Mention add_dist_script in meson dist docs

[omoerbeek]

See PowerDNS/pdns#15048

[eli-schwartz]

――――――――――――――――――――――――――――――――――――― ✀  ―――――――――――――――――――――――――――――――――――――
"GNOME Builder" https://gitlab.gnome.org/GNOME/gnome-builder 500
"GNOME Web" https://gitlab.gnome.org/GNOME/epiphany 500
"Nautilus" https://gitlab.gnome.org/GNOME/nautilus 500
"Pango" https://git.gnome.org/browse/pango/ 500
"Sysprof" https://git.gnome.org/browse/sysprof 500
――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――


Summary of Failures:

2/2 validate_links FAIL            4.84s   exit status 1

lol... unsure how this could be reliably detected/avoided.

[eli-schwartz]

Maybe in combination with the above suggestion, remove "this"?

[eli-schwartz]

Or actually maybe remove the line entirely.

Modification to the checked out files like generating files or
setting version info can be done with dist scripts:
```meson
[[#meson.add_dist_script]]('distscript.sh')
```

[jsoref]

Link checkers should not treat 5xx errors as failures. Only 4xx responses are user failures worth failing over.

[eli-schwartz]

On the one hand, yes. On the other hand, let's say an upstream site goes "out of business", their backend database gets unplugged one day, and the whole domain simply returns http 500 for the next 5 years because it's still being served by nginx on a box that is also doing something totally different and not website related.

I've definitely seen a few sites like that. It's arguably a broken link, albeit much less common.

[jsoref]

Oh, it's fine to warn for that, I'm just opposed to error-ing for it.

[jpakkane]

Please keep the commits squashed down to one. Extraneous commits like "better phrasing" are just noise in the commit history.

[eli-schwartz]

^^ what he said

In theory I could squash this locally, update your PR (you have left "maintainers are allowed to edit this pull request" enabled) and then merge the result, but it would have to wait until I get back to my laptop...