>> Again: license issues, as far as I know (but not from Apple in this >> case). > > Are you sure? It seems to me that it just wants clang from XCode. > I can't imagine what else could be useful for building LilyPond and > provided by XCode.
Well, there are no pre-built 'lilypond' (or 'lilypond-devel') binary packages available on https://packages.macports.org, so yes, it seems that LilyPond must be always compiled. I apparently misremembered, sorry. >> There are some programs, libraries, etc., that MacPorts is not >> allowed to distribute in binary form (see thread >> https://lists.macports.org/pipermail/macports-users/2022-September/051424.html >> for more information). Instead, they must always be compiled on >> the target computer (i.e., the computer you are actually using). I >> currently don't know which one, but at least one of LilyPond's >> prerequisites apparently is affected by that. > > All of LilyPond's dependencies definitely are GPLv3-compatible (or > we could not distribute official binaries on lilypond.org). I don't have access to my old Mac in the next days, but there is a possibility to actually check that in more detail. One of the MacPorts maintainers, Joshua Root, wrote the following some time ago on the 'macports-users' list: > List which ports do and don't have an archive available (for your > current OS version and arch): > <https://gist.github.com/jmroot/f524dcfe5fdadcd8b7a9c2e46151e0d4> > > List which ports are and aren't considered distributable: > <https://gist.github.com/jmroot/f84c329919356bfb1ed2d8425f3cfebb> Maybe somebody who is using MacPorts can run those two TCL scripts and report back, especially the second one? If LilyPond doesn't appear in the non-distributable list it should be possible to ask the MacPorts maintainers to actively switch on binary-package generation. If this can be done, installation of LilyPond should be *much* faster on MacPorts. Werner
