Hey there, thanks for doing some editing around Baltimore. I noticed you deleted one of our new protected cycletracks. I reverted this changeset because it used old Bing Imagery. If you're editing Maryland, use the "MD Latest 6 Inch Aerial Imagery" in iD. It is higher resolution, leaf-off imagery that is updated regularly. As for this changset, we have a number of new protected bike tracks that are separate from the sidewalk and the car lanes. Watch out for those. Thanks!

user_5359: the OP and I are colleagues and so I walked him through changing this to a more appropriate tag (leisure=common). Cheers!

Phil, I've been looking for a dataset for adding hydrology to OSM that is both high-resolution and accurate. This changeset matches those characteristics, but the NHD data (even NHDPlusv2) isn't very accurate in my experience. For instance, look at https://i.imgur.com/a4Z9hTw.png and see how off the NHD flowline is. Are you manually correcting each flowline with imagery, or is there some other NHD source that I am missing. Thanks!

changeset under discussion on OSM US Slack: https://osmus.slack.com/archives/C1YP8DZ0A/p1502233346926171

Hello there. The planned reservoir will not be an open reservoir as you mapped but rather three large tanks. Since the reservoir isn't constructed yet, and will not look like this, I went ahead and undid your change. Reverted in changeset/52212556

Jan: I'm glad that you found this and commented. I think your idea is superb. I was trying to decide if I should use an existing key like barrier but you make a good point that it could be interpreted by a routing engine as a generic barrier. Better to start something totally fresh.

What do you think about using a colon vs an underscore to separate subtypes? I like the way the colon looks. Is there any functional difference?

I cannot think of any other physical objects that manifest a symbolic border in a religious context. However, I think that a border fence has the same connotation for political divisions. The fence is observable so it gets the fence tag, whereas the boundary is typically set by relation. Once I find the entire length of this eruv (it has to be continuous so that should make it easier to follow), I'd add all the parts as eruv=* and then add some kind of relation. I've seen a few (1-4 on the planet) where people are using a boundary=religious for these. I think that such a boundary might be appropriate, though are usually used for papal states and such with a government. The eruv is somewhat less formal, though I suppose it does have a maintainer and legal backing, so religious boundary might be good for relation.

Jan,

Sure, an eruv is a physical object that serves as a symbolic boundary in the Jewish Orthodox faith. Read more:https://en.wikipedia.org/wiki/Eruv

There isn't a great tag for this yet on OSM, so I added this under the premise of the "Any tags you like" principle (osm.wiki/Any_tags_you_like) until I come up with a formal proposal. Do you have any ideas?

It is a physical thing, a rope or wire but it doesn't serve as a mobility barrier, it doesn't carry power or communication. This sets it apart from other OSM tags. Fellow OSM'er MDRoads and I have been discussing how best to map these. There are a few others on the map but people are taking the approach of using a boundary relation, which I think is OK but I'm more interested in mapping the physical thing that can be observed out in the world. Here are some photos of the Eruv in Baltimore. It is something you can see and touch, unlike most boundaries.

https://goo.gl/photos/9TgKDkDyjWvmTQGG7

Best,
Elliott

CORRECTION: Just one request though: when a sidewalk reaches the end, connect *it* back to the road

Hi there. Welcome to the OSM project! It looks like you have been mapping sidewalks. These look good so far. Just one request though: when a sidewalk reaches the end, connected back to the road (unless it truly ends with no connection). This will help with pedestrian/detour routing. You can use the "crosswalk" preset for mapping the connection. Thanks!

A few issues with this changeset that will prevent it from being accepted. 1) Don't match the edge of a landuse with a road (Wellington Valley)--it is OK to cross a road but not to touch it. Landuse polygons rarely share a boundary with a line or point. 2) super general changeset comments are discouraged. Try to write a sentence about what the purpose of the edit was at least and try to vary it from one edit to the next.

Awesome, there's a link to join on the Maryland wiki osm.wiki/Maryland

That sounds about right about the town/village issue, especially considering the shops along York Road in lutherville go back decades. So both L/T as town then?
I think that the USPS areas are more identifiable. The only issue is that some people do think of Mays Chapel as a place. I know folks that would definitely say that they live in MC, especially around that Grauls on Padonia. Lutherville or Timonium might be too generic for them. Happy to discuss more. If anything, the point for MC should probably be up closer to what I think of as MC, the area bounded by Timonium, 83, Jennifer and then Mays Chapel Elementary.

Indeed, you'll find some oddities in this data, as it is based on the legal land transfers, which sometimes are combined and thus the "plat" is not the boundary that you'd consider on the ground. It ends up being important for official purposes. These subdivisions are good for reference, though sometimes the local mapper knows more than the data tells.

Alright, it is split up. Generally, in OSM we use points to map stores if there is more than one use (store/office/apt.) per building. When there is just one store/office/etc. in a building, we'd just apply the relevant category to the building. If you're interested in mapping the shops there, add those places restaurants, shops, etc. as points.

Perfect! I'll chop up the building and then you can apply any shops or the like. The relation editing is best accomplished via the advanced JOSM editor.

Hey, coincidentally, Baltimore County GIS has published data just today that you can use for mapping subdivisions. This map shows "Conveyences" which are groups of plats (aka subdivision plats) in County GIS jargon. The boundaries are official, however. In the past, I received permission to use County data for OSM (such as all of the buildings on the map). I believe this still applies, so feel free to reference this map for mapping subdivisions. http://data-bc-gis.opendata.arcgis.com/datasets/217c84d9ed914b01afff255f8abf950f_2?geometry=-76.682%2C39.422%2C-76.602%2C39.433

Looks good! I was just over here the other day and scouted a tiny path between that northern edge of the parking lot and The Meadows Neighborhood. Those types of things are useful to add--unofficial type paths, as it helps ped routing software find routes.

Would you say that the way the pavilions are split in this photo is representative of how you'd split them? https://goo.gl/photos/wEbpoGr8T5jCX5GU9

This situation is pretty unique to Baltimore County and that the mapping community, unfortunately, does not have a great answer for. Baltimore County is somewhat unique in the US in that there are no incorporated towns in it. Thus, things like Towson and Timonium don't have any official existence at the local government level. However, two federal entities define these places differently: the USPS and the Census. The USPS ZIP Codes associated towns (the list you cited) are how people identify with a place here. However, the Census, strangely, has decided to break up the 21093 area into these three zones, and the boundaries are not the same. A few years ago I attempted to streamline the hodge-podge list of towns and villages in Baltimore County objectively, by population. However, as you point out, this numbers-only definition doesn't exactly mirror the situation on the ground.
I think that we should go ahead and make MC a village, and then Timonium a town. I always thought of "Lutherville" as the area south of Ridgely that is mostly residential, thus maybe that should be a village. Timonium, meanwhile, does have an urban core of sorts around the fairgrounds. What do you think?

TJ, that is a great point, and I'm glad you made it. I am happy to help out. Place classification is a contentious thing in the community so I do apologize if it seems harsh. Often times we'll see edits where someone will reclassify every town or do some mechanical edits. In the past, we had Lutherville-Timonium combined as one place. The census chopped them up and now classifies Mays Chapel separately. However, as you pointed out, there is no real "urban centre" in Mays Chapel. At best, you might call it the shops along Padonia. Given this point, I think that perhaps the population alone shouldn't be the deciding factor. Do you want to join our Slack channel to discuss? I think you'd find lots of helpful information there, and good people to chat with about the map. Cheers - Elliott