[PD] Pd-L2ork Features

[Jonathan Wilkes]

>________________________________
> From: András Murányi <muranyia at gmail.com>
>To: pd-list <pd-list at iem.at> 
>Sent: Friday, December 14, 2012 2:38 PM
>Subject: Re: [PD] Pd-L2ork Features
> 
>
>On Fri, Dec 14, 2012 at 6:15 PM, Jonathan Wilkes <jancsika at yahoo.com> wrote:
>
>>________________________________
>>> From: Hans-Christoph Steiner <hans at at.or.at>
>>>To: Jonathan Wilkes <jancsika at yahoo.com>
>>>Cc: Ivica Ico Bukvic <ico at vt.edu>; 'pd-list' <pd-list at iem.at>
>>>Sent: Friday, December 14, 2012 11:53 AM
>>
>>>Subject: Re: [PD] Pd-L2ork Features
>>>
>>>
>>
>>>That's nice for people who are used to doing that from Max/MSP.  I've never
>>>found a reason to want that feature, so I haven't implemented it.  Can someone
>>>explain how they use it?
>>
>>
>>You partly answered your own question.  Another reason is if you are mucking
>>around in the source you can just copy-paste it into a new patch instead of
>>closing, opening the file selector and choosing that file again.
>>
>>Another reason is you can select the code off the browser and immediately
>>play with a patch off svn instead of saving it to a location and opening it.
>>
>>Another reason is that you can throw text up on a pastebin for people
>>who you don't necessarily have an email address for (e.g., on IRC), and they can paste it
>>into a new patch instead of copy-pasting it into a text editor, choosing save,
>>choosing a save location, choosing "Open" in Pd, navigating to that same
>>location where they saved it, selecting it, and eventually deleting the patch
>>when their download directory is full of junk.
>>
>>
>Another reason is that it makes super easy to exchange code snippets in email and on forums.


BTW-- for simple patch snippets, what about making an ascii-art parser to paste
the art as actual object chain in a patch?



---Begin ASCII Snippet---
[880( #comment
|
[osc~ 440] #Some other comment
|
[*~ 0.1]
|
[dac~]

---End ASCII Snippet---

Above, it's possible to parse in tcl-- everything that's an object should be
enclosed in [], [(, and maybe [\ for atombox.  Maybe number-sign to
set off comments.

---Begin ASCII Snippet---

[f]x[+ 1]
    |
    [send wherever]

---End ASCII Snippet---

"x" = 1) outlet0 of object to the left --> inlet0 of object to the right,
2) outlet0 of object to the right --> inlet 1 of object to the left.  Nice
quick-n-dirty shortcut for a counter.


Since Pd ascii art is monospaced we can see that the xpos of
the wire = xpos of [+ 1] above and xpos of [send wherever] below,
so connect outlet0 to inlet0.

---Begin ASCII Snippet---


[route 0 1 2 3 4]
|   |   |   |   |
[5( [6( [7( [8( [9(


---End ASCII Snippet---

For each wire from/to an object, increment the xlet number by one.  This
gets you ancillary wire connections for contiguous connections (i.e., you
can't skip an outlet).

If we close objects off by looking for [, (, or \, followed by a space or linebreak, this lets
us also use those characters inside objects like [expr (32+12)*4] without ambiguity.

Writing the above ascii is _way_ easier than doing the same in Pd-gui and connecting
all the wires.  In fact I'd rather click <ctrl-t> and enter that into a text widget than do that.

General solutions for crossed-wires, slanted wires, etc. are harder but I
think this would already be quite useful.


-Jonathan


>
>András
>
>_______________________________________________
>Pd-list at iem.at mailing list
>UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
>
>
>