Hi all,
I re-merged Iohannes's changes, found the "-noadc" problem with ALSA, and fixed the netreceive object to handle UDP correctly... it's on
http://crca.ucsd.edu/~msp/software.html
(for linux only, sorry...) It's still just an attempt to collect all the changes people have proposed together. The interesting stuff I want to add for 0.35 comes later...
cheers Miller
hi,
-noadc problem is gone, and resampling works great!
What bothers me now, is that among the new sources I found g_numbox.c (not merged yet). Is this one going to replace number box we currently have? I hope not...
Anyway, I checked [my_numbox] aka [nbx] (compiled as an external) and found it to be very useful -- especially the ability to have it really big and in colors. Also log-scale feature is nice, but I think it needs to be adjusted to work better if used as a `fader' connected to [*~ 0] (replacing [gatom]->[dbtopow]). I mean, when tracking mouse movements, it better be more mouse-friendly and use pow(10,0.1) as a base, i.e. 1.25892541179, instead of current radix 1.01.
Another problem is hardcoding of a default font size to 10pt, instead of consulting a canvas with canvas_getfont(canvas_getcurrent()) call in creation routine (making this right means also calculating a height value, instead of using hardcoded default of 12 pixels).
There is no auto-resizing and no alt-click (at least alt-click can easily be borrowed from gatom code).
There is also a small bug in my_numbox_list(). Arrow-movement, as it
seems, applies to a ghost' object, or, strictly speaking, there is no attempt made to redraw the object (thel' flag variable handling should
be corrected).
Btw. why do we need all those class creators for gui objects? After all there is only _one_ namespace for internals, externals and abstractions. We have now: bng, hsl, hslider, hdl, rdb, radiobut, radiobutton, tgl, toggle, cnv, my_canvas, vsl, vslider, vu, and many more comming. They are stored with short names in patch files, so what is the purpose of the longer names? Moreover, virtually nobody types them in, but rather uses menu/shortcuts to put them on a page.
Krzysztof
Miller Puckette wrote: ...
I re-merged Iohannes's changes, found the "-noadc" problem with ALSA, and fixed the netreceive object to handle UDP correctly... it's on
...
changes people have proposed together. The interesting stuff I want to add for 0.35 comes later...
... and the suspense grows :)
Hi Krzysztof,
I'm planning to test g_numbox.c (make sure it works ok on NT and OSX) and if I can't see any reason not to, just to use it. One thing I haven't looked at is how Thomas Musil (who wrote it) is dealing with font sizes...
I also have to think about the problem of everything seeming to have two names... shouldn't be necessary...
I don't like auto-resizing very much (although I have to use it myself sometimes) -- it might make it harder to port to a future toolkit. I had forgotten about alt click, thanks for reminding me about that one..
cheers Miller
On Fri, Nov 30, 2001 at 08:09:49PM +0100, Krzysztof Czaja wrote:
hi,
-noadc problem is gone, and resampling works great!
What bothers me now, is that among the new sources I found g_numbox.c (not merged yet). Is this one going to replace number box we currently have? I hope not...
Anyway, I checked [my_numbox] aka [nbx] (compiled as an external) and found it to be very useful -- especially the ability to have it really big and in colors. Also log-scale feature is nice, but I think it needs to be adjusted to work better if used as a `fader' connected to [*~ 0] (replacing [gatom]->[dbtopow]). I mean, when tracking mouse movements, it better be more mouse-friendly and use pow(10,0.1) as a base, i.e. 1.25892541179, instead of current radix 1.01.
Another problem is hardcoding of a default font size to 10pt, instead of consulting a canvas with canvas_getfont(canvas_getcurrent()) call in creation routine (making this right means also calculating a height value, instead of using hardcoded default of 12 pixels).
There is no auto-resizing and no alt-click (at least alt-click can easily be borrowed from gatom code).
There is also a small bug in my_numbox_list(). Arrow-movement, as it seems, applies to a
ghost' object, or, strictly speaking, there is no attempt made to redraw the object (thel' flag variable handling should be corrected).Btw. why do we need all those class creators for gui objects? After all there is only _one_ namespace for internals, externals and abstractions. We have now: bng, hsl, hslider, hdl, rdb, radiobut, radiobutton, tgl, toggle, cnv, my_canvas, vsl, vslider, vu, and many more comming. They are stored with short names in patch files, so what is the purpose of the longer names? Moreover, virtually nobody types them in, but rather uses menu/shortcuts to put them on a page.
Krzysztof
Miller Puckette wrote: ...
I re-merged Iohannes's changes, found the "-noadc" problem with ALSA, and fixed the netreceive object to handle UDP correctly... it's on
...
changes people have proposed together. The interesting stuff I want to add for 0.35 comes later...
... and the suspense grows :)
hi,
first of all, I do like [nbx] and do see its usefulness. But there are few concerns:
Dangers of having hidden send/receive in a core object class.
Compatibility across releases (is the parameter set of [nbx] ready
to be freezed?).
Portability across environments (mainly font issues).
Adjustability to other gui frontends (well, parting from tcl/tk
seems more and more improbable, anyway...)
If they are somehow addressed, then we will have another nice tool.
Btw. performance is not an issue -- or, rather, [nbx] should perform better, because update requests for gatom are passed through a lot of generic text object routines, much too many to be efficient.
Krzysztof
Miller Puckette wrote: ...
I'm planning to test g_numbox.c (make sure it works ok on NT and OSX) and if I can't see any reason not to, just to use it. One thing I haven't
Krzysztof Czaja hat gesagt: // Krzysztof Czaja wrote:
There is no auto-resizing and no alt-click (at least alt-click can easily be borrowed from gatom code).
Just a quick question: What does alt-click normally do? I can't use it, because my window manager catches it to move any window.
ciao,
__ __
Frank Barknecht ____ ______ ____ __ trip\ \ / /wire ______
/ __// __ /__/ __// // __ \ / / __ \ ___\
/ / / ____/ / / / // ____// /\ \ ___\____ \
/_/ /_____/ /_/ /_//_____// / \ \_____\_____
/_/ _\
I read:
Just a quick question: What does alt-click normally do? I can't use it, because my window manager catches it to move any window.
it resets number boxes 0
regards
x
... and this raises a harder question... I just found out that "gnome" grabs alt-left-clicks, which I had always thought a polite window manager should pass to the application. But the "data structure" editor also uses alt-clicks, in a much more essential way than the number box does (alt clicking a number box acts like a "mute", and toggles between zero and the last nonzero value.)
I don't know of a good way to prevent window managers from taking over various clicks... is there any standard way that applications and window managers can negotiate this? Or any standard notion of which mouse actions should belong to an application?
thanks to anyone who can shed light on this one... Miller
On Sat, Dec 01, 2001 at 12:39:07PM +0100, Frank Barknecht wrote:
Krzysztof Czaja hat gesagt: // Krzysztof Czaja wrote:
There is no auto-resizing and no alt-click (at least alt-click can easily be borrowed from gatom code).
Just a quick question: What does alt-click normally do? I can't use it, because my window manager catches it to move any window.
ciao,
__ __Frank Barknecht ____ ______ ____ __ trip\ \ / /wire ______ / __// __ /__/ __// // __ \ / / __ \ ___\ / / / ____/ / / / // ____// /\ \ ___\____ \ /_/ /_____/ /_/ /_//_____// / \ \_____\_____
/_/ _\
hi miller, hi list,
kde for example also uses alt-leftclick to move windows by clicking elsewhere in the window (and not to bring them to front) at least in kde this keybinding can changed, though its the default. what about "triple" combinations ?? (i.e. ctrl-shift-click) ?
greets,
chris
Am Samstag, 1. Dezember 2001 21:07 schrieb Miller Puckette:
... and this raises a harder question... I just found out that "gnome" grabs alt-left-clicks, which I had always thought a polite window manager should pass to the application. But the "data structure" editor also uses alt-clicks, in a much more essential way than the number box does (alt clicking a number box acts like a "mute", and toggles between zero and the last nonzero value.)
I don't know of a good way to prevent window managers from taking over various clicks... is there any standard way that applications and window managers can negotiate this? Or any standard notion of which mouse actions should belong to an application?
thanks to anyone who can shed light on this one... Miller
On Sat, Dec 01, 2001 at 12:39:07PM +0100, Frank Barknecht wrote:
Krzysztof Czaja hat gesagt: // Krzysztof Czaja wrote:
There is no auto-resizing and no alt-click (at least alt-click can easily be borrowed from gatom code).
Just a quick question: What does alt-click normally do? I can't use it, because my window manager catches it to move any window.
ciao,
__ __Frank Barknecht ____ ______ ____ __ trip\ \ / /wire ______ / __// __ /__/ __// // __ \ / / __ \ ___
/ / / ____/ / / / // ____// /\ \ ___\____
/_/ /_____/ /_/ /_//_____// / \ \_____\_____
/_/ _\
hmm...
i use gnome and the sawfish window manager and left clicks work for me. it's probably because a long time ago i setup sawfish to use the (otherwise useless) windows key as the meta-key for window actions (it seemed to make more sense and is very handy). if anyone is interested, they can do this in gnome configuration manager -> sawfish -> shortcuts.
pix.
On Sat, 1 Dec 2001 12:07:54 -0800 Miller Puckette mpuckett@man104-1.ucsd.edu wrote:
... and this raises a harder question... I just found out that "gnome" grabs alt-left-clicks, which I had always thought a polite window manager should pass to the application. But the "data structure" editor also uses alt-clicks, in a much more essential way than the number box does (alt clicking a number box acts like a "mute", and toggles between zero and the last nonzero value.)
I don't know of a good way to prevent window managers from taking over various clicks... is there any standard way that applications and window managers can negotiate this? Or any standard notion of which mouse actions should belong to an application?
thanks to anyone who can shed light on this one... Miller
On Sat, Dec 01, 2001 at 12:39:07PM +0100, Frank Barknecht wrote:
Krzysztof Czaja hat gesagt: // Krzysztof Czaja wrote:
There is no auto-resizing and no alt-click (at least alt-click can
easily
be borrowed from gatom code).
Just a quick question: What does alt-click normally do? I can't use
it,
because my window manager catches it to move any window.
ciao,
__ __Frank Barknecht ____ ______ ____ __ trip\ \ / /wire ______ / __// __ /__/ __// // __ \ / / __ \ ___\ / / / ____/ / / / // ____// /\ \ ___\____
/_/ /_____/ /_/ /_//_____// / \\_____\_____\
/_/ \_\
im new to pd, but have been using max.msp for quite some time... i have an issue with the fact that there is no undo button, especially since the method for creating patch cords is so akin to the method to delete them.... am i missing something??? thanks _lgb
Do You Yahoo!? Buy the perfect holiday gifts at Yahoo! Shopping. http://shopping.yahoo.com
until now undo is not implemented,
but you can prevent cords from being (accidentally) deleted
if you activate "Protect" under the the "Edit" menu.
you can also toggle the protect mode with the keyoard
shortkey "ctrl"+"r".
norbert
Get your FREE download of MSP Pure Data at http://iem.kug.ac.at/pd
On Sat, 1 Dec 2001, logrey beam wrote:
im new to pd, but have been using max.msp for quite some time... i have an issue with the fact that there is no undo button, especially since the method for creating patch cords is so akin to the method to delete them.... am i missing something??? thanks _lgb
Do You Yahoo!? Buy the perfect holiday gifts at Yahoo! Shopping. http://shopping.yahoo.com
-
Christian Klippel -
CK -
Frank Barknecht -
Krzysztof Czaja -
logrey beam -
Miller Puckette -
Norbert Math -
pix