Steve Harris wrote: >On Fri, Dec 07, 2001 at 12:00:02 -0000, Richard W.E. Furse wrote: >> GUIS AS PART OF THE PLUGIN > >As Paul pointed out it isn't actually neccesary to change LADSPA to >implement this. Earlier we we thinking of letting hte plugin specify a UI, >but that isn't really neccesary (or desirable).
agree, that would take the 'S' out of it. >> CHANGING CONTROL PARAMETERS DURING PROCESSING > >This has never bitten me because my plugins (nearly) always do: > > LADSPA_Data gain = *(plugin_data->gain); > >But this is an important issue and needs to be sorted out. i think that how to deal with this should be left to the host; possibly with a LADSPA_HINT_UNBUFFERED_CONTROL. >Yep, I agree with this. I was forgetting the S! However I would like to >see the defaults (they should be back compatible) and the propsed GUI >doesn't add to the complexity of LADSPA. yeah, let's keep it real simple and export: const char * xml_spec; and we can translate all discussions about ontologies, guis etc into a less harmful discussion about the semantics of this xml_spec, which i'd imagine to be roughly like: <typeOf> waveshaper frequency-effect </typeOf> and future enhancements may come at a lesser price than a revised <ladspa.h>. if a host does not understand a tag, he may just skip it. 't would also allow for the 'quick action wanted' in the subject line :) tim
