>> * provide a toolkit that is implemented on top of >> existing (popular) toolkits (i.e. host X using toolkit Y would link >> against the Y version of the GMPI toolkit). > >Why not as above, but "Host X using toolkit Y will load the version of the >plugin linked for Y"? sure, that's another alternative. but it would lead to paralysis on *nix systems. of the major sequencer-style apps that exist right now, ardour is written with gtk, muse and rosegarden are written with qt, and some others exist that use other toolkits. the likelihood of GMPI plugins that rely on featureful GUIs being implemented for more than 1 X Window toolkit seems to be close to zero (granted the chance of it being implemented for even 1 X toolkit is rather small too :). >> * wait for toolkits to fix this godawful mess. > >How soon is that likely to be? i don't know. the problem is that this is not a problem most programs face. "plugin" in the *nix world generally refers to things that are just code, no GUI. there are lots of examples of this. i can't think of anywhere outside of the audio/midi realm where a plugin is not only a chunk of code to be executed, but also a GUI. as a result, the toolkit authors see us as a bizarre niche and don't have much interest in fixing things. the sad part of it? it isn't that hard to fix. there are several approaches, such as providing direct access to an entry point in the toolkit that just accepts X events that are associated with windows controlled by the toolkit. another is to remove the insufferably lame assumption on the part of the toolkits that they own the one and only connection to the X server. >> conceptually yes. technically and workwise it would be a real pain. i >> suspect that most plugins have about a 60:40 to 80:20 split between >> GUI code and DSP code. the GUI code has to reimplemented for each >> framework. and if the developer doesn't feel like doing it, it doesn't >> get done, and their plugins continue to be platform-specific. > >That's the developer's own problem at the end of the day. Same as if he >or she can't be bothered to port it to the Mac. Obviously we should try to >avoid creating unnecessary platform divisions, but when the platforms >really are significantly different, what can you do? vstgui is already an example of what can be done. win and mac GUI APIs are not at all alike (syntactically anyway), and so its an example of someone deciding to bridge the gap, reduce developer workload and enhance end-user satisfaction. i suspect we can do the same, but i'm not sure yet of the best way to do this. ---------------------------------------------------------------------- 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