Here's my proposal: horizontal connections should only be used when the upper object has only one outlet, and the lower object has only one inlet. Otherwise there will be always be ambiguity.
With your example of all outlets pointing to the same inlet, there is still ambiguity as to whether or not all the outlets are actually connected. That's fine if your patching style is to connect them first, and only line them up after. But if someone is debugs a patch they made that includes your abstraction, they would have to click and move objects to be certain all the objects are actually connected.
There is one way to be a little less conservative, but it requires a usage agreement: if a horizontal connection travels from left to right, it must come from the right-most outlet of the upper object and must connect to the left-most inlet of the lower object. Reverse if the connection travels from right to left. The problem is there's no good way to tell if a patch adheres to the agreement.
Matt
Matt, I like the usage agreement idea, as long as there's visual feedback. For example, if the inlets and outlets in pd-ext protruded one or two pixels, you'd be able to see the connection crossing over to the "wrong" outlet.
-Jonathan
--- On Sun, 3/22/09, Matt Barber brbrofsvl@gmail.com wrote:
From: Matt Barber brbrofsvl@gmail.com Subject: [PD] Horizontal Connections (Was: Re: style guide idea: [send To: pd-list@iem.at Date: Sunday, March 22, 2009, 5:52 AM
Here's my proposal: horizontal connections should
only be used when the upper object has only one outlet, and the lower object has only one inlet. Otherwise there will be always be ambiguity.
With your example of all outlets pointing to the same
inlet, there is still ambiguity as to whether or not all the outlets are actually connected. That's fine if your patching style is to connect them first, and only line them up after. But if someone is debugs a patch they made that includes your abstraction, they would have to click and move objects to be certain all the objects are actually connected.
There is one way to be a little less conservative, but it requires a usage agreement: if a horizontal connection travels from left to right, it must come from the right-most outlet of the upper object and must connect to the left-most inlet of the lower object. Reverse if the connection travels from right to left. The problem is there's no good way to tell if a patch adheres to the agreement.
Matt
Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
Jonathan Wilkes wrote:
Matt, I like the usage agreement idea, as long as there's visual feedback. For example, if the inlets and outlets in pd-ext protruded one or two pixels, you'd be able to see the connection crossing over to the "wrong" outlet.
Maybe better if outlets with connections were coloured differently.
Martin
Martin, That's a great idea. You would get visual feedback for all relevant cases. And I now realize that the protruding inlets/outlets don't resolve the ambiguity of the case you mentioned, whereas colored connections do.
-Jonathan
--- On Sun, 3/22/09, Martin Peach martin.peach@sympatico.ca wrote:
From: Martin Peach martin.peach@sympatico.ca Subject: Re: [PD] Horizontal Connections (Was: Re: style guide idea: [send To: "Jonathan Wilkes" jancsika@yahoo.com Cc: pd-list@iem.at, "Matt Barber" brbrofsvl@gmail.com Date: Sunday, March 22, 2009, 4:35 PM Jonathan Wilkes wrote:
Matt, I like the usage agreement idea, as long as
there's visual feedback. For example, if the inlets and outlets in pd-ext protruded one or two pixels, you'd be able to see the connection crossing over to the "wrong" outlet.
Maybe better if outlets with connections were coloured differently.
Martin
On Sun, 22 Mar 2009, Martin Peach wrote:
I like the usage agreement idea, as long as there's visual feedback.For example, if the inlets and outlets in pd-ext protruded one or two pixels, you'd be able to see the connection crossing over to the "wrong" outlet.
Maybe better if outlets with connections were coloured differently.
Why not just colour the wire itself!
Even if you connect the other outlets in various horizontal ways, you'll still be able to see the one wire crossing over a bunch of outlets... if it's only one wire.
If you need to see several wires... they could be coloured using random colours.
_ _ __ ___ _____ ________ _____________ _____________________ ... | Mathieu Bouchard - tél:+1.514.383.3801, Montréal, Québec
Mathieu Bouchard wrote:
On Sun, 22 Mar 2009, Martin Peach wrote:
I like the usage agreement idea, as long as there's visualfeedback. For example, if the inlets and outlets in pd-ext protruded one or two pixels, you'd be able to see the connection crossing over to the "wrong" outlet.
Maybe better if outlets with connections were coloured differently.
Why not just colour the wire itself!
Even if you connect the other outlets in various horizontal ways, you'll still be able to see the one wire crossing over a bunch of outlets... if it's only one wire.
If you need to see several wires... they could be coloured using random colours.
Exactly. I already proposed a patch to do that a couple of years ago but it didn't get picked up.
Martin
On Sun, 22 Mar 2009, Matt Barber wrote:
must connect to the left-most inlet of the lower object. Reverse if the connection travels from right to left. The problem is there's no good way to tell if a patch adheres to the agreement.
There's never a good way to do that. I remember losing an hour or so trying to debug a patch to finally realise I had two identical objects exactly on top of each other and that my patch was using the first's inlets and the second's outlet. No wonder no message got sent through.
_ _ __ ___ _____ ________ _____________ _____________________ ... | Mathieu Bouchard - tél:+1.514.383.3801, Montréal, Québec
-
Jonathan Wilkes -
Martin Peach -
Mathieu Bouchard -
Matt Barber