IETF
paws@jabber.ietf.org
Tuesday, November 5, 2013< ^ >
psa has set the subject to: PAWS WG | http://tools.ietf.org/wg/paws/
Room Configuration
Room Occupants

GMT+0
[18:48:24] Andy Lee joins the room
[20:45:37] mccap joins the room
[22:23:18] Rosen, Brian joins the room
[22:23:28] <mccap> Hello.
[22:23:37] <Rosen, Brian> hi
[22:23:39] <mccap> This is Pete McCann
[22:23:43] <Rosen, Brian> I'll jabber scribe
[22:24:02] <Rosen, Brian> if you have questions to raise, I will bring them to the room
[22:24:03] dseomn joins the room
[22:25:19] <Rosen, Brian> anyone with ipr issues on the two declarations?
[22:25:35] <mccap> Peter seemed to indicate that the new IPR applied to spectrum slope specifications.
[22:25:35] resnick joins the room
[22:28:41] <Rosen, Brian> okay, I'll double check with Peter
[22:28:57] <Rosen, Brian> I thought he was just updating the IPR statement to cover the actual draft
[22:30:04] <mccap> His e-mail said he was updating due to the new non-zero slope encoding getting incorporated into the WG draft.
[22:48:58] <resnick> Make sure to get the "Nokia Internal Use Only" off the bottom of the slides at some point.
[22:49:13] <resnick> (Oops. ;) )
[22:50:16] <Rosen, Brian> I've already mentioned it to Gabor!
[22:51:04] <resnick> No big deal, but we probably don't want it in the official proceedings.
[22:52:32] <Andy Lee> FCC does allow multiple channel use
[22:52:46] <Andy Lee> Ofcom also allows multi-channel use
[22:53:06] <mccap> And OOB emissions would need to be specified on either side of the group of contiguous channels.
[22:53:23] <mccap> We should not be using the term "channel" in the protocol.
[22:53:36] <Andy Lee> Agreed
[22:54:03] <resnick> So is this just a matter of properly specifying the semantics of this?
[22:54:33] <Andy Lee> Two issues:
[22:54:41] <Andy Lee> 1) should we allow slope encoding
[22:54:50] <Andy Lee> 2) what are the semantics around band edge
[22:55:05] <resnick> Seems like we want to answer 2 before we answer 1, eh?
[22:55:24] <Andy Lee> Yes.  I'd prefer to separate the two
[22:57:15] <Andy Lee> Yes, Ofcom does allow this
[22:57:40] <mccap> Ask this: is the device required to put a gap between two contiguous channels?
[22:58:15] <mccap> I would expect not.
[23:00:09] <mccap> The mask is not a per-channel mask.  It is a mask for the whole range of spectrum.
[23:01:57] <Andy Lee> In Ofcom, a device will only get the spectrum profile appropriate to its device class.
[23:02:18] <resnick> Please preface comments with "MIC:" if you want them read in the room.
[23:02:59] <resnick> Did either of you want those comments repeated for the room?
[23:03:28] <Andy Lee> MIC: Under Ofcom rules, the spectrum profile is already adjusted according to the device class (OOB emission limits)
[23:03:38] <mccap> MIC: because when you combine contiguous channels, the resulting mask is not the same as the sum of the channel masks.
[23:06:46] <Andy Lee> How about enable encoding of sloped now, but defer semantics of OOB interpretation.
[23:08:33] <Andy Lee> Who was speaking?
[23:08:52] <mccap> I am in favor of generic slope encoding.
[23:09:29] <mccap> Being in the current version.
[23:10:15] <resnick> Doug Otis was the previous speaker.
[23:10:22] <Andy Lee> thx
[23:10:27] dharasty joins the room
[23:14:12] <Andy Lee> The DB *only* sends the availability mask.
[23:14:28] <Andy Lee> I was not expecting the DB to tell the device its transmit mask.
[23:19:27] <mccap> Cannot hear the speaker.
[23:19:59] <mccap> Mic is not on.
[23:20:11] <Rosen, Brian> is that better?
[23:20:26] <mccap> Better, thanks.
[23:24:11] <Rosen, Brian> For my notes, mccap, are you Peter McCann?
[23:24:18] <mccap> Yes.
[23:24:21] <Rosen, Brian> thanks
[23:25:21] <Andy Lee> MIC: We've heard from Ofcom that the UK listing service needs to be consulted, but it doesn't need to be the *first* server contacted.
[23:27:05] <resnick> Brian has seen it and will relay.
[23:29:18] <mccap> It needs to map geographic location to regulator server address.
[23:29:46] <Andy Lee> Don't know if the IANA servers can handle calls from 100's of millions of devices.
[23:30:45] <mccap> Just put the nation-state map on a CDN.
[23:39:48] <Rosen, Brian> the timezone database handles quite a lot of queries
[23:40:08] <Rosen, Brian> But I agree that 100s of millions every 24 hours is a pretty large load
[23:40:26] <mccap> I thought it was a file that was only seldom updated.
[23:40:42] <resnick> Yeah, I think it may be downloaded infrequently.
[23:40:56] <mccap> Like, only with OS updates.
[23:41:16] <resnick> (The other thing that popped into my head was something DNS based. I hit myself in the back of the head immediately.)
[23:41:22] <mccap> :)
[23:41:54] <Andy Lee> Trust model is the other elephant in the room
[23:42:17] <mccap> Is IANA in a position to arbitrate the geographic boundaries of nation-states? ;)
[23:42:36] <resnick> It's willing to point that folks who are. :)
[23:42:52] <mccap> We need a UN working group on this.
[23:42:54] <resnick> to point to folks who are.
[23:43:12] <mccap> ITU-R. ;)
[23:43:38] <resnick> @mccap: :-o *That's* why I'm inclined to throw something back over the wall to the IAB to handle.
[23:43:42] <Rosen, Brian> please, no!
[23:44:18] <mccap> IAB needs to think about it.  I think ITU-R is one of the folks who should potentially be consulted.  Seriously.
[23:44:24] <Andy Lee> MIC: My impression from discussion with Ofcom is that they are OK with databases being contact before the listing service verification, so long as the database doesn't transmit until everything checks out.
[23:44:41] <Andy Lee> Sorry, device doesn't transmit
[23:45:07] <mccap> A device that isn't in the UK shouldn't be required to contact the UK listing service at all.
[23:49:54] <dharasty> If I step off the plane.... I'm probably not carrying a master device.  I'm probably carrying a slave, right? So I'm not using PAWS anyways... I'll just find the nearest master.
[23:50:28] <mccap> Not the fault of the authors.  We need IAB action.
[23:50:41] <Rosen, Brian> for what?
[23:50:53] <mccap> To liaise with regulators.
[23:51:15] <Rosen, Brian> ah, okay, that's what Pete Resnick is saying
[23:51:20] <mccap> Yes.
[23:52:34] resnick leaves the room
[23:52:43] mccap leaves the room
[23:53:56] Andy Lee leaves the room
[23:55:49] dharasty leaves the room: offline
Powered by ejabberd Powered by Erlang Valid XHTML 1.0 Transitional Valid CSS!