OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

 


Help: OASIS Mailing Lists Help | MarkMail Help

cti message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]


Subject: Re: [cti] Location Object


Jason Keirstead wrote this message on Sun, Aug 12, 2018 at 22:44 -0300:
> I agree with this proposal. Is this really needed, and realistically how 
> often will anyone provide it, and if so how often would consumers process 
> it. 
> 
> (FWIW, you can't even populate this value from Maxmind originated location 
> data since it's definition disagrees with how they communicate accuracy, 
> which is based on a radius)

We have defined precision AS a radius.  Per WD02:
"The actual Location may be anywhere up to precision meters from the defined point."

Here is MaxMind's definition:[1]
"The approximate accuracy radius, in kilometers, around the latitude and longitude for the geographical entity (country, subdivision, city or postal code) associated with the IP address."

Our definition is the same as MaxMinds except that we use meters instead
of kilometers.  (If MaxMind were to indicate precision of 100 meters,
they would use .1, while we would use 100.)

That MaxMind includes precision in their data shows that we should include
it in ours.  It is a simple matter of multiplying by 1000 their accuracy
radius to populate the precision property.

Removing this optional property significantly degrades the ability to
convey location data, and decreases the confidence in the accuracy
of the provided data.

[1] https://www.maxmind.com/en/geoip2-precision-insights

> From:   Bret Jordan <Bret_Jordan@symantec.com>
> To:     "cti@lists.oasis-open.org" <cti@lists.oasis-open.org>
> Date:   08/12/2018 09:06 PM
> Subject:        [cti] Location Object
> Sent by:        <cti@lists.oasis-open.org>
> 
> 
> 
> All,
> 
> When we first worked on the Location object, there was some discussion 
> about adding a precision property for lat/long.  Back then, I went along 
> with the request and claims being made.  However, after letting this sit 
> for a while (6 months), I now think that precision for lat/long is just 
> not worth it. The amount of extra complexity it will require on 
> implementations and systems not knowing what to do with it, makes this a 
> solid candidate for a custom property.
> 
> We have a history in the TC about debating precision on timestamps, and I 
> feel like this falls in to that same bucket. We decided to elide timestamp 
> precision and I think we should do the same here with lat/long precision 
> in the Location object. 
> 
> If we find in the future that we honestly really do need it, then sure, we 
> can see about adding it.  I just do not feel like it is worth it right now 
> and would strongly propose that we remove precision from Location before 
> it is included in 2.1. 
> 
> Thanks
> Bret
> 
> 
> 
> 
> 
> 

-- 
John-Mark


[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]