At least in Pd-l2ork, clicking "New" in the "File" menu sends a series of messages to Pd to set the filename/dir,
create a new canvas, and "pop" the new canvas.  So I'm guessing Chris has figured out he can send these
messages in Pd-extended 0.42 to manually create a new patch with whatever name/directory he wants.

Same thing happens in Pd-Vanilla, except it's a single "menunew" message sent to Pd and the canvas gets named,
created and pop'd as a result.

Pd-l2ork has [preset_node] and [preset_hub], but Pd-l2ork currently only runs on GNU/Linux.

-Jonathan




On Thursday, September 24, 2015 10:24 AM, IOhannes m zmoelnig <zmoelnig@iem.at> wrote:


On 2015-09-24 16:05, Chris wrote:
> Thanks Jonathan, menunew did the trick!
>
> IOhannes: it creates and opens a new patch -- first argument is the
> filename, second argument is the path.

according to the source code it does not.
it simply sets the filename and path for the next to-be-created abstraction.


> It turns out this is very handy for
> dynamically creating abstractions, e.g. for state-saving/presets :)

hmm, why don't you just use a text-file (or multiples thereof) for presets?
dynamic patching can generally be avoided.

e.g. marian's [kollabs] is a very powerful Pd-vanilla state-saving
system, that doesn't rely on any undocumented features.


fgmsdr

IOhannes






[kollabs] https://github.com/m---w/kollabs


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