-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA256 On Fri, 2018-10-26 at 16:46 -0400, Unit 193 wrote: > > Does this mean we'll lose any gsettings parameter stored in dconf? I don't > > think it's acceptable. > > As mentioned above, yes. I suppose that could be another reason to ship it > in > another package, but I'd be interested in making it easier to enable. I > presume > you'd be against injecting the setting into the user's environment if > xfconf-gsettings-backend is installed?
It depends on what are upstream plans. I don't think the env var is really a long term solution, but it's nice for testing. I think we should use a different package (so go ahead and commit it), but don't touch the env var (an user can still play with it if needed). When upstream considers it stable, then I assume env var should be dropped and the backend enabled by default (when compiled in and the package is installed). But then before actually enabling it in unstable, we would need a migration script, which should be run the first time xfconf is run with the new option, in order to get all gsettings parameters from dconf and copy them to xfconf backend. Upstream should provide that migration script, ideally (like was done for migration from MCS to xfconf). Regards, - -- Yves-Alexis -----BEGIN PGP SIGNATURE----- iQEzBAEBCAAdFiEE8vi34Qgfo83x35gF3rYcyPpXRFsFAlvTgZAACgkQ3rYcyPpX RFskEQf9HRyoGbPlEhkyayI/LFMcX6aagCAaLYT0PRFjqPfmOIdQTGjRMHTgnZCe jcXluTAyz7UoCvHT2PXU7PqDKoDkpPbhGlJ6Nxti/g3DpSFJgG62qFv2h9fvjifx 40trl1ZDlUDoTkLet8yNmlC+a0Op0wZB1rN4SxFdIM1KrsVFfaP0t50K7KgVsuec mYKL4Xh19akZ967Z7EWEMOxKyHvY+B0Vr8YDZ5mPYCNNbIg7i26RrIogQMkcmzVm nieaScZi5T4vkDbU1OXlfxQlvcPgYiUnSPzbhgQaOXI4bJKJcDQzZUF6NeG63ukC C8Mr/+qB36ZBNp7clcAKanvPxyFHww== =Sii7 -----END PGP SIGNATURE-----