ティージェーグレェ on Nostr: Oh, MacPorts has a pretty convoluted set of compilers available (some versions of OS ...
Oh, MacPorts has a pretty convoluted set of compilers available (some versions of OS X are so old, they're before Apple moved from gcc to llvm/clang), and some versions have bugs. So the Pull Request there I think was basically trying to exclude known buggy compilers which seems to have done the trick. For example, this earlier Pull Request also apparently fixed snac for MacPorts users who were using version 4.2 of gcc: https://github.com/macports/macports-ports/pull/23999
To be honest, some of that stuff is a bit over my head, there was a rather involved discussion on the macports-dev mailing list recently to try to determine which versions of gcc should be abandoned, which should be maintained, and which ones might require other versions for boot strapping. Clearly, there are some MacPorts devs who really care about supporting older versions of OS X, and I think that is great! But I only have so many spoons personally. Regardless, I still prefer the MacPorts approach to say, Homebrew (which only supports the current Apple macOS release and two previous OS releases).
Moreover, Homebrew also has "analytics" on by default and I don't like that either as it is invasive and anti-privacy in nature. There are other package managers which are similarly bad, e.g. Microsoft's vcpkg. MacPorts does allow for analytics, if users want they can opt-in by installing mpstats (based on https://ports.macports.org/port/mpstats/ it appears as if roughly 1100 people have it installed?) to provide some statistics to the MacPorts project, but it is a sub-sampling, at best. Currently, MacPorts has only 1 snac user listed, which means that only one of those 1100 or so MacPorts users who has installed mpstats has also installed snac? I am not that user, and given that I have also not been the only one submitting Pull Requests related to snac in MacPorts, I am guessing that there might be more snac users using MacPorts than merely 1? Just a guess though! I am OK not knowing and leaving MacPorts' users their personal privacy on by default.
To be honest, some of that stuff is a bit over my head, there was a rather involved discussion on the macports-dev mailing list recently to try to determine which versions of gcc should be abandoned, which should be maintained, and which ones might require other versions for boot strapping. Clearly, there are some MacPorts devs who really care about supporting older versions of OS X, and I think that is great! But I only have so many spoons personally. Regardless, I still prefer the MacPorts approach to say, Homebrew (which only supports the current Apple macOS release and two previous OS releases).
Moreover, Homebrew also has "analytics" on by default and I don't like that either as it is invasive and anti-privacy in nature. There are other package managers which are similarly bad, e.g. Microsoft's vcpkg. MacPorts does allow for analytics, if users want they can opt-in by installing mpstats (based on https://ports.macports.org/port/mpstats/ it appears as if roughly 1100 people have it installed?) to provide some statistics to the MacPorts project, but it is a sub-sampling, at best. Currently, MacPorts has only 1 snac user listed, which means that only one of those 1100 or so MacPorts users who has installed mpstats has also installed snac? I am not that user, and given that I have also not been the only one submitting Pull Requests related to snac in MacPorts, I am guessing that there might be more snac users using MacPorts than merely 1? Just a guess though! I am OK not knowing and leaving MacPorts' users their personal privacy on by default.