2011/3/30 David Robillard <[email protected]>: > On 29/03/11 02:59 PM, Tim Goetze wrote: >> >> [Philipp �berbacher] >> >> >>> >>> Excerpts from Stefano D'Angelo's message of 2011-03-28 22:59:46 +0200: >>> >>>> >>>> This means, if you change the port signature and maintain the same >>>> UniqueID, we would have incompatibilities in the LV2 world. If you >>>> create a new plugin or don't touch ports, instead, everything's fine. >>>> >>>> Stefano >>>> >>> >>> I'd say you'd even have incompatibilities in LADSPA world. Even fixes in >>> LADSPA plugins would sometimes need a new ID (This was discussed a while >>> ago regarding a LADSPA that has an unintuitive port order). >>> >> >> Lacking sufficient knowledge of all the LADSPA hosts out there, I'm >> unable to judge how many will cope with the addition of a port to an >> existing plugin and how many will not. >> > > /Adding/ a port is probably fine, since hosts can just use the default value > or connect it to silence. > > However, "adding" here really means "appending": the new ports must be added > on to the end of the ports (by index). Definitely do NOT change existing > indices for ports, that will definitely break a lot of things in horrible > ways! It's not even possible to properly cope with that situation.
Mmm.. so I could maybe add the port number to the URI and add dc:replaces, like this: <urn:ladspa:1234:5> a lv2:Plugin ; dc:replaces <urn:ladspa:1234> . Yet, I should send patches to those plugin authors providing both LV2 and LADSPA versions of their plugins (the same holds true for DSSI, I guess). At the moment I don't seem to be able to come up with a better solution. :-S Stefano _______________________________________________ Linux-audio-dev mailing list [email protected] http://lists.linuxaudio.org/listinfo/linux-audio-dev
