Re: 2026Q1 x11/xephem segfaults amd64

[Greg Troxel <gdt%lexort.com@localhost>]

I have applied Martin's upstream PR.

There are a very large amount of issues, some minor, some perhaps
leading to bad behavior.  I don't want to maintain even a temporary soft
fork of xephem with local fixes for all of them, when upstream appears
to be functioning.   It feels like a large amount of effort to make
something not used very much functional a little sooner, when the
problems are not pkgsrc accomodation.

So it would be great if people who care would try to build upstream
master from git, see the warnings, and submit upstream PRs to fix, them,
and then those fixes will end up in pkgsrc after a release.

I ran a build after the #126 patch, and there remain a significant
number of issues: ctype UB, printf specifiers, sprintf size.

I might consider an xephem-snapshot in wip, to track master, so people
can get fixes sooner.  We will have to see how the two PRs filed
upstream go.

As maintainer, I always ask that no patches (that belong upstream) be
applied unless an upstream PR is filed and referenced.  Unfiled packages
needlessly make future updates harder.  (This is a documented norm
already, not a special gdt rule.)