On Fri, Nov 07, 2003 at 03:09:37 -0800, Tim Hockin wrote: > > ---> Some central DB, which could be nothing more than an "index" file at > > the root of a directory. Simply copying into a directory presents > > versioning problems, and having to scan at startup if a huge inconvenience > > to users. I dont see how an index file is any different to a directory. You still have the issue that you have to remove files refering to plugins that no longer exist and add entries for plugins that are added, but a directory removes synchronisation problems. > I think the requirements should include something like: > * there is no system-global registration needed for plugins > * plugin installation can be as simple as dropping a file into a dir > * hosts should examine the $GMPI_PLUGINS_PATH environment variable and > recursively scan each entry found That is more or less how LADSPA+RDF works. I dont really know wether its succesful or not. > As I see it, sample-rate is set at plugin-instance creation and never again. > To change sample-rate, destroy and re-create the plugin. I don't know if > anything else needs this level of treatment. Sample-rate is a big deal. If > we make plugins support dynamic sample rates, the code gets harder. Agreed, personally I dont want ot have to deal with that. It can be hidden by the host requesting a state dump, reinstntiating the plugin and making it restore from the last state. > My first thought was that each dll (or .so) file could indicate 1 or more > plugins that existed in that file. The full list of plugins that the host > could access was the set of all plugins in all dll files. ... > When probed, vst_wrapper examines the VSTPlugins path and returns > all the plugins that it can find. Yes, this is what the LADSPA/VST and LADSPA/jMax wrapper plugins do. It works well enough. > > ---> Yes, we do. I think we should do this by a type definition, that's > > defined globally for the plugin "profile." In 99% of the plugins, we'll > > have "typedef SAMPLE_TYPE float". For non float systems, it can be "typedef > > SAMPLE_TYPE int32" (or whatever). Then its up to the plugin code to manage > > the math. I would love to be able to support ProTools TDM, which means we > > need this. > > So do we just provide the typedef and no operations? Does that make float > parameters become int32, too? What about double parameters? We would need to define low- and high-precision reals if we need float and double parameters. - Steve ---------------------------------------------------------------------- Generalized Music Plugin Interface (GMPI) public discussion list Participation in this list is contingent upon your abiding by the following rules: Please stay on topic. You are responsible for your own words. Please respect your fellow subscribers. Please do not redistribute anyone else's words without their permission. Archive: //www.freelists.org/archives/gmpi Email gmpi-request@xxxxxxxxxxxxx w/ subject "unsubscribe" to unsubscribe