net-p2p/bitcoind should provide systemd unit file. Reproducible: Always
Created attachment 348144 [details, diff] bitcoind-0.8.1.ebuild.patch
Created attachment 348146 [details] bitcoind.service
(In reply to comment #2) > Created attachment 348146 [details] > bitcoind.service Without "-daemon=0" option users could add "daemon=1" in bitcoind.conf and this will break the unit (daemon=1 require Type=forking).
Surely there should be a USE flag controlling this? I don't want systemd clutter on my system, at least, and I doubt many others do.
(In reply to comment #4) > Surely there should be a USE flag controlling this? I don't want systemd > clutter on my system, at least, and I doubt many others do. Yeah, that's the debate on gentoo-dev@. I'm thinking unconditionally install and limit with INSTALL_MASK. Leave this open until the dust settles.
(In reply to comment #4) > I don't want systemd clutter on my system, at least, and I doubt many others do. You likely already have several systemd units installed on your system. ;)
I will commit in a week if nobody disagrees (also CCing systemd team to let them review the unit files)
(In reply to Pacho Ramos from comment #7) > I will commit in a week if nobody disagrees (also CCing systemd team to let > them review the unit files) I still disagree, and see no reason to change that...
(In reply to Luke-Jr from comment #8) > (In reply to Pacho Ramos from comment #7) > > I will commit in a week if nobody disagrees (also CCing systemd team to let > > them review the unit files) > > I still disagree, and see no reason to change that... This disagreement isn't valid as the question of installing a single file depending on a USE flag has been discussed a lot and what you need to do is to use INSTALL_MASK
(In reply to Pacho Ramos from comment #9) > (In reply to Luke-Jr from comment #8) > > (In reply to Pacho Ramos from comment #7) > > > I will commit in a week if nobody disagrees (also CCing systemd team to let > > > them review the unit files) > > > > I still disagree, and see no reason to change that... > > This disagreement isn't valid as the question of installing a single file > depending on a USE flag has been discussed a lot and what you need to do is > to use INSTALL_MASK That's why we have USE=bash-completion?
(In reply to Luke-Jr from comment #10) > (In reply to Pacho Ramos from comment #9) > > (In reply to Luke-Jr from comment #8) > > > (In reply to Pacho Ramos from comment #7) > > > > I will commit in a week if nobody disagrees (also CCing systemd team to let > > > > them review the unit files) > > > > > > I still disagree, and see no reason to change that... > > > > This disagreement isn't valid as the question of installing a single file > > depending on a USE flag has been discussed a lot and what you need to do is > > to use INSTALL_MASK > > That's why we have USE=bash-completion? In fact, not using USE flags just because it's a single file adds up to a LOT of "single files" after a while... Unlike INSTALL_MASK, USE flags can be disabled by default.
(In reply to Luke-Jr from comment #10) > That's why we have USE=bash-completion? Because that packages are using obsolete eclass instead of new eclass that don't add that flag, the same occurs for logrotate case: https://bugs.gentoo.org/show_bug.cgi?id=198901
+*bitcoind-0.8.3-r1 (27 Jul 2013) + + 27 Jul 2013; Pacho Ramos <pacho@gentoo.org> +bitcoind-0.8.3-r1.ebuild, + +files/bitcoind.service: + Install unit file #469670 +