Re: RFC: New release and abandoning MacOS?

Greg Troxel <[email protected]> Tue, 27 Jun 2023 14:45:04 -0400
Newsgroups gmane.linux.printing.gimp-print.devel
Message-ID <[email protected]>
Solomon Peachy via Gimp-print-devel
<[email protected]> writes:

> I'd _really_ like to see a proper, full release go out; I'm still 
> getting bug/support issues about 5.3.3, which is the last release we 
> actually announced... nearly four years ago.  While we did cut a 5.3.4 
> tarball, we never announced it to the world, holding back due to the
> lack of a sane/working MacOS installer.
>
> In the subequent 2.5 years, things are even worse today on the MacOS 
> front, as Apple has continuted to iterate on their platform vision in a
> way we've just not been able to keep up with.
>
> Meanwhile, there have been more printer additions and even more 
> bugfixes, and I think it's past the time we cut a new release so maybe 
> our first answer to nearly every support request won't be "uninstall 
> whatever you got from your distro and compile Gutenprint maually"
>
> Additionally, since we can't meaningfully support MacOS at all any more, 
> IMO it makes more sense to just walk away from it entirely.  Once we get 
> a "Gutenprint Printer Application" working for the brave new 
> IPP-Everywhere paradigm, we can revisit releasing that for MacOS if it 
> still makes sense to do so.

Speaking as a packager (pkgsrc) and user (on NetBSD), this seems
reasonable.

I am unclear on the mac issues; I wonder if a command-line build is
workable, skipping the entire GUI world, so that
e.g. pkgsrc/homebrew/fink could build it on macOS.  But if even that is
too much, just marking it unsupported (while not actively withdrawing)
makes sense.

Speaking as a Free Software community member, I think it's 100% ok to
simply decline to support a non-Free system that is becoming more
difficult.  mac users are welcome to ask Apple to contribute fixes, to
do the work themselves, or to hire someone to do it.


Separately, I remember some issues with 5.3.4 on NetBSD that I didn't
manage to figure out, so my suggestion would be to tag an alpha (with a
release tarball) as soon as reasonably practical to ask packagers and
users to start testing.

Greg