[PD] propertybang-help

[Jonathan Wilkes]

--- On Thu, 4/1/10, IOhannes m zmoelnig <zmoelnig at iem.at> wrote:

> From: IOhannes m zmoelnig <zmoelnig at iem.at>
> Subject: Re: [PD] propertybang-help
> To: "Jonathan Wilkes" <jancsika at yahoo.com>
> Cc: "pd-list" <PD-list at iem.at>
> Date: Thursday, April 1, 2010, 7:31 PM
> On 2010-04-01 18:11, Jonathan Wilkes
> wrote:
> > 
> > 
> > 
> > That's currently the problem with your help
> patch.  The behavior of the 
> > object is clear from the usage description, but then
> you decide to put 
> > "per abstraction" in all caps, which made me wonder
> whether you mean 
> > "per abstraction instance"-- the desired behavior-- or
> "per abstraction class"-- in which case I would take it to
> mean right-clicking one instance 
> > sends a bang to _all_ instances.  Then you said
> at the bottom that "you 
> 
> it does the "desired" behaviour.
> 
> > cannot have [propertybang]s per subpatch," which
> confirms the latter (see 
> > "2.7. subpatches"),
> 
> "subpatch" in the help patch means what is commonly called
> "subpatch"
> (aka [pd]) which is a "one-off subpatch" in the docs.
> 
> >  which would render the object useless and make
> me 
> > think it's just not finished yet.  So I have to
> build my own abstraction 
> > and test the object to see whether the it does
> something useful, which 
> > defeats the whole purpose of having the help patch in
> the first place.
> 
> i cannot follow.
> i don't like help-patches that are not self-contained (at
> least for what
> they are trying to document); in older versions
> [propertybang] obviously
> did not work for "one-off subpatches" (hence the
> documentation about
> this), which made it a bit hard to show without
> abstractions.
> otoh, creating an abstraction yourself and test whether the
> object
> indeed does what _you_ want it do, is not so complicated.

The vast majority of help patches are self-contained (excluding for the 
moment those that have objects from other libraries which may or may 
not exist).  For those few patches that warrant an abstraction in the 
example, it's easier (on the reader of the subpatch) to include an 
example abstraction than to describe what the behavior would be were an 
abstraction included (which may or may not be up to date).  Right-clicking 
"Properties" on one [foo] object and noticing that a bang does not come 
out a different [foo] object takes a few seconds to comprehend, unlike 
your "PER ABSTRACTION" paragraph which takes a message to a mailing list 
and revisions to correct/clarify the meaning.

Also, please note that I'm talking about about an example of the core behavior of [propertybang], not an example of how it might be useful in a 
real (albeit simplified) patch.  Only the latter belong in the 
"examples" folder. I.e., "5.reference" = what it does, "examples" = how 
it might be used.

-Jonathan