Issues with this road:
1) It was supposed to be changed from FL 9B to I 795, but I cannot find any evidence online that it officially has been changed.
2) Even so, it has been changed in OSM, but only partially; some sections still say 9B.
3) It is "I 795", not "I-795" with a hyphen.
I don't want to go through the process of correcting it until I am sure if it has been officially made I 795 or not.

Not to mention the landuse values are just plain wrong. Gas stations and florists are not 'commercial'. Residential does not cover houses, zig zags through properties as if the person who placed them was drunk. Source says RIGIS. Haphazard at best.

In the effort to paint every square inch of RI with landuse values, there seems to have been an unhealthy disregard to the street grid. Nodes placed to create these overly complex TIGER-like landuse features have yanked streets out of alignment, particularly when differing landuses were forcibly made to contact each other for no good reason. Where egregious, my method is to detach the landuse features from each other and the streets, then align them with the properties off the streets' sides.

It appeared that the international border was accidentally skewed by being erroneously attached to a natural=wood feature. I could not find any evidence that the border was anything but straight. Aligned border by lining up Monument 19A and 19C, and moving 19B in line.

justification: these ramps were specifically built to facilitate truck traffic between parkway and 21 St / Talleyrand, and guidance signs relay this. They are NOT connectors for Phoenix and 21 St (i.e. not tertiary links).

Justification for extending I 10 eastbound, and why the southernmost roadway: 1) the mainline clearly continues to the right at the split over the RR tracks; 2) two more exits signed on I 10 east actually exit from the southernmost roadway, with the last exit reachable from I 10 (and not I 95) for Forest St; 3) signs on Stockton St indicate the onramp to this roadway is for I 10 East / To I 95

@cubbe8 My reasoning in switching the classification is that Washington Street is not directly accessible from the Newton Corner interchange, Park/Tremont by design is the route from Newton Corner to Oak Square, and the bus route is along Park/Tremont.

This subdivision was added to the map in 2019, but no aerial shots before or since show this was built or even ever existed. Moreover, it appears the town line was shifted on the map. Searching the internet did not reveal any plans or discussion on such a development or a shifting of the boundary. The town line was reestablished using the MassGIS L3 Parcels overlay, and the buildings and roads were tagged "proposed" to maintain them as hidden placeholders.

Changed canal=qanat to man_made=pipeline which is more appropriate IMHO

Declassified from tertiary to residential. Mainly, it is not a connector for 1A South and Revere Beach Parkway, those movements are handled by direct ramps and MA 16. No signage directs drivers to use this road as a connector. Also, dense housing, narrow ROW, street parking on both sides and residential speed limits. Not tertiary.

A situation commonly found in New England, residential roads turned into highway ramps or vice versa. Could be mapped with a classification like tertiary, as trunk_link, as residential, etc. I chose "unclassified" because its classification is vague. I named it because the house addresses are Revere Beach Parkway, and a driver turning right from Wilson St would turn onto what street? Revere Beach Parkway. Put another way, this is more a frontage road than a ramp. Surely this will raise eyebrows due to its appearance, but I feel this treatment is technically more appropriate and not mapping for the renderer.

This is an odd intersection for many reasons. The main issue was determining the road classifications and names, and admittedly there are other ways to do it. I chose to keep the main tertiary way on the split, and the section leading to and from East Street as tertiary_link. I named the latter segment Cambridge Street because turning right from 28 South leads to that street and I feel it is technically not part of East Street.

This was not named explicitly by me, I just kept the name that had been tagged previously. Did not change it at the time because I was unsure. In retrospect, I suspect it was vandalism.

You'll be happy to know that I have abandoned my old practice of making all T intersections perpendicular, and I have adopted a case-by-case approach to better represent how the roads actually angle.

I actually added split the Carisbrooke Res into 2 parcels, added the Bates parcel and retagged them to be consistent. Many of these open spaces had been tagged as landuse=conservation, and my convention when encountering such areas is boundary=protected_area, landuse=recreation_ground (if applicable), and protect_class=5. If the leisure tag is present I normally would keep it, I must have accidentally overlooked it here. Certainly add it to any parcel you see fit. Admittedly, how to appropriately apply that tag is particularly vague to me, and so I leave it when it is applied and I don't apply it explicitly myself.

I had edited this a long time back and when I revisited this I did not realize someone else had properly configured it. Sounds like you have much more reliable sources than me, so thank you for reverting this and sorry for overstepping.

I had edited this a long time back and when I revisited this I did not realize someone else had properly configured it. Sounds like you have much more reliable sources than me, so thank you for reverting this and sorry for overstepping.

Message received. See my comment on 109425654 and 109429933. Last time I edit in NH, thank you.

Thanks for informing me on the effort to explicitly tag road surfaces. I will keep this in mind next time and revert what I can here.

I thought surface=asphalt was implied, but I have come to find that actually surface=paved is implied. I will revert that change here.