On 02/18/2017 05:38 AM, Ivica Ico Bukvic wrote:
If you had a print inside the abstraction it would print. If you dynamically create objects, chances are you will first dynamically create the abstraction which can initialize itself using a loadbang, then the print, then the connection. In this respect its behavior is synonymous to, for example [f 42] which initializes itself to 42 but does not output anything until asked to do so. This way initialization of individual abstractions is handled gracefully, whereas the example you are suggesting would require manual interaction regardless whether you are using an abstraction foo or an [f 42]. In other words, pd-l2ork/purr-data's loadbang implementation allows for parity between built-in objects and abstractions.
the thing is, that pd-vanilla's [loadbang] has slightly different semantics: it allows you to implement [init 666] as an abstraction (rather than just only [f 42]).
you are of course free to change the semantics of any built in object. however, it comes at the cost of deliberatly breaking compatibility.
gfmards IOhannes