[Talk-us-massachusetts] Why Barnstable is such a mess, and other MADness (Wayne Emerson, Jr.)

[Greg Troxel]

Yury Yatsynovich <yury.yatsynovich at gmail.com> writes:

> Sure, if adding villages as admin_level=8 is relatively easy, I would
> prefer this solution instead of adding addr:suburb.

I really am having a hard time understanding where you are coming from.
You see to be accepting things as axioms that I believe are not true.
And you seem to be willing to do things that are wrong, specifically
using addr:suburb and changing level8 boundaries to things that are not
actually level8 entities.  I will assume that this is due to not really
being clear on how things work in terms of governments and addresses.

First: addr:suburb is just wrong in massachuetts.  It refers to a
concept, explained well by Bill Ricker, that exists in other countries,
where a town is somehow subordinate to a bigger town.  Here, we use the
word suburban as an adjective, but it is entirely disconnected from both
government and addressing.  We might say "Belmont is a suburb of
Boston".  But Belmont is a level8 entity, and so is Boston.  It is not
contained in Boston.  And addresses in Belmont do not use the word
Boston.   We use addr:town=Belmont in Belmont addresses.

The problem with suburb here is that you can say this about Belmont and
be right, and Lexington, but Concord might bristle a little, Littleton
would wonder, and by the time you get to Winchendon they will think you
are just confused.  By Erving it is crazy talk.  But every one town is
much like the previous, and there is no good place to cut.  But more
importantly, nobody outside of OSM would think of formal suburb
labeling, only in this fuzzy sense of describing things that exist as
towns.

I'm not sure you understand how towns work, given your comments.  It is
certainly not obvious and cannot be figured out by first principles, so
I'll explain:

  In Massachusetts, there are entities called (formally, in state law)
  cities and towns.  They are the same thing, except that cities have a
  city council and often a mayor (nitpicking, there are 5ish types of
  cities that have slightly different rules, but city council is a fair
  if slightly fuzzy generalization) and towns have a town meeting and a
  board of selectmen.  Henceforth I will call these all towns, as cities
  are just towns with a city council instead of a town meeting.  Every
  bit of land is in one town or another.  Towns are level8 entities,
  because that's what level8 is supposed to be about, and Mass is so
  rigorous about town/not and every land in a town that this is pleasing
  and neat.  Towns have planning boards, zoning bylaws, and usually
  police departments, fire departments, etc.  People pay taxes to towns,
  and register to vote with their town clerk.

Village is another confusing term.  In OSM, there is a notion of the
hierarchy of populated places, which is a geography term not aligned
with admin boundaries.  In this scheme, a village has fewer people than
a town and more than a hamlet.  There is no implication of inclusion of
one in the other, Suburb, in OSM, is another one of these populated
places tags, and is sort of a town that is subordinate to a city.

In Mass, we use village in this populated place name way occasionally.
It shows up in minor ways in place names, not with any formality; Stow
has "Lower village", which is a name for a place of indeterminate extent
with zero legal standing.  It is not an admin_boundary.

More relevant to addressing, we also use village as a named subdivision
of a town/city, where there really is a boundary line (if only on an
official town map).  In this way, these village names can be said to be
level9 entities (or we could call them 10, but they feel like 9 to me,
so I agree with Alan, and IMHO this list gets to decide).

This use of village is very different from named areas of towns that
don't have boundaries.   Examples of non-boundary place names are "West
Acton", "Waverley" (Belmont), "Harvard Square".

I am only aware of three towns that really have villages in a formal
sense: Boston, Newton, and Barnstable.  Of those, I am only aware of
Barnstable using the village names in addresses.  Perhaps Bill can
comment on Boston -- while he said the post office would deliver mail
addressed three ways, I would bet that there is one form which is the
official legal address, that appears on tax bills, that you get in jury
duty summons, that they want on your driver's license, etc.  I think in
Boston this is "1 Elm Street, Boston, MA" even if you are in Back Bay or
Roxbury or some other named area.  For these, there is no problem;
addresses have a town name that is the level8 entity.


So then there is the question of Barnstable.  First, we need to talk
about what I think you are assuming, wrongly, is an axiom, which is that
the thing that seems to be where you expect a town name in an address
must match the name of the level8 admin boundary.   This is just not a
true statement about how our world works.  Here is the official  page of
the town of Barnstable.

  https://townofbarnstable.us/Departments.asp

It gives the town hall address as

  Town Hall 
  367 Main Street
  Hyannis MA 02601

Calling the boundary of Hyannis as level8 would be just wrong; Hyannis
is not a town.

But, putting address tags on the town hall as addr:city=Hyannis is fine;
that is *in terms of addressing* treated as if it were a town.

So I don't see anything to resolve - there is no rule that the town
field and the level8 entity has to match.  We simply can allow QA
passing for addressses in specific level8 entities where the addr:city
field has specific values known to be sub-parts of those towns and that
are known to be used in addresses.

> When mentioning "not allowed" to import zipcodes, I was expressing my
> suspicions that MassGIS may probably not have rights to distribute the
> zip-code data, though, it is just my guess

OK.   I'm not sure the PO cares any more, but I don't object to leaving
it out.