So. First step - Sussex county? It should be possible to look at the Cities Overlay KML file in WME and/or in GoogleMyMaps and make some decisions.

The issue of whether to use the "expected city name" (usually identical to the mailing address city) or the "official" name as primary needs to be laid to rest before beginning.

While the details are being hashed out, I propose to the New York State community a simple guideline that can be added to the wiki page for the reference of all editors, and especially new editors and visiting editors from other states:

* State Manager approval needed for any changes to polygon boundaries (segment primary city name).
* State Manager approval needed for the addition of any new city names (used in segment primary or secondary, RPP address, or public place address)
* State Manager approval needed for addition of alt names to segments to fix address search problems related to HN or RPP. MC must be used for this information.

I think a simple google sheet could be used when these types of changes are made - problem. solution type (rpps, change to primary name, change to alt name, fix in google maps). Whether it was successful. This could help inform the discussion of developing guidance for the benefit of SMs and everyone else.

The situation is (mostly) much less complicated in PA, but based on our experience there, I think it can be useful to think about the two different uses for city names quite separately.
1) segment primary name used to create "city polygons"
* in which place names float around the map
* which appear in hazard reports ("near Newark, DE")
2) city names used in address search and displayed Place addresses
* added in alt addresses on segments so HN and RPP search can work properly
* used in RPP in cases where HN won't work, or in cases where wazers add RPPs with photos, so that address search by mailing address will retrieve an RPP when appropriate
* used in public point and area places so that wazers can choose the chain location they actually want to drive to.

This is still being actively worked on. The problem is the variability with how different parts of the state are structured and it's hard to just "draw a line" on which parts of the state follow which guidance. For now, Long Island is using the Hamlet names instead of Hempstead.

This discussion doesn't work nationally or even at the Northeast level, because this pattern varies state to state.

The more I look at this, the more I see this not scaling up to a regional approach and we may even have to divide NY state into multiple sections. The basemap import brought all of the Long Island (Nassau and Suffolk County) city and village boundaries, with the hamlet boundaries filling in the rest. The incorporated town names are not used here at all, except when a village shares the name of the town (e.g., Hempstead, Oyster Bay) and the name is representing that village. Searching by the incorporated town name has come up with the correct hamlet with the examples I have tried to search.

These hamlets dwarf their upstate counterparts in population density and I don't think it's a fair comparison to make a blanket rule for the whole state. This type of use works well for densely populated counties such as those in Long Island, as well Rockland County and possibly Westchester. But beyond those areas, CDPs and Hamlets do not play as important of a role and do not have as defined of boundaries as they are in these counties. Upstate areas seem to have more conflict with USPS crossing incorporated town lines than CDP/hamlet names, but that gets into a point beyond what I want to post right here.

As for NYC, I need to defer to local guidance. I think neighborhoods will work similarly to the hamlets of Long Island, where a search for a Brooklyn address should still return a neighborhood address when setup, if there's no overlap in names and HNs. One of the boroughs would be a good testing ground, as it seems Long Island is already setup in this aspect.

Do you have some examples of hamlets that aren't mapped yet? Looking at GIS, it looks like almost everything is in place for Nassau and Suffolk. It seems roads were used as borders for many, so there may be a big alt name project. But overall I think the names are there and boundaries mostly correct.

After a bit of research, it seems Rockland is the only other county south of I-84 that has the same official use of CDPs and hamlets as the Long Island counties. Orange, Westchester and Putnam do not have any official hamlet boundaries. This trend continues northward.

One more wrench into using CDPs elsewhere: their evolution over time. I haven't been able to find a list, but there were 24 new CDPs added between 2010 and 2018 in the state. In areas like Long Island and Rockland, the general density of housing is fairly stable and I expect that list hasn't changed. I am currently gathering the old lists and new lists to see what changed and will report back when I get a chance to thoroughly compare.

I also know of at least 2 villages that have also dissolved in that timeframe, making them hamlets that have quasi boundaries.

At the moment, the county is already named after most of the hamlets. You might switch it to just those few town and villages, but the locked polygon layer will still show those dozens of other shapes and names.

Edit: Just realized you said app. So it would be a good time to check that out now and see if it's too crowded.

So after some research, I think we need to stop considering the use of ZIP codes alt names and continue to review city name URs on a case by case basis. In some situations, we have ZIP codes the size of towns, covering several hamlets and even multiple villages. And these do not necessarily match the names of the towns. We run the risk of duplicate house numbers fairly quick, so I think it's best to leave search to handle this end.

However, I think we should move forward and test hamlets in Suffolk County. The few towns here are large, and there is either a village or hamlet that uses this town name within the town. To avoid having to use a name format of "Town of", "Village of", or "Hamlet of", I think sticking with villages and hamlets are the best bet. According to the Suffolk GIS viewer, villages share the boundaries of their respective hamlets, so there are no border issues geometry-wise.

Based on this and JS's previous work, I proposed a small test project with Sulfolk hamlets to: 1) Fill in the few areas with "no city" and 2) Check the borders of the hamlets. Many borders are actually divided down the center of roads, so these segments need to be checked to ensure there are both hamlet names on the street. The primary can be decided with discretion, based on what is present on that road.

I have swept through and there are a few CDPs listed in Waze that are not drawn on the hamlet map. Thoughts on keeping them? North Bay Shore and Northampton (2). We also have the Shinnecock Reservation in Waze, but it is not included in the hamlets. I do believe that should stay, as we keep other reservations names.

DrivingWithBill wrote:

DrivingWithBill wrote:I was following up to see if there was any updates on the hamlet testing in Suffolk County, its been several months and I was hoping that there was more of a clear picture or feedback available at this time?

It’s been two weeks and no feedback from ANY New York State Managers. Disappointing to try and advance this topic and get complete radio silence.

I sent you a message during this timeframe explaining the situation and that testing is not complete. I have not had a chance to write up a more formal set of results for a full forum post.