hi,
i've transformed my former "filterbank resynthesis patch" so that you can now set the number of filters used for filtering/resynthesis.
by making this, i discovered the followings faulty behaviours with messaging :
a/ "find" & "cut" messages make PD crash if the patch is not displayed.
the message is : "Consistency check failed: glist_findrtext"
some people reported that it crashes "sometimes", i guess that's because the patch whose messages are sent to were not open.
b/ "pop" message doesn't work :
the message is : "Consistency check failed: gstack_pop"
c/ some errors is Henry Damien's doc ( which was a precious help, i must say ) :
- there's no "editmode 0" or "editmode 1",
it's rather a toggle and the message is only "editmode"
** unlike "editmode", "vis" message requires an argument : "vis 0" or "vis 1".
i'm using test17.
cheers,
yves
hi Yves,
Yves Degoyon wrote: ...
faulty behaviours with messaging :
a/ "find" & "cut" messages make PD crash if the patch is not displayed.
the find feature has been fixed in test24.
While I am not sure why would you want to send 'cut' to a closed
window (in which there is nothing selected), this should not
crash Pd. But it is probably hard to fix properly, esp. prior to
making glist'-handling stabilize for good (properly: taking into account canvas' self-destruction too)... but well, invisibleglists' are real pain for me (vide rabin~)...
b/ "pop" message doesn't work :
the message is : "Consistency check failed: gstack_pop"
it does, but one has to push things first to be able to pop them later! Actually, 'pop' message is used for creating a new main canvas via: '; pd filename <fname> <dir>; #N canvas; pop 1'
- there's no "editmode 0" or "editmode 1",
it's rather a toggle and the message is only "editmode"
more precisely: there is 'editmode 1' (always entering editmode), but 'editmode 0' is interpreted as 'editmode' (a toggle) -- seems like a bug, esp. taking into account handling of a 'dirty flag'...
Krzysztof
hi list, i am living now with Pd for quite a long time. one of the bothersome things is the search for externals, esp. for something like a general collection of all externals... sometimes new externals are only posted once and hidden on a website. (shame on me, but i didŽnt even realize that there was a iem-lib2). so i would highly support any work that is done in this direction, to collect, update and distribute externals for the different OS, maybe give short description. once this is done, only new externals would have to be added (which still is enough / to much work for one person). i think maybe iem could do this, or pure-data.org, or i will do it myself, ... or was this already done by someone? sme.
i do beleive that this is the goal of pure-data.org
On Wed, 29 May 2002, sme wrote:
hi list, i am living now with Pd for quite a long time. one of the bothersome things is the search for externals, esp. for something like a general collection of all externals... sometimes new externals are only posted once and hidden on a website. (shame on me, but i did´nt even realize that there was a iem-lib2). so i would highly support any work that is done in this direction, to collect, update and distribute externals for the different OS, maybe give short description. once this is done, only new externals would have to be added (which still is enough / to much work for one person). i think maybe iem could do this, or pure-data.org, or i will do it myself, ... or was this already done by someone? sme.
Josh .. Yoshi .. Joschi .. http://eds.org/~joschi/vitriolix
On pure-data.org, you can post your objects (externals) as well as externals you find of other people. I post as many as I can find, myself, but this is far from complete. Also, not all of the objects posted work with modern versions of Pd.
The pure data base at IEM is much more ambitious. Not only do they seek out *every* object they can find, their indexing scheme allows you to find specific objects within libraries.
On Mon, 2002-06-24 at 16:00, Josh Steiner wrote:
i do beleive that this is the goal of pure-data.org
On Wed, 29 May 2002, sme wrote:
hi list, i am living now with Pd for quite a long time. one of the bothersome things is the search for externals, esp. for something like a general collection of all externals... sometimes new externals are only posted once and hidden on a website. (shame on me, but i did´nt even realize that there was a iem-lib2). so i would highly support any work that is done in this direction, to collect, update and distribute externals for the different OS, maybe give short description. once this is done, only new externals would have to be added (which still is enough / to much work for one person). i think maybe iem could do this, or pure-data.org, or i will do it myself, ... or was this already done by someone? sme.
jfm3 jfm3@ouroboros-complex.org wrote:
On pure-data.org, you can post your objects (externals) as well as externals you find of other people. I post as many as I can find, myself, but this is far from complete. Also, not all of the objects posted work with modern versions of Pd.
The pure data base at IEM is much more ambitious. Not only do they seek out *every* object they can find, their indexing scheme allows you to find specific objects within libraries.
This (http://www.soundobject.org/software/impacts.tgz) doesn't seem to be included yet.
mik
FWIW, I'd like to see consistent documentation of the posted externals, as possible.
The docs I wrote and posted a few days ago might serve as a template. There would be clear information as to the external's inlets, outlets, and arguments (something i realized after writing and posting those), and a brief description of what it does. Such documentation would go a long way toward helping users (especially new ones) wade through the abundance of available objects to find the ones they need.
The documentation would also be quite useful for the built-in items, since I've found going through the online help items a bit cumbersome. I know someone here (I forget whom) is working on updating the help texts. Perhaps these projects could work together.
Does that seem like a good idea?
I could start soon on documenting some... after I finish driving cross-country to move to Berkeley, CA, over the next couple of weeks.
On Mon, 2002-06-24 at 16:00, Josh Steiner wrote:
i do beleive that this is the goal of pure-data.org
On Wed, 29 May 2002, sme wrote:
hi list, i am living now with Pd for quite a long time. one of the bothersome things is the search for externals, esp. for something like a general collection of all externals... sometimes new externals are only posted once and hidden on a website. (shame on me, but i did´nt even realize that there was a iem-lib2). so i would highly support any work that is done in this direction, to collect, update and distribute externals for the different OS, maybe give short description. once this is done, only new externals would have to be added (which still is enough / to much work for one person). i think maybe iem could do this, or pure-data.org, or i will do it myself, ... or was this already done by someone? sme.
Josh .. Yoshi .. Joschi .. http://eds.org/~joschi/vitriolix
the problem is even worse on my side, in fact, 0.24 crashes with most of my patches, especially with filterbank here attached.
sme wrote:
hi yves, krzystof, i also thought it was fixed, but is not. pd still crashes. sÃme.
the find feature has been fixed in test24.
Krzysztof Czaja wrote:
hi Yves,
Yves Degoyon wrote: ...
faulty behaviours with messaging :
a/ "find" & "cut" messages make PD crash if the patch is not displayed.
the find feature has been fixed in test24.
i dunno, as said before, pd crashes when loading my patches thus i cannot test the find feature.
While I am not sure why would you want to send 'cut' to a closed window (in which there is nothing selected),
krysztof ! ! of course there's a "find" message before the "cut", so something is or should be selected.
.
b/ "pop" message doesn't work :
the message is : "Consistency check failed: gstack_pop"
it does, but one has to push things first to be able to pop them
???? couldn't find any "push" message. here's what pd says :
error: canvas: no method for 'push' ... you might be able to track this down from the Find menu.
cheers,
yves
hi Yves
Yves Degoyon wrote: ...
krysztof ! ! of course there's a "find" message before the "cut", so something is or should be selected.
but you have written about cutting in an invisible window, while 'find' always opens the window containing an object found.
???? couldn't find any "push" message.
Yves! -- of course there is no 'push' message, it is the 'canvas' message that does the pushing (you are a fluent code reader -- search for pd_pushsym() call in canvas_new()...)
Krzysztof
Krzysztof Czaja wrote:
hi Yves
Yves Degoyon wrote: ...
krysztof ! ! of course there's a "find" message before the "cut", so something is or should be selected.
but you have written about cutting in an invisible window, while 'find' always opens the window containing an object found.
curious, it does not open the window on test17 ( maybe, it tries and crashes ?? )
???? couldn't find any "push" message.
Yves! -- of course there is no 'push' message, it is the 'canvas' message that does the pushing (you are a fluent code reader -- search for pd_pushsym() call in canvas_new()...)
ok, but i was speaking from a patch-maker point of view and i simply thought "pop" message would actually "pop-up" the window.
i'm not currently speaking pd's internal code language as you do.
Krzysztof
hi,
i've finally got test24 up and running ! !
the error was dued to the loading of former iemgui library which is now included into pd.
iemlib-bundled version still compiles with test24 BUT it provokes lots of crashes, so, i guess it should be removed from delivery.
so, i could be able to test dynamic creation and it's true it's been somewhat fixed ( thank you, Miller btw ) :
it doesn't crash anymore and "find" opens the target patch, ok.
but, the "cut" messages sent to the patch BEFORE it's actually open are lost : it takes some time between the first find message and the opening of the window, in that time, you can't do any "cut" operations....mm, still annoying.
you can test this with the patch i've posted ( ten times ?? ) here.
cheers,
yves.
Yves Degoyon wrote:
Krzysztof Czaja wrote:
hi Yves
Yves Degoyon wrote: ...
krysztof ! ! of course there's a "find" message before the "cut", so something is or should be selected.
but you have written about cutting in an invisible window, while 'find' always opens the window containing an object found.
curious, it does not open the window on test17 ( maybe, it tries and crashes ?? )
???? couldn't find any "push" message.
Yves! -- of course there is no 'push' message, it is the 'canvas' message that does the pushing (you are a fluent code reader -- search for pd_pushsym() call in canvas_new()...)
ok, but i was speaking from a patch-maker point of view and i simply thought "pop" message would actually "pop-up" the window.
i'm not currently speaking pd's internal code language as you do.
Krzysztof
-
jfm3 -
Joseph Zitt -
Josh Steiner -
Krzysztof Czaja -
mik -
sme -
Yves Degoyon