16:02:00 <kalev> #startmeeting FESCO (2017-05-05)
16:02:00 <zodbot> Meeting started Fri May  5 16:02:00 2017 UTC.  The chair is kalev. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:02:00 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
16:02:00 <zodbot> The meeting name has been set to 'fesco_(2017-05-05)'
16:02:00 <kalev> #meetingname fesco
16:02:00 <kalev> #chair maxamillion dgilmore jwb nirik jforbes jsmith kalev sgallagh Rathann
16:02:00 <zodbot> The meeting name has been set to 'fesco'
16:02:00 <zodbot> Current chairs: Rathann dgilmore jforbes jsmith jwb kalev maxamillion nirik sgallagh
16:02:03 <kalev> #topic init process
16:02:12 <kalev> who's around for a fesco meeting?
16:02:12 <nirik> morning
16:02:16 <kalev> morning
16:02:29 <zdohnal> good evening :)
16:02:45 <sgallagh> .hello sgallagh
16:02:46 <zodbot> sgallagh: sgallagh 'Stephen Gallagher' <sgallagh@redhat.com>
16:03:25 <kalev> should be a short meeting with just one ticket on the agenda
16:03:26 <jforbes> .hello jforbes
16:03:27 <zodbot> jforbes: jforbes 'Justin M. Forbes' <jforbes@redhat.com>
16:03:36 <jsmith> .hello jsmith
16:03:37 <zodbot> jsmith: jsmith 'Jared Smith' <jsmith.fedora@gmail.com>
16:03:39 <zdohnal> .hello zdohnal
16:03:40 <zodbot> zdohnal: zdohnal 'Zdenek Dohnal' <zdohnal@redhat.com>
16:04:32 <dgilmore> hola
16:04:46 <kalev> ok, let's get started, looks like we have plenty of people around today
16:04:50 <kalev> #topic #1701 Disable cups service by default in most editions
16:04:52 <kalev> .fesco 1701
16:04:53 <zodbot> kalev: Issue #1701: Disable cups service by default in most editions - fesco - Pagure - https://pagure.io/fesco/issue/1701
16:05:26 <sgallagh> zdohnal: So I hope you will forgive me if I ask you to clarify some things, because I couldn't parse all of your most recent reply.
16:05:54 <zdohnal> sgallagh: ok
16:06:00 <sgallagh> What I don't understand is why remote clients who tried to use the printer wouldn't just trigger it to autostart.
16:06:15 <sgallagh> Why it would have to be up and running and consuming resources all the time for them to use it.
16:06:34 * kalev had the same question.
16:08:18 <kalev> my understanding of this is that starting cups.service starts the actual service on boot, whereas when cups.socket is enabled then it's systemd listening on the cups socket and cups gets autostarted when there's an actual request coming in
16:08:32 <kalev> making booting faster
16:08:38 <zdohnal> Because cupsd is sharing printer queue in this setting - it is showing this queue for all systems that can access cups on server
16:08:40 <dgilmore> sgallagh: I guess for the use case or running as a network print server
16:08:50 <kalev> ahhh, right
16:09:05 <jforbes> Sure, but that has to be configured anyway doesn't it?
16:09:09 <sgallagh> OK, so it has to be running for autodetection to work?
16:09:13 <dgilmore> and advertising the print queue
16:09:19 <sgallagh> jforbes: That was my next question :)
16:09:30 <sgallagh> CUPS doesn't do this by default, it requires configuration.
16:09:54 <sgallagh> So if that's the case, is it really that much more work to add `systemctl enable cups.service` to the steps?
16:09:58 <zdohnal> yes, that cupsd is advertising the print queue, as dgilmore said
16:10:46 <dgilmore> zdohnal: does that use case work without configuration?
16:10:54 * nirik thinks we are only talking about server now. I cant see much use for it in cloud or docker... although I am sure there's some oddly named openstack project to provide printers.
16:10:55 <zdohnal> sgallagh: it is not, but someone can complain about it - if they use it this way
16:11:15 <dgilmore> zdohnal: I plug a printer into a server, what do I have to do to share it to the world?
16:12:07 <zdohnal> gdilmore: when you are installing that printer by lpadmin, you need to add '-o printer-is-shared=yes' option
16:12:16 <zdohnal> dgilmore ^
16:13:01 <zdohnal> nirik: I only thought about edition which can be used as print server in the most probability
16:14:01 <nirik> well, you would need to go to some pretty far lengths to get a printer exposed to a container or cloud instance IMHO.
16:14:13 <dgilmore> zdohnal: so maybe only for server we leave cups.service enabled
16:14:28 <dgilmore> or we just say you have to enable the service
16:14:46 <dgilmore> but I would say that enabling the socket by default globally is correct
16:15:08 <dgilmore> and then its up to the server WG to decide what they want to do there
16:15:18 <zdohnal> dgilmore: agree
16:15:24 <sgallagh> dgilmore: For the record, the socket is already enabled globally
16:15:30 <sgallagh> But so is the .service file :)
16:15:49 <dgilmore> sgallagh: so the change would be to change cups.* to cups.socket
16:15:55 <sgallagh> So we're actually discussing *removing* the global enablement of cups.service from 90-default.preset
16:16:03 <jforbes> Right, I just don't see why it even needs to be enabled by default anywhere. Most printers these days have their own print servers, I wonder what percentage of server installs are using cups to share printers
16:16:03 <dgilmore> sgallagh: right
16:16:31 <sgallagh> dgilmore: Well, cups.socket and cups.path, but close enough
16:16:32 <dgilmore> jforbes: likely only in corporate settings for accounting and auditing
16:16:58 <Southern_Gentlem> jforbes, all of mine
16:17:38 <jforbes> I am not saying users don't exist, I am asking what percentage of users actually use the feature
16:18:39 <jforbes> Either way, I suppose that is a better question for the Server WG to decide
16:19:01 <dgilmore> jforbes: indeed
16:19:09 <sgallagh> jforbes: +1
16:19:17 <dgilmore> I think its a small amount of capable users
16:19:35 <zdohnal> kalev: When you have cups-browsed turn on, cups.service will be still running - because cups-browsed will start it - so there will be maybe little faster, but still cups.,service will be running when you are using network printers
16:20:00 <sgallagh> Proposal: Drop `cups.service` from the 90-default.preset file. Defer to Server WG on whether to leave it enabled on that Edition.
16:20:14 <jforbes> sgallagh: +1
16:20:17 <kalev> sgallagh: +1
16:20:23 <zdohnal> sgallagh +1
16:20:27 <nirik> +1
16:20:30 <kalev> sgallagh: will you bring it up with the Server WG?
16:20:35 <sgallagh> kalev: yes
16:20:37 <kalev> awesome
16:21:13 <jsmith> sgallagh: +1
16:21:24 <sgallagh> zdohnal: If cups-browsed pulls in cups.service anyway, I'm not sure there's any value to cups.service being enabled on its own at all.
16:21:37 <dgilmore> proposal #agreed globally we will enable cups.path and cups.socket, sgallagh to take enabling cups.server to the server WG
16:21:45 <dgilmore> +1 to sgallagh
16:21:55 <kalev> #agreed Drop `cups.service` from the 90-default.preset file. Defer to Server WG on whether to leave it enabled on that Edition. (+1:7, 0:0, -1:0)
16:22:07 <dgilmore> ignore my proposal
16:22:12 * dgilmore was too slow
16:22:12 <sgallagh> #action sgallagh to take the question to Server WG
16:22:26 <zdohnal> sgallagh: sorry, I talked about workstation edition in that comment
16:22:26 <kalev> zdohnal: you ok with the outcome?
16:22:44 <zdohnal> kalev: ok with that
16:22:49 <kalev> excellent
16:23:08 <kalev> moving on.
16:23:13 <kalev> #topic Next week's chair
16:23:20 <kalev> who wants to chair next week?
16:23:32 <jsmith> I'll be on a plane :-/
16:23:38 <jsmith> Otherwise, I would have volunteered
16:23:39 * nirik will not be here next week
16:24:30 * dgilmore will do it
16:24:42 <dgilmore> though I hope we have enough
16:24:44 <kalev> #info dgilmore to chair next week
16:24:50 <kalev> great, thanks dgilmore
16:24:57 <kalev> #topic Open Floor
16:25:30 <kalev> anything to discuss here?
16:26:31 <dgilmore> reminder to get f27 requests in
16:26:47 <dgilmore> releng just about has everything planned for f27 deliverables
16:27:09 <zdohnal> kalev: I would like to ask you more about faster booting later in PM, so go on
16:27:15 <dgilmore> if you want something from us you need to get it in now or you will miss f27 and have to wait to be considered in f28
16:28:45 * kalev nods.
16:29:10 <kalev> ok, thanks for coming everyone! see you next week
16:29:13 <kalev> #endmeeting