Acee, (Thanks for the plain ascii...)
> Hi Peter, > > Thanks for your detailed response. One thing we decided since my original > query was to move the YANG geo-location type out of the > ietf-routing-types.yang model since we thought that we didnbt have enough > review and experience to model it correctly. We are going to put it in a > separate YANG model and, based on your response, there may be more than > one YANG type. I'm happy to contribute on this, just put me in the loop. A generic Yang model for describing a location would be very useful. Just remember it's not as simple as it first looks. I ended up digging into this when we needed to describe the network as a graph, and you need to describe anything down to the location of a connector on a panel/device in a building and then what content/chanel is actually in that fiber/wire . > We will be discussing this in the future. Please keep me in the loop. --Peter (ps: we can use a yang model to tell time anywhere in universe to...) > Thanks, > Acee > > On 8/23/17, 10:40 AM, "Peter Lothberg" <[email protected]> wrote: > > >Acee, > > > >I missed this message as it was just text encoded, no attachments, my > >obsolete mail ui can't handle it... > > > >>Hi Peter, > >> > >>We wanted to assure that milliseconds of latitude/longitude offer > >>enough granularity for your applications. This slightly more than 3 > >>cm granularity for latitude and slightly less the 3 cm granularity > >>for longitude. > >> > >>Thanks, > >>Acee > > > >We have Geo coordinates to form a graph over the network, from cables > >and ducts in the ground to the wiring inside a CO/POP/Datacenter. > > > >As Vint Cerf says we need to bring IPv6 to the planet as soon as > >March, what we need to future proof this is a format that allows for > >the following; > > > >([Galaxy,Solar System,Planet],)* [Operator], [World], [Lat, Long, > >Height], [Time of Observation UTC in MJD with decimals] > > > > (MJD - Modified Julia Date) > > (Operator is the originator/maintainer of the data.) > > > >The proposal is a good start but you lack some important things; > > > > - You need to be ale to express arbitrary resolution for Lat/Long. > > > > Depending on the use, you might want to truncate the resolution to > > not reveal the exact location of infrastructure, and then again you > > may want to express exactly where things are. > > > > You need to be able to express what we can measure for lat/long > > as an example, so you have to be able to represent this in your > > format: > > > > LAT 59.317647949 North > > LON 18.023390835 East > > > >- When it comes to height, we really have a problem; > > > > - Different countries have different ways of measuring the "sea > > level". > > - The sea level are rising > > - The land is also rising, like our Stockholm location is moving > > about 11mm up and 7mm NE in average every year, so you need to > > have a date on the observation of the position. > > - The local observations have different ages > > - GPS presented height is done with a theoretical model of the earth > > and calculated based on the distance from the GPS satellite. > > > > (Neighboring countries in some cases have more than 100cm > > difference in their definition of sea level.) > > > >Height is a signed value, 1mm resolution, with a source identifier > >(you need way to add new identifiers for the future); > > - WGS84 > > - Local official survey > > - Local height (maybe from basement of building) > > - Local sequence number > > > >- In some cases you want to describe something other than the natural > >world > > itself. e.g., a VPN described as a flat plane, or military / private > > tectical graph that is a different "plane" / "overlay" etc. > > > > So we ended up with a Integer that represents the "world" you are > > describing and the vale of 0 represent the natural world we all live in > > and have as a common reference. > > > >(*) - The astronomers refers to planets in our solar system with a > >coordinate system with the sun as 0,0,0. Everything moves, continental > >drift to ephemerides of planets and satellites. I chose to leave those > >models out of scope as there is enough data in the proposed format to > >go to other sources and augment the positions. > > > > > >-Peter > > > > > > > >> --_000_D576E8B0B65A3aceeciscocom_ > >> Content-Type: text/plain; charset="utf-8" > >> Content-Transfer-Encoding: base64 > >> > >> > >>SGkgUGV0ZXIsDQoNCldlIHdhbnRlZCB0byBhc3N1cmUgdGhhdCBtaWxsaXNlY29uZHMgb2Ygb > >>GF0 > >> > >>aXR1ZGUvbG9uZ2l0dWRlIG9mZmVyIGVub3VnaCBncmFudWxhcml0eSBmb3IgeW91ciBhcHBsa > >>WNh > >> > >>dGlvbnMuIFRoaXMgc2xpZ2h0bHkgbW9yZSB0aGFuIDMgY20gZ3JhbnVsYXJpdHkgZm9yIGxhd > >>Gl0 > >> > >>dWRlIGFuZCBzbGlnaHRseSBsZXNzIHRoZSAzIGNtIGdyYW51bGFyaXR5IGZvciBsb25naXR1Z > >>GUu > >> DQoNClRoYW5rcywNCkFjZWUNCg== > >> > >> --_000_D576E8B0B65A3aceeciscocom_ > >> Content-Type: text/html; charset="utf-8" > >> Content-ID: <[email protected]> > >> Content-Transfer-Encoding: base64 > >> > >> > >>PGh0bWw+DQo8aGVhZD4NCjxtZXRhIGh0dHAtZXF1aXY9IkNvbnRlbnQtVHlwZSIgY29udGVud > >>D0i > >> > >>dGV4dC9odG1sOyBjaGFyc2V0PXV0Zi04Ij4NCjwvaGVhZD4NCjxib2R5IHN0eWxlPSJ3b3JkL > >>Xdy > >> > >>YXA6IGJyZWFrLXdvcmQ7IC13ZWJraXQtbmJzcC1tb2RlOiBzcGFjZTsgLXdlYmtpdC1saW5lL > >>WJy > >> > >>ZWFrOiBhZnRlci13aGl0ZS1zcGFjZTsgY29sb3I6IHJnYigwLCAwLCAwKTsgZm9udC1zaXplO > >>iAx > >> > >>NHB4OyBmb250LWZhbWlseTogQ2FsaWJyaSwgc2Fucy1zZXJpZjsiPg0KPGRpdj5IaSBQZXRlc > >>iwm > >> > >>bmJzcDs8L2Rpdj4NCjxkaXY+PGJyPg0KPC9kaXY+DQo8ZGl2PldlIHdhbnRlZCB0byBhc3N1c > >>mUg > >> > >>dGhhdCBtaWxsaXNlY29uZHMgb2YgbGF0aXR1ZGUvbG9uZ2l0dWRlIG9mZmVyIGVub3VnaCBnc > >>mFu > >> > >>dWxhcml0eSBmb3IgeW91ciBhcHBsaWNhdGlvbnMuIFRoaXMgc2xpZ2h0bHkgbW9yZSB0aGFuI > >>DMg > >> > >>Y20gZ3JhbnVsYXJpdHkgZm9yIGxhdGl0dWRlIGFuZCBzbGlnaHRseSBsZXNzIHRoZSAzIGNtI > >>Gdy > >> > >>YW51bGFyaXR5IGZvciBsb25naXR1ZGUuJm5ic3A7PC9kaXY+DQo8ZGl2Pjxicj4NCjwvZGl2P > >>g0K > >> > >>PGRpdj5UaGFua3MsPC9kaXY+DQo8ZGl2PkFjZWUmbmJzcDs8L2Rpdj4NCjwvYm9keT4NCjwva > >>HRt > >> bD4NCg== > >> > >> --_000_D576E8B0B65A3aceeciscocom_-- > >> > > > > > > > _______________________________________________ Isis-wg mailing list [email protected] https://www.ietf.org/mailman/listinfo/isis-wg
