[PD-dev] "modular" compositions of object classes

[mescalinum at gmail.com]

Jakob Leben wrote:
> I guess everybody has missed the point of the second option I gave.
> Indeed it is an extension to the first, but it doesn't have to do
> anything with pd's arrays. The point of it is in binding your own
> externals to pd symbols, thus allowing "yourself" (or themselves) to
> pass data between them (yes, using the same technique as for accessing
> g_array data, but that doesn't matter here). So considering that you
> want to "invisibly patch" something, and that you thought of copying
> the g_array code - I thought this could be a solution. Instead of
> doing insivible patch connections between some objects, you access
> data from one into the other within their own code.

yep works well for externals you are writing, or externals that already
support this mechanism

> Now, if passing data between *your own* externals is not what you
> want, but you really truly want to programmatically control patching
> to other objects, then my proposal is not the solution.

you got it

> Is your final goal to offer to a user who wants to write pd externals
> in tcl the option to create other pd objects and connect them from the
> tcl code of the external?

YES! :)

> Or was that only a part of a solution to another problem that could be
> solved differently?

perhaps.
since it's an externals toolkit, that would be just giving the developer
one more option (hopefully that would not be the main option, but still
an option, to be used when other solutions suck more)

-- 
Federico Ferri

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 261 bytes
Desc: OpenPGP digital signature
URL: <http://lists.puredata.info/pipermail/pd-dev/attachments/20090924/cca23a82/attachment.pgp>