Re: [PATCH] edo.eclass: support EAPI 9

>>>>> On Wed, 19 Aug 2026, Florian Schmaus wrote:

> On 19/08/2026 00.14, Sam James wrote:
>> This isn't right. First, edo is already provded in EAPI 9, so its
>> definition would have to be guarded.

> Oh, right. Now I remember that we already tried to bump edo to EAPI 9.

> That said, was it considered to simply not guard the edo definition?
> Since the eclass would simply override a EAPI 9 function with
> identical behavior, nothing should break and we shouldn't get the
> spurious pkgcheck warning.

No, it wasn't considered. Eclasses shouldn't override package manager
commands. Especially not when eclass and Portage have different
implementations (as is the case for edo).

> Factoring out edob in a new eclass, as you suggest below, is, of
> course, the cleaner solution.

Yes. 

>> I think it makes sense for edob to be split into its own eclass
>> instead.

> Sounds sensible. Shall I do it, or do you want to do it as the eclass'
> author and member of the maintaining project (qa)?

I think this question is for sam to answer.

-----BEGIN PGP SIGNATURE-----

iQFfBAEBCABJFiEEtDnZ1O9xIP68rzDbUYgzUIhBXi4FAmqFhYEbFIAAAAAABAAO
bWFudTIsMi41KzEuMTIsMiwyDxx1bG1AZ2VudG9vLm9yZwAKCRBRiDNQiEFeLtj0
CACgtmBXJcuDsBX3VNfNtRqi/nPYNaMNU3lpxIn171pB6PunzBrCBuN913iMVquW
iDqOVe1fv0sbiM9iMtaDeuNHsZ9PP2pxEkA7u/+MQSSbG0rNJHBABAdnzuY0QcHl
53fAB6/nwEA2VLtdSnX55uyLyepmij/6ba2Gn83O4Rk6cVRn+taz5Bpv5VSxIlwS
XxakLaLm4tHAvKXZfCssd+qrltDJQpp3g8DzBZNhjF/gXg0gNfYKq5WzwSIrv0FL
5ofbaRERXTWJZVgAwZRiUZkDWWT5oDMvRcQIlYpDxvP3m98Q1+1+tLaOjbjVYspr
g2wC/crp2tN7Mv3bjIngXJPY
=KVyF
-----END PGP SIGNATURE-----