On 10/02/2025 10.36, Gleb Mazovetskiy wrote: > I've subscribed to the mailing list just to post this. This is in > response to https://lists.macports.org/pipermail/macports-dev/2025- > January/046014.html <https://lists.macports.org/pipermail/macports- > dev/2025-January/046014.html>, which I noticed because Tiger support has > been ripped out of macports-base in https://github.com/macports/ > macports-base/commit/fe87793b8e5bb4a05bc432a3c3b17b3fb870b0c7 <https:// > github.com/macports/macports-base/commit/ > fe87793b8e5bb4a05bc432a3c3b17b3fb870b0c7>. > > I think dropping support for Tiger in this way is quite heavy-handed. > I'd like to ask for a compromise, where Tiger stuff is maintained on a > volunteer best-effort per-issue basis rather than entirely ripped out. > > I understand that maintaining Tiger causes some overhead. Only a subset > of packages build on Tiger and that's OK. Trac issues can be closed as > wontfix or ignored until somebody volunteers to fix them. Breaking Tiger > accidentally shouldn't block anyone but please do not break it > deliberately.
What you describe is exactly what has been the state of things for the last ~10 years. We kept things for older systems in place in base as long as they did not break anything. However, it is finally time to remove support for Mac OS X 10.4 Tiger. It has not even seen security updates by Apple in over 15 years. At this point, there is little practical justification to use such an unsupported OS other than for purely nostalgic reasons. And if you want it for that, you will not even need current software on top of it to recreate an authentic experience. If you do not want to throw out the hardware, use one of the *BSDs or a Linux distribution. But even those are slowly removing support for i386, while 32-bit PowerPC has already been dead for a few years in the major distributions. Even larger (development) communities have already moved on from these platforms. Preserving old machines and software is valuable for historical purposes, but maintaining support in a project like MacPorts will not be sustainable. > Moreover, I think porting to Tiger has recently become much easier > because of 2 things: > > 1. gcc14 is now the default for Tiger less than 2 months ago. A modern > compiler makes many previously used hacks unnecessary. https:// > github.com/macports/macports-ports/pull/26655 <https://github.com/ > macports/macports-ports/pull/26655> > 2. There has been incredible progress in legacy-support for Tiger, > merged today. This brings Tiger much closer to modern macOS SDK, and > makes . https://github.com/macports/macports-ports/pull/27619 > <https://github.com/macports/macports-ports/pull/27619> > > These two changes make it much easier to support Tiger. I am well aware that this support of legacy machines is actually the motivation for some to contribute at all. In a volunteer-based open source project, people will have different incentives to come together. That also means that the effort currently spent on legacy systems cannot be redirected to work on other areas of MacPorts. I would have wished that all this effort went into improving modern open source software instead of desperately trying to keep legacy systems on life support. By that I really do not care if that time went into MacPorts or any other open source software. The time contributed by volunteers is a valuable resource for all projects on a global scale. In my opinion, putting it into legacy systems is a complete waste. That said, I do not want to diminish the effort people put into working on what interests them. Open source projects thrive on people that "scratch their own itch" and everyone contributes where they find personal value. Please continue to contribute. After all, regardless of your motivation, your effort is what keeps open source communities alive. Rainer
