Hi list, I am trying since two days to understand why it doesnt work. I make a comparison of two values and want to bang an action if the signal is lower/higher/equal a tolerance value. The operator objects give the right values. but everytime the status changes (eg. from lower to higher) the old status would be printed once (eg. lower printed even if its higher, higher will be printed,too). I dont know why, even after reading the help patches. Maybe somebody of you could help me solving this problem. I attached the the .pd and would really appreciate your help. thanks in advance
Christian Klotz Kunsthochschule Kassel / School of Arts Kassel ::digitalpool
Use more a lot more |trigger| objects in your patches. I think the source of your problem resides in how your control signal is routed and distributed.
tom
----- Original Message ----- From: "Christian Klotz" list_pd@deviation.de To: "PD List" pd-list@iem.at Sent: Monday, August 16, 2004 3:31 PM Subject: [PD] operator/spigot problem - action repeats unintentional
Hi list, I am trying since two days to understand why it doesnt work. I make a comparison of two values and want to bang an action if the signal is lower/higher/equal a tolerance value. The operator objects give the right values. but everytime the status changes (eg. from lower to higher) the old status would be printed once (eg. lower printed even if its higher, higher will be printed,too). I dont know why, even after reading the help patches. Maybe somebody of you could help me solving this problem. I attached the the .pd and would really appreciate your help. thanks in advance
Christian Klotz Kunsthochschule Kassel / School of Arts Kassel ::digitalpool
Hallo, Thomas Ouellet Fredericks hat gesagt: // Thomas Ouellet Fredericks wrote:
Use more a lot more |trigger| objects in your patches. I think the source of your problem resides in how your control signal is routed and distributed.
It surely is. The behaviour of the collision detection region in the patch is totally undefined. Like the first example in the attached patch: No way to tell what's going to happen.
Triggers are a must here, as they are everywhere where numbers are added.
Frank Barknecht _ ______footils.org__
Frank Barknecht wrote:
Triggers are a must here, as they are everywhere where numbers are added.
i'll second this and would even say: "Triggers are a must everywhere where more than one message-connection is made from a single outlet"
ever since trigger has known the "anything" directive this is true for sure. (there *were* reasons for not using [trigger] before, but these times are gone)
i guess 90% of all buggy pd-patches are buggy because of the not-use of [trigger]
probably we should add a "[trigger] is your friend" line to the pd-splash screen.... or the possibility to connect several objects to one message-outlet should be banned entirely - or at least make it the default behaviour and allow the user to switch it off (like the famous "prevent line deletion" short-key)
mfg..asdr IOhannes
probably we should add a "[trigger] is your friend" line to the pd-splash screen.... or the possibility to connect several objects to one message-outlet
btw, does there exist an equivalent of max's "pak" in pure-data? it doesnt seem to be in maxlib or cyclone (that is, a "pack" that outputs upon recieving something from any inlet..)
should be banned entirely - or at least make it the default behaviour
cdr wrote:
probably we should add a "[trigger] is your friend" line to the pd-splash screen.... or the possibility to connect several objects to one message-outlet
oops, this should of course read "to connect one message-outlet to several objects" (although i think it is clear what i meant)
btw, does there exist an equivalent of max's "pak" in pure-data?
none that i know of. i guess this is because there is no *real* use for this object apart from getting the code more readable. but probably i'll add it to zexy to be a companion of [nop~]...
mfg..ar IOhannes
IOhannes m zmoelnig wrote:
cdr wrote:
btw, does there exist an equivalent of max's "pak" in pure-data?
none that i know of. i guess this is because there is no *real* use for this object apart from getting the code more readable. but probably i'll add it to zexy to be a companion of [nop~]...
oops, i misunderstood the concept of [pak] completely... thanks frank for opening my eyes.
mfg.as.arde IOhannes
hi i too have compiled the OSCx external but i am unsure how to test it with another application. Does anyone have a good way to test OSCx in pd with Supercollider or MAX/msp?
cheers
Patrick Pagano,B.S. M.F.A candidate http://www.digitalworlds.ufl.edu http://www.screwmusicforever.com/SHREESWIFT
Hallo, Patrick Pagano hat gesagt: // Patrick Pagano wrote:
i too have compiled the OSCx external but i am unsure how to test it with another application. Does anyone have a good way to test OSCx in pd with Supercollider or MAX/msp?
Max has OSC objects using the same syntax as the Pd OSC externals. Supercollider is OSC all the way. To test Pd with them, just open the OSC.pd help file and edit it to suit your (port and hostname) needs.
Frank Barknecht _ ______footils.org__
On Tue, 17 Aug 2004, IOhannes m zmoelnig wrote:
cdr wrote:
btw, does there exist an equivalent of max's "pak" in pure-data?
none that i know of. i guess this is because there is no *real* use for this object apart from getting the code more readable.
And, as we know, because Pd users are real men, they can put up with the reduced readability, and they don't complain. [pak] is for sissies.
...
Sorry, but readability is for *real*, no matter how intangible it is. I hope the community becomes more sensitive to those issues in the future.
Mathieu Bouchard http://artengine.ca/matju
Mathieu Bouchard wrote:
On Tue, 17 Aug 2004, IOhannes m zmoelnig wrote:
And, as we know, because Pd users are real men, they can put up with the reduced readability, and they don't complain. [pak] is for sissies.
i have to object to being called either "sissy" or "real man". my gender is not an issue to this list.
Sorry, but readability is for *real*, no matter how intangible it is. I
as for readability, you are surely right. that is why i use [trigger]s everywhere (even if message-subtrees are "asynchronous"), this is why there is a [nop~] object, --- my, am i cool -- and this why i really have considered including a [pak] object "like i understood it at the first glance": like an upside-down trigger (e.g. several inlets that immediately are output to one outlet (not forming a list; just multiplexing the messages))
but thank you for pointing this out and please accept my apologies for my improper way of writing.
mfg.adsr IOhannes
Hallo, cdr hat gesagt: // cdr wrote:
btw, does there exist an equivalent of max's "pak" in pure-data? it doesnt seem to be in maxlib or cyclone (that is, a "pack" that outputs upon recieving something from any inlet..)
I don't know anything like that, but you can build it using [t b a] objects. You could put this into an abstraction using [pack $1 $2 ...] if you need it more than once. Or build several abstractions, pak2, pak3, pak4, ... for paks using various argument counts.
Frank Barknecht _ ______footils.org__
On Tue, 17 Aug 2004 11:37:20 +0200, Frank Barknecht fbar@footils.org
wrote:
I don't know anything like that, but you can build it using [t b a] objects. You could put this into an abstraction using [pack $1 $2 ...] if you need it more than once. Or build several abstractions, pak2, pak3, pak4, ... for paks using various argument counts.
Ciao
IT would be nice if you could build abstractions with a variable number of
inlets/outlets, depending on the arguments
ciao Gerard
On Tue, 17 Aug 2004, Frank Barknecht wrote:
cdr hat gesagt: // cdr wrote:
btw, does there exist an equivalent of max's "pak" in pure-data? it doesnt seem to be in maxlib or cyclone (that is, a "pack" that outputs upon recieving something from any inlet..)
I don't know anything like that, but you can build it using [t b a] objects. You could put this into an abstraction using [pack $1 $2 ...] if you need it more than once. Or build several abstractions, pak2, pak3, pak4, ... for paks using various argument counts.
In GridFlow there is a [@pack] that does it like that because I realized that most of the time I actually wanted all inlets to be hot. Then you prolly want to convert the resulting grid to a list using [@export_list]. Note that this will only handle integers. GridFlow 0.7.7 allows an extra argument to [@pack] for specifying floats, but you can't make it handle symbols.
Mathieu Bouchard http://artengine.ca/matju
hi cdr,
cc has no pak, because pak is one of the new, 4.5 additions, and I have not yet got around checking how these actually behave.
However, there is bondo in cc, so you can use the equivalent (unless you wanted initializers...) [bondo]->[pack].
Krzysztof
cdr wrote: ...
btw, does there exist an equivalent of max's "pak" in pure-data? it doesnt seem to be in maxlib or cyclone (that is, a "pack" that outputs upon recieving something from any inlet..)
... ok, I found the prob. Probably the basic newbie thing - the wrong sequence of connection. thank you anyway
Christian Klotz Kunsthochschule Kassel / School of Arts Kassel ::digitalpool
Hallo, Christian Klotz hat gesagt: // Christian Klotz wrote:
... ok, I found the prob. Probably the basic newbie thing - the wrong sequence of connection. thank you anyway
No, please, it's not the sequence or order of connections. This still will be undefined, although it will seem to work for now. Please use triggers everywhere the order of execution matters. triggers are very important and common that's why they can be shortened to "t". They are as important as "f" or "bang".
Frank Barknecht _ ______footils.org__
yeah, it's sad to see this suddenly coming back into the pd vernacular.
where are people learning this? sure you can deduce it experimentally, but if it's written down somewhere, there isn't so much documentation for pd that it wouldn't be hard to track down instances of this being suggested, and delete (or burn) them ;)
it would be great if multiple connections from a point were actually evaluated in separate threads so that execution order was also undefined in practice, and maybe even beneficial (only on an smp machine, probably) to give up in instances where it doesn't matter to the end result. although determining when this is safe could be pretty daunting for a newbie.
but, the most pedagogical approach would probably just be to have pd flag multiple connections from a non-signal output as an error (that mentions the trigger object).
are there other cases i'm not thinking of?
pix.
Quoting Frank Barknecht fbar@footils.org:
Hallo, Christian Klotz hat gesagt: // Christian Klotz wrote:
... ok, I found the prob. Probably the basic newbie thing - the wrong sequence of connection. thank you anyway
No, please, it's not the sequence or order of connections. This still will be undefined, although it will seem to work for now. Please use triggers everywhere the order of execution matters. triggers are very important and common that's why they can be shortened to "t". They are as important as "f" or "bang".
Ciao
Frank Barknecht _ ______footils.org__
PD-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://iem.at/cgi-bin/mailman/listinfo/pd-list
This mail sent through IMP: http://horde.org/imp/
hi
On Aug 18, 2004, at 12:10 AM, Frank Barknecht wrote:
No, please, it's not the sequence or order of connections. This still will be undefined, although it will seem to work for now. Please use triggers everywhere the order of execution matters.
ok, now I got it, I guess ;) - But I think that it would be a good idea to have a warning or better, an automatic insert of a trigger if multiple connections are made to avoid this basic problem, especially for people starting using pd. But for pros I think would it be nice to switch this feature of. cheers
Christian Klotz Kunsthochschule Kassel / School of Arts Kassel ::digitalpool
hello list,
I am guessing this has come around a few times so my appologies. But my laptop screen has died, i use(d) a sony vaio, boo. It was getting old and cranky anyway.
But having respect for this list and its contributors I want to ask about which laptops you use and what you would recomend I buy as a replacement. It needs to make PD roar, that is my main goal. So in my case it has to run linux fc1 ccrma and do all the grunt work GEM requires for the graphics + pdp and have firewire+ tv out or s-video etc. Big wish list I know.
thanks
dominic
On Mon, 2004-08-16 at 21:39, spoonbender wrote:
hello list,
I am guessing this has come around a few times so my appologies. But my laptop screen has died, i use(d) a sony vaio, boo. It was getting old and cranky anyway.
But having respect for this list and its contributors I want to ask about which laptops you use and what you would recomend I buy as a replacement. It needs to make PD roar, that is my main goal. So in my case it has to run linux fc1 ccrma and do all the grunt work GEM requires for the graphics + pdp and have firewire+ tv out or s-video etc. Big wish list I know.
thanks
dominic
PD-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://iem.at/cgi-bin/mailman/listinfo/pd-list