 08/27/13 12:30, Max wrote:
There is one little annoyance: if you select a GUI object, right click fro the properties and then go to change on of the fields in the dialog, you can't hit backspace, because this deletes the object and closes the properties dialog. Somehow the backspace takes effect on the canvas instead of the numbers of say changing the values of a slider.
not on my pd-0.45.0. this is on debian with xfce4, focus-follows-mouse policy.
Are you sure the object is selected? the error is occurring for me on OS X 10.8, haven't tried other platforms. m.
Aha - I can make this malfunction... don't know what's causing it yet. It's quite n abupt surprise when it happens :)
On Tue, Aug 27, 2013 at 08:37:25PM +0200, Max wrote:
Am 27.08.2013 um 18:57 schrieb IOhannes m zmölnig zmoelnig@iem.at:
On 08/27/13 12:30, Max wrote:
There is one little annoyance: if you select a GUI object, right click fro the properties and then go to change on of the fields in the dialog, you can't hit backspace, because this deletes the object and closes the properties dialog. Somehow the backspace takes effect on the canvas instead of the numbers of say changing the values of a slider.
not on my pd-0.45.0. this is on debian with xfce4, focus-follows-mouse policy.
Are you sure the object is selected? the error is occurring for me on OS X 10.8, haven't tried other platforms. m.
Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
it happens consistently in os x 10.7.5.
2a) If it is selected and I try to delete the "5" in "width" it deletes the number box.
2b) If it is not selected, it won't even delete the "5" in "width".
J
On Aug 27, 2013, at 10:53 PM, Miller Puckette wrote:
Aha - I can make this malfunction... don't know what's causing it yet. It's quite n abupt surprise when it happens :)
On Tue, Aug 27, 2013 at 08:37:25PM +0200, Max wrote:
Am 27.08.2013 um 18:57 schrieb IOhannes m zmölnig zmoelnig@iem.at:
On 08/27/13 12:30, Max wrote:
There is one little annoyance: if you select a GUI object, right click fro the properties and then go to change on of the fields in the dialog, you can't hit backspace, because this deletes the object and closes the properties dialog. Somehow the backspace takes effect on the canvas instead of the numbers of say changing the values of a slider.
not on my pd-0.45.0. this is on debian with xfce4, focus-follows-mouse policy.
Are you sure the object is selected? the error is occurring for me on OS X 10.8, haven't tried other platforms. m.
Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
OK.... I think I have a 'fix' but I'm not sure about it (for some reason Pd was throwing key events to the "last canvas on which a text editor had opened" but now I can't see why one would do that, and anyway the 'feature' was broken since 0.43 and I never saw anything wrong... s oI think it's safe just to take that 'feature' out. I think I can just release the fix as 0.45-1 but am going to wait until I'm wider awake. Here's a patch if anyone wants to try it:
From fff0b83e6fb500672a4706487358d612db7fe543 Mon Sep 17 00:00:00 2001
From: Miller Puckette msp@ucsd.edu Date: Tue, 27 Aug 2013 22:01:22 -0700 Subject: [PATCH] Took out the canvas_editing business to try to fix the backspace-in-dialog- getting-sent-to-canvas bug
src/g_canvas.c | 3 --- src/g_canvas.h | 1 - src/g_editor.c | 4 ++-- src/g_rtext.c | 1 - 4 files changed, 2 insertions(+), 7 deletions(-)
diff --git a/src/g_canvas.c b/src/g_canvas.c
index 3d38f1c..54079e3 100644
--- a/src/g_canvas.c
+++ b/src/g_canvas.c
@@ -42,7 +42,6 @@ desktops because the borders have both window title area and menus. */
extern t_pd *newest;
t_class *canvas_class;
int canvas_dspstate; /* whether DSP is on or off */
-t_canvas *canvas_editing; /* last canvas to start text edting */
t_canvas *canvas_whichfind; /* last canvas we did a find in */
t_canvas *canvas_list; /* list of all root canvases */
@@ -725,8 +724,6 @@ void canvas_free(t_canvas *x) t_gobj *y; int dspstate = canvas_suspend_dsp(); canvas_noundo(x);
canvas_editing = 0;
if (canvas_whichfind == x)
canvas_whichfind = 0;
glist_noselect(x);diff --git a/src/g_canvas.h b/src/g_canvas.h index ba6d979..3871444 100644 --- a/src/g_canvas.h +++ b/src/g_canvas.h @@ -336,7 +336,6 @@ struct _parentwidgetbehavior #define CURSOR_EDITMODE_RESIZE 7 EXTERN void canvas_setcursor(t_glist *x, unsigned int cursornum);
-extern t_canvas *canvas_editing; /* last canvas to start text edting */ extern t_canvas *canvas_whichfind; /* last canvas we did a find in */ extern t_canvas *canvas_list; /* list of all root canvases */ extern t_class *vinlet_class, *voutlet_class; diff --git a/src/g_editor.c b/src/g_editor.c index 85628b9..90bb78a 100644 --- a/src/g_editor.c +++ b/src/g_editor.c @@ -2675,8 +2675,8 @@ static void canvas_texteditor(t_canvas *x)
void glob_key(void *dummy, t_symbol *s, int ac, t_atom *av) {
/* canvas_editing can be zero; canvas_key checks for that */
/* canvas_key checks for zero */
}
void canvas_editmode(t_canvas *x, t_floatarg state) diff --git a/src/g_rtext.c b/src/g_rtext.c index 02f7400..f7aa451 100644 --- a/src/g_rtext.c +++ b/src/g_rtext.c @@ -449,7 +449,6 @@ void rtext_select(t_rtext *x, int state) t_canvas *canvas = glist_getcanvas(glist); sys_vgui(".x%lx.c itemconfigure %s -fill %s\n", canvas, x->x_tag, (state? "blue" : "black"));
}
1.7.11.7
cheers Miller
On Tue, Aug 27, 2013 at 11:55:58PM -0400, J Oliver wrote:
it happens consistently in os x 10.7.5.
- create a number box.
2a) If it is selected and I try to delete the "5" in "width" it deletes the number box.
2b) If it is not selected, it won't even delete the "5" in "width".
J
On Aug 27, 2013, at 10:53 PM, Miller Puckette wrote:
Aha - I can make this malfunction... don't know what's causing it yet. It's quite n abupt surprise when it happens :)
On Tue, Aug 27, 2013 at 08:37:25PM +0200, Max wrote:
Am 27.08.2013 um 18:57 schrieb IOhannes m zmölnig zmoelnig@iem.at:
On 08/27/13 12:30, Max wrote:
There is one little annoyance: if you select a GUI object, right click fro the properties and then go to change on of the fields in the dialog, you can't hit backspace, because this deletes the object and closes the properties dialog. Somehow the backspace takes effect on the canvas instead of the numbers of say changing the values of a slider.
not on my pd-0.45.0. this is on debian with xfce4, focus-follows-mouse policy.
Are you sure the object is selected? the error is occurring for me on OS X 10.8, haven't tried other platforms. m.
Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
On 08/27/13 20:37, Max wrote:
Am 27.08.2013 um 18:57 schrieb IOhannes m zmölnig zmoelnig@iem.at:
On 08/27/13 12:30, Max wrote:
There is one little annoyance: if you select a GUI object, right click fro the properties and then go to change on of the fields in the dialog, you can't hit backspace, because this deletes the object and closes the properties dialog. Somehow the backspace takes effect on the canvas instead of the numbers of say changing the values of a slider.
not on my pd-0.45.0. this is on debian with xfce4, focus-follows-mouse policy.
Are you sure the object is selected? the error is occurring for me on OS X 10.8, haven't tried other platforms.
since you mentioned "slider" in the original bugreport, i did my initial test with [hsl], and i *am* sure that the slider is selected when deleting the, say, width-value of the slider in the properties dialog. no problems here.
i now did another test with the traditional numberbox and there i *can* reproduce the problem.
fmasdr IOhannes
Am 28.08.2013 um 09:33 schrieb IOhannes m zmölnig zmoelnig@iem.at:
On 08/27/13 20:37, Max wrote:
Am 27.08.2013 um 18:57 schrieb IOhannes m zmölnig zmoelnig@iem.at:
On 08/27/13 12:30, Max wrote:
There is one little annoyance: if you select a GUI object, right click fro the properties and then go to change on of the fields in the dialog, you can't hit backspace, because this deletes the object and closes the properties dialog. Somehow the backspace takes effect on the canvas instead of the numbers of say changing the values of a slider.
not on my pd-0.45.0. this is on debian with xfce4, focus-follows-mouse policy.
Are you sure the object is selected? the error is occurring for me on OS X 10.8, haven't tried other platforms.
since you mentioned "slider" in the original bugreport, i did my initial test with [hsl], and i *am* sure that the slider is selected when deleting the, say, width-value of the slider in the properties dialog. no problems here.
i now did another test with the traditional numberbox and there i *can* reproduce the problem.
Interesting. I have the phenomenon with any atom which has a properties window.
There is another odd thing: While the properties window opens you can see that "Pd" in the Menu bar changes into "Apple" (?!) Also: the first time you open a properties window after the lauch of Pd it takes significantly longer than subsequent times. (around 5 seconds vs. 0.1 second)
m.
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
Le 27/08/2013 12:30, Max a écrit :
There is one little annoyance: if you select a GUI object, right click fro the properties and then go
to change on of the fields in the dialog, you can't hit backspace, because this deletes the object and closes the properties dialog. Somehow the backspace takes effect on the canvas instead of the numbers of say changing the values of a slider.
but generally: amazing, thank you miller. :)
m.
Am 23.08.2013 um 21:00 schrieb Miller Puckette msp@ucsd.edu:
Hi all,
Pd version 0.45-0 is available on http://crca.ucsd.edu/~msp/software.htm or via git from sourceforge: git clone git://git.code.sf.net/p/pure-data/pure-data cd pure-data git checkout -b 0.45
As always I'm sure there will be problems here and there - you're welcome to report them on the Pd mailing list (pd-list@iem.at) which is
always the
fastest way to get me to see them.
cheers Miller
Pd-announce mailing list Pd-announce@iem.at http://lists.puredata.info/listinfo/pd-announce
Pd-list@iem.at mailing list UNSUBSCRIBE and account-management ->
http://lists.puredata.info/listinfo/pd-list
Pd-list@iem.at mailing list UNSUBSCRIBE and account-management ->
http://lists.puredata.info/listinfo/pd-list Hello,
I had the same problem with one of my patch today : 'properties' on canvas then backspace on a field deleted the canvas. But I can't reproduce it. ++
Jack
Le 27/08/2013 19:15, Jack a écrit :
Le 27/08/2013 12:30, Max a écrit :
There is one little annoyance: if you select a GUI object, right click fro the properties and then
go to change on of the fields in the dialog, you can't hit backspace, because this deletes the object and closes the properties dialog. Somehow the backspace takes effect on the canvas instead of the numbers of say changing the values of a slider.
but generally: amazing, thank you miller. :)
m.
Am 23.08.2013 um 21:00 schrieb Miller Puckette msp@ucsd.edu:
Hi all,
Pd version 0.45-0 is available on
http://crca.ucsd.edu/~msp/software.htm
or via git from sourceforge: git clone git://git.code.sf.net/p/pure-data/pure-data cd pure-data git checkout -b 0.45
As always I'm sure there will be problems here and there - you're
welcome
to report them on the Pd mailing list (pd-list@iem.at) which is
always the
fastest way to get me to see them.
cheers Miller
Pd-announce mailing list Pd-announce@iem.at http://lists.puredata.info/listinfo/pd-announce
Pd-list@iem.at mailing list UNSUBSCRIBE and account-management ->
http://lists.puredata.info/listinfo/pd-list
Pd-list@iem.at mailing list UNSUBSCRIBE and account-management ->
http://lists.puredata.info/listinfo/pd-list Hello,
I had the same problem with one of my patch today : 'properties' on canvas then backspace on a field deleted the canvas. But I can't reproduce it. ++
Jack
I forgot : the problem occurs on : Ubuntu 13.04 Pd 0.45.0test 2 ++
Jack
On 08/23/13 21:00, Miller Puckette wrote:
Hi all,
Pd version 0.45-0 is available on http://crca.ucsd.edu/~msp/software.htm
cool. i started updating the Debian packages...
As always I'm sure there will be problems here and there - you're welcome to report them on the Pd mailing list (pd-list@iem.at) which is always the fastest way to get me to see them.
two smallish notes.
#1 CFLAGS when manually including the Debian-patch "usercflags" (that makes sure that user-specified CFLAGS are honoured by prepending the configure-detected flags rather than appending), a small oversight has occured:
change was made for *hurd* (aka "GNU"). i think this was simply a confusion of "GNU" and "linux".
the attached "usercflags.patch" hopefully gets this right. it /also/ fixes the CFLAGS append/prepend order for cygwin & mingw (though i admit that i have not tested this)
#2 BUILD FAILURES with "-Werror=format-security" when trying to build with the above error-flag (which pays some extra attention to argument passing), the build fails due to two problems. a) a call of "error(buf)" withing bug() is rejected, as the 'error'-function really reads "error(const char *fmt, ...)" and 'buf' is not a format, resp. there are no varargs. the fix is simply to call error("consistency check failed: %s", buf) and not construct the "consitency-failed" string beforehand. cool, this makes the code a little more readable! b) more serious, the new [text] object uses 'pd_error("ouch %s", str)', when pd_error() really needs a pointer to a pd-object... this is a potential crasher bug, as it accesses string memory as objects.
the attached patch "fix_format-security.patch" fixes this as well.
gfmasdr IOhannes
OK, these should be applied and 'pushed' both to branch 0.45 and to 'master'.
cheers Miller
On Tue, Aug 27, 2013 at 07:21:06PM +0200, IOhannes m zmölnig wrote:
On 08/23/13 21:00, Miller Puckette wrote:
Hi all,
Pd version 0.45-0 is available on http://crca.ucsd.edu/~msp/software.htm
cool. i started updating the Debian packages...
As always I'm sure there will be problems here and there - you're welcome to report them on the Pd mailing list (pd-list@iem.at) which is always the fastest way to get me to see them.
two smallish notes.
#1 CFLAGS when manually including the Debian-patch "usercflags" (that makes sure that user-specified CFLAGS are honoured by prepending the configure-detected flags rather than appending), a small oversight has occured:
- on *linux*, the configure-CFLAGS are still appended. instead the
change was made for *hurd* (aka "GNU"). i think this was simply a confusion of "GNU" and "linux".
- otoh. the *hurd* CFLAGS still include "-O6"
the attached "usercflags.patch" hopefully gets this right. it /also/ fixes the CFLAGS append/prepend order for cygwin & mingw (though i admit that i have not tested this)
#2 BUILD FAILURES with "-Werror=format-security" when trying to build with the above error-flag (which pays some extra attention to argument passing), the build fails due to two problems. a) a call of "error(buf)" withing bug() is rejected, as the 'error'-function really reads "error(const char *fmt, ...)" and 'buf' is not a format, resp. there are no varargs. the fix is simply to call error("consistency check failed: %s", buf) and not construct the "consitency-failed" string beforehand. cool, this makes the code a little more readable! b) more serious, the new [text] object uses 'pd_error("ouch %s", str)', when pd_error() really needs a pointer to a pd-object... this is a potential crasher bug, as it accesses string memory as objects.
the attached patch "fix_format-security.patch" fixes this as well.
gfmasdr IOhannes
Author: Paul Brossier piem@debian.org Description: do not overwrite user cflags, add them *after* hardcoded ones --- puredata.orig/configure.ac +++ puredata/configure.ac @@ -42,7 +42,7 @@ if test "x${ANDROID}" = "xno"; then LINUX=yes portaudio=yes
CFLAGS="$CFLAGS -O3 -funroll-loops -fomit-frame-pointer"
fi EXTERNAL_CFLAGS="-fPIC" EXTERNAL_LDFLAGS="-Wl,--export-dynamic -shared -fPIC"CFLAGS="-O3 -funroll-loops -fomit-frame-pointer $CFLAGS"
@@ -50,7 +50,7 @@ ;; *-*-gnu*) HURD=yes
- CFLAGS="-O6 -funroll-loops -fomit-frame-pointer $CFLAGS"
- CFLAGS="-O3 -funroll-loops -fomit-frame-pointer $CFLAGS" EXTERNAL_CFLAGS="-fPIC" EXTERNAL_LDFLAGS="-Wl,--export-dynamic -shared -fPIC" EXTERNAL_EXTENSION=pd_linux
@@ -62,7 +62,8 @@ # to make the final linking phase use g++ # asio=yes portaudio=yes
- CFLAGS="$CFLAGS -O3 -funroll-loops -fomit-frame-pointer -DWINVER=0x0501 -D_WIN32_WINNT=0x0501"
- CFLAGS="-O3 -funroll-loops -fomit-frame-pointer -DWINVER=0x0501
+-D_WIN32_WINNT=0x0501 $CFLAGS" # ASIO is a C++ library, so if its included, then use g++ to build CC=g++ EXTERNAL_CFLAGS="-mms-bitfields" @@ -73,7 +74,7 @@ WINDOWS=yes CYGWIN=yes portaudio=yes
- CFLAGS="$CFLAGS -O3 -funroll-loops -fomit-frame-pointer"
- CFLAGS="-O3 -funroll-loops -fomit-frame-pointer $CFLAGS" EXTERNAL_CFLAGS= EXTERNAL_LDFLAGS="-Wl,--export-dynamic -shared -lpd" EXTERNAL_EXTENSION=dll
Author: IOhannes m zm??lnig Description: printf-like varargs functions must have a proper format; use "error('%s', str);" rather than "error(str);" also pd_error() requires an instance-pointer as the first argument --- puredata.orig/src/s_print.c +++ puredata/src/s_print.c @@ -282,12 +282,11 @@ va_list ap; t_int arg[8]; int i;
- strcpy(buf, "consistency check failed: "); va_start(ap, fmt);
- vsnprintf(buf+strlen(buf), MAXPDSTRING-1, fmt, ap);
- vsnprintf(buf, MAXPDSTRING-1, fmt, ap); va_end(ap);
- error(buf);
- error("consistency check failed: %s", buf);
}
/* this isn't worked out yet. */
--- puredata.orig/src/x_text.c +++ puredata/src/x_text.c @@ -1174,7 +1174,7 @@ } else {
pd_error("text sequence: unknown flag '%s'...",
pd_error(x, "text sequence: unknown flag '%s'...", argv->a_w.w_symbol->s_name); } argc--; argv++;
Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
Am 23.08.2013 um 21:00 schrieb Miller Puckette msp@ucsd.edu:
As always I'm sure there will be problems here and there - you're welcome to report them on the Pd mailing list (pd-list@iem.at) which is always the fastest way to get me to see them.
There is a small bug for users of multiple monitors or resolutions (tested only on osx, but i guess that's platform independent).
Some Pd versions ago the patches saved on a bigger resolution screen were opening off-screen when opened with a smaller resolution or only one monitor. This has been addressed since.
In 0.45 if you have the main monitor on the right and a secondary on the left (thus having negative x values) if you save this patch on the secondary and open it again it will appear on the primary monitor, even if you have the second still attached.
Subpatches and abstractions do show an even stranger behavior: if you have a patch open on the primary monitor and drag a subpatch on the secondary monitor (placement relative to the primary monitor doesn't matter here) then save and close the subpatch, open the subpatch again: you can't. It has become inaccessible. In a performance situation this is not what you hope for, because you will only get that window back by re-opening the patch.
m.
these bugs were also there on windows and ubuntu, I've always experienced them. They're small things, but it does bother a little, because pd patches are limited to 1 monitor - unless one likes draging windows every time they get open.
Am 23.08.2013 um 21:00 schrieb Miller Puckette msp@ucsd.edu:
As always I'm sure there will be problems here and there - you're welcome to report them on the Pd mailing list (pd-list@iem.at) which is always the fastest way to get me to see them.
There is a small bug for users of multiple monitors or resolutions (tested only on osx, but i guess that's platform independent).
Some Pd versions ago the patches saved on a bigger resolution screen were opening off-screen when opened with a smaller resolution or only one monitor. This has been addressed since.
In 0.45 if you have the main monitor on the right and a secondary on the left (thus having negative x values) if you save this patch on the secondary and open it again it will appear on the primary monitor, even if you have the second still attached.
Subpatches and abstractions do show an even stranger behavior: if you have a patch open on the primary monitor and drag a subpatch on the secondary monitor (placement relative to the primary monitor doesn't matter here) then save and close the subpatch, open the subpatch again: you can't. It has become inaccessible. In a performance situation this is not what you hope for, because you will only get that window back by re-opening the patch.
m.
Well, not having a way to get a dual screen Mac I can't test this but here's a guess - it might still be that Pd refuses to allow negative window locations and so throws the abstraction below the visible portion of the smaller of the two screens.
I don't know if Tcl/TK has any way of dealing with this.... I'll have a look but probably can't fix this very quickly.
cheers Miller
On Wed, Aug 28, 2013 at 07:56:58PM +0200, Max wrote:
Am 23.08.2013 um 21:00 schrieb Miller Puckette msp@ucsd.edu:
As always I'm sure there will be problems here and there - you're welcome to report them on the Pd mailing list (pd-list@iem.at) which is always the fastest way to get me to see them.
There is a small bug for users of multiple monitors or resolutions (tested only on osx, but i guess that's platform independent).
Some Pd versions ago the patches saved on a bigger resolution screen were opening off-screen when opened with a smaller resolution or only one monitor. This has been addressed since.
In 0.45 if you have the main monitor on the right and a secondary on the left (thus having negative x values) if you save this patch on the secondary and open it again it will appear on the primary monitor, even if you have the second still attached.
Subpatches and abstractions do show an even stranger behavior: if you have a patch open on the primary monitor and drag a subpatch on the secondary monitor (placement relative to the primary monitor doesn't matter here) then save and close the subpatch, open the subpatch again: you can't. It has become inaccessible. In a performance situation this is not what you hope for, because you will only get that window back by re-opening the patch.
m.
Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list