add tests for invalid input

inet_net_test: add a test for an overlong prefix

i'm okay with this on the basis that it turns the pkg problem into a release thing and not something that affects all developers. we definitely need a better way to do this, but i think everyone is aware of that at this point :-)

i'm still not super fond of this since it's just vendoring with extra steps, but it does fix my major complaints with the original change.

i had to rebase this to fit with Colin's change in b9afd7f6a5b1cf1a6cbcef7a55aede7e71a0babc: https://git.le-fay.org/freebsd/src/commit/?h=main&id=9cdd5f0088308e68f71323737c6b3044c9b485bc

per IRC discussion, also move kyua to its own package, and add a depdendency from -tests to -set-base

In D52562#1201142, @kevans wrote:

Yeah, I'm happy with that, thanks. I recognize this is maybe not ideal for more flexible system construction (piece together your own based on a set and removing some), but IMO we should err on the side of caution and consider whether we can do something to enable better behavior with the final form of pkg groups.

it turns out installing set-devel always installs tests via a transitive dependency set-devel -> tests-dev -> tests, D52597 fixes this.

as i mentioned on IRC, i would prefer to land this myself along with D52558 (bsdinstall changes) so i can MFC both together, since they both require each other.

if possible, please MFC this to stable/15 immediately after landing it since it's required for D52558 and i'd like to get all the pending MFC changes into stable/15 in time for 15.0-ALPHA3 on Friday.

In D52558#1201065, @ifreund_freebsdfoundation.org wrote:

I haven't had a chance to test this out myself locally yet, will try to get to that tomorrow.

only use a blacklist for the package selection dialogue

address minor review feedback

this is a proof-of-concept for jail sets, so there's only one for now; we can add more later.

use path_glob in the trigger rather than path

based on feedback and discussion, mark all sets as vital

reword the comment about installing the minimal set to be more clear

In D52558#1200643, @jrtc27 wrote:

Is the expectation that we ensure all release media have all supported package sets present in an offline package repo? Because disc1/memstick won't have the real src package, for example, and bootonly/mini-memstick won't have any real packages at all. So I think for this to work even bootonly/mini-memstick needs to have a package repo of the package sets (as a pkgbase analogue to how it still has the MANIFEST file that lists all distribution sets, just the sets themselves aren't present)?

enable all libcompats by default

minor fixes

address review feedback

In D52558#1200313, @ifreund_freebsdfoundation.org wrote:

I'm not convinced that making both base and minimal selectable with an error when neither is selected is clearer.

i'm not sure what's going on here, but it seems like my local system isn't building the src(-sys) package at all, which may be why i didn't notice this. in any case i like having sets for both tests and src so we can use them in the installer.

In D52562#1200173, @kevans wrote:

I think my complaint is that, iirc, all of the metapackage dependencies are marked as automatic. As soon as they break the set as you suggest, the remainder would be subject to autoremove

note: to test this you need a pkg(8) with the patch at https://github.com/freebsd/pkg/pull/2516. however if the user doesn't have that patch, everything will still work fine, they just won't get the mandoc database built, which is the same as the current behaviour.

In D52562#1200123, @jrtc27 wrote:

I think they should all be vital, in the sense that if you've installed that metapackage then you don't want your system to ever remove it without very explicit confirmation, no?

i've been thinking about this and i'm inclined to continue with this change as-is, mainly since this means we still list all available rc.conf options in /etc/defaults and the user doesn't have to look at /etc/rc.conf. if no one objects, i will land this in main and stable/15 along with D51854, then open another review to convert the remaining rc.conf entries to rc.conf.d.

fine by me, except please ensure we refer to Rocky Linux(tm) by its official name.

In D52558#1200053, @jrtc27 wrote:

Please make base the default. bsdinstall defaults should be such that you can mash enter without thinking and get a sensible system. I suspect minimal is not what most want.