Location Types Registry
draft-ietf-geopriv-location-types-registry-06

IANA Comments:
Upon approval of this document the IANA will create a new registry for
location types. New tokens and updates are assigned after Expert Review by the IETF GEOPRIV working group or its designated successor.  IANA will populate with the initial registrations in the document.

[Ballot comment]
Based on Gen-ART review by John Loughney, but backing up David's Discuss.
The types are quite soft in nature and not algorithmically precise.
The scope for interpretation is wide. This will make both the expert
reviewer's task very vague, and operational choices very unclear.

"I found many of the location types confusing, for
example:

  water:

      The person is on water, such as an ocean, lake, river, canal or
      other waterway.

watercraft:

      The person is traveling in a boat or ship.

What about a swimmer or SCUBA diver?  What if the boat is not moving, is
it still traveling? My gut feeling is that the place needs to be
separate from the action.  Addtionally, prepositions should be separate
from the place (i.e. - on, in, above, under, etc.).

In summary - the location places should be just locations, there should
be no linkage to the activity of the target with respect to the
location."

[Ballot comment]
Based on Gen-ART review by John Loughney, but backing up David's Discuss.
The types are quite soft in nature and not algorithmically precise.
The scope for interpretation is wide.

"I found many of the location types confusing, for
example:

  water:

      The person is on water, such as an ocean, lake, river, canal or
      other waterway.

watercraft:

      The person is traveling in a boat or ship.

What about a swimmer or SCUBA diver?  What if the boat is not moving, is
it still traveling? My gut feeling is that the place needs to be
separate from the action.  Addtionally, prepositions should be separate
from the place (i.e. - on, in, above, under, etc.).

In summary - the location places should be just locations, there should
be no linkage to the activity of the target with respect to the
location."

[Ballot discuss]
Based on Gen-ART review by John Loughney:

The IANA Considerations say:

  To ensure widespread usability across protocols, tokens should follow
  the character set restrictions for XML Names.

This needs a reference. Also, is it really 'should'?

[Ballot discuss]
Just like Bill's DISCUSS, this is a real discuss DISCUSS.

I personally have a different idea what the meaning is of many of the terms in this document and would like to offer some of my own ideas.

Normally, I would have entered a COMMENT for this part of the DISCUSS. However, the number of issues that I have with the definitions is that large that I believe that it is good idea to take second look at the definitions before proceeding.

You will however be able to convince me rather easily as with all human
language issues, opinions can vary widely and I have no intention to override
the opinion of the working group (though I wonder how much review this document
actually got in the working group).

First some general remarks:

unknown:

      The type of place is unknown.

This seems inconsistant with the rest of the document. Everywhere else the definitions are like:

The entity is in a ...

(in this case, the entity is in an unknown location)

Also, I am a bit confused that sometimes you are talking about 'the entity' is in a particular location, while at other times the entity becomes a real person
and you are talking about 'the person' is in a particular location. You might
want to make this more consistant.

Another thing that I noticed that many definitions where defined using examples
instead of defining the term (your aircraft definition is a good example of this)
I believe it is more appropriate to describe the definitions and mention a few examples in addition to the definition (I included an example below in the aircraft section). One could require the same for the IANA registration with an example field.

Also, the definitions will sound less awkward if the 'the entity is in' part is simply omitted and that you only describe the location.

Now for the definitions:

aircraft:

      The entity is in a plane, helicopter or balloon.
                        ^^^^^

                    s/plane/airplane/

As an aviator, I believe that there are better definitions for the
term aircraft. The FAA defines an aircraft as follows:

---
aircraft:
    means a device that is used or intended to be used for flight in
    the air.
---

Examples are (these are my own):

    helicopters, airplanes, gyroplanes, gliders and lighter-than-air
    devices like balloons

So my defintion would be:

aircraft:
    entity is in a device that is used or intended to be used for flight in
    the air.

    Examples:

    helicopters, airplanes, gyroplanes, gliders and lighter-than-air
    devices like balloons

====


bus:

      The entity is traveling in a public or charter bus.

I don't understand the distinction between 'public' or 'charter'.
As far as me concerned, a 'bus' is vehicle used for mass
transportation of people.

=====

cafe:

      The entity is in a cafe or coffeeshop.

A cafe is something different from a coffeeshop in US english (in my limited experience). A cafe
is usually a restaurant and people look really strange at you if you
ask for just a coffee (you better don't drink a coffee in a 'cafe'
either because it usually looks like a dark colored water and has no
taste). Occasionally, somebody tries to have the cachet of a european
coffeeshop and calls their coffeeshop a cafe. This is all very
confusing to me and I recommend to check with a more authentic
american than me. To make things even more fun, people will look really
strange at you if you go to a coffeeshop in Amsterdam and order a
coffee.

===

cycle:

      The entity is riding a bicycle, motorcycle or similar vehicle.

I personally feel that bicycle should have it's own category.

===

club:

      The person is in a dance club or discotheque.

While this is certainly one definition of a club, I believe the more
common meaning is:
The building, room, or other facility used for the meetings of an
organized group.

===


prison:

      The person is in a prison, penitentiary, jail, brig, or criminal
      mental institution.

Different countries have different ways of treating people with mental
illnesses who committed crimes. In some countries the emphasis is on
the 'prison' part, in other countries it is on the 'mental
institution' part. I don't believe it is good idea to connect the term
mental institution with a prison.

===

restaurant:

      The entity is in a restaurant, coffee shop or other public
      dining establishment.

I consider a restaurant soemthing different than a coffee shop, eg. I
don't go to a restaurant when I want to drink a coffee, neither do I
go to a coffee shop to enjoy dinner. Basically, there is some overlap between
restaurants and coffee shops but most people mean something very different
when they are talking about a restaurant versus a coffee shop.

My dictionary says:

restaurant:

  - a place where meals are served to the public.
  - a building where people go to eat

(I admit that I found a dictionary entry that defined a coffee shop as a
small restaurant ;-))

I also noticed that my spell checker insists that it is actually:
'coffee shop' and not 'coffeeshop' (see your entry for 'cafe:')

===

water:

      The person is on water, such as an ocean, lake, river, canal or
      other waterway.

So how about people who are in the water ?

[Ballot discuss]
This is a discuss-discuss.  What's the intended relationship between the names in the location types registry and the RPID place-type?  RPID says that values can be registered in place-type (although that's not explicitly listed in the IANA considerations, just mentioned in the place-type section); does that mean I can register "foobar" in RPID's place-type registry to mean "The presentity is upside-down hanging from a cliff by his toes", and in geopriv-location-types-registry to mean "The entity is lying on a comfortable couch, being fed grapes"?  Given the overlap in the initial values, I'd bet that people would assume that the same values had the same meanings in both places, and might not bother to register a given value in one or the other registry. I'd feel much better if there was a way to have a single registry and a pointer.

[Ballot comment]
To my great surprise, I'm a no-objection on this document rather than
a discuss or abstain.  The authors have done a good job of explaining
how the registry might be used in enough detail that protocol
designers can determine if this registry is appropriate.  Also, the
IANA considerations are much improved.  Thanks for the great work with
last call comments.

At first cut, a discussion of an applicability statement of the registry may be required and a reinforcement that this is about tokens, rather than internationalizable strings.

[Note]: 'There were significant discussions during IETF Last Call; discussion with the WG Chairs and authors is needed to determine how to address the objections.  At first cut, a discussion of an applicability statement of the registry may be required' added by Ted Hardie