OpenStreetMap logo OpenStreetMap

Changeset When Comment
57034975 almost 8 years ago

No -- do you think it should have been?

I'm not sure exactly what your concerns are, but (a) the use of the DfE data for UK Schools is well-established, and (b) I wouldn't regard the changeset as a mechanical edit as each change was examined individually before being included in the changeset.

The changes were made based on official data from https://get-information-schools.service.gov.uk/ that said a school with a now-definct ID had a successor institution with the new ID. I used the data extracted by my tool in the table at
http://robert.mathmos.net/osm/schools/matching-queries.html#no-ref-successor
to manually check that the school still had essentially the same name, and was in roughly the correct location. For such objects, I concluded the ID change was just administrative, and so could be updated safely. Given the nature of the changes (updating the ID and in some cases changing the name to the official name supplied by the school to the DfE) I felt this was sufficient verification for each change. (The currently remaining entries in the table above are the ones I rejected in my initial pass as needing a fuller investigation before making any changes. I haven't got round to looking into those yet.)

55573374 almost 8 years ago

Well spotted! Yes, it should have been thanks. I've now fixed it. (It was a 'repeat tags' error using potlatch. I noticed it on another way, and thought I only had one to fix. I must have missed that one.)

11709297 almost 8 years ago

Given the position of node/1766526044 I wonder if it is supposed to be a post box (rather than a post office).

If it is a Post Office, then should it be located on the other side of one of the roads?

54506868 almost 8 years ago

Three ways way/545809287 way/545809288 and way/545809289 created in this changeset have been tegged as designation=public_footpath .

But they don't appear to be recorded as such in the official council data: http://robert.mathmos.net/osm/prow/progress/norfolk/south-norfolk/ditchingham/ . Are you sure this route is actually a Public Footpath, rather than e.g. just being a footway with the highway boundary of the road?

48159001 almost 8 years ago

Oops -- I didn't notice that. To be fair though the place=town, name=Wallasey had already been destroyed by the previous edit. Anyway, I've now restored the node in question to its prior state: node/26703036 Thanks for spotting this.

54132168 about 8 years ago

I think you've mis-pasted the addr:postcode value into way/543125303 ...

53597248 about 8 years ago

Hi, I've just noticed that on the Post Box object at node/5217230757 you tagged it with collection_times=Royal Mail as well as operator=Royal Mail. Did you mean to put something different for the collection_times?

53613994 about 8 years ago

Hi! Could I ask where you got the addr:portcode value of "PO180J1" for way/539245375 from? The final three characters of a postcode should be a number followed by two letters. But the 1 you have can't be a typo for an I, as I's aren't allowed in the final two places either.

48666498 about 8 years ago

Yes Rjw62 on the wiki is me. I wouldn't take the wiki as gospel -- it's as much to document current practice as it is to set out guidelines. In particular, there appears to be little information there about how or why highway=no is marked as deprecated. From what is on the wiki, I would suspect that its presence in the list is really to deprecate the (IMO clearly incorrect) use of highway=no to indicated a physical highway over which access is forbidden -- which is (IMO) correctly tagged instead as highway=* + access=no. But that's not what I've used it for.

I think historic:highway would be useful to record what type of highway used to be there in the past, but I don't see it as a replacement for a mechanism to record that there is no highway there now. For instance, you might want to use historic:highway=primary + highway=service for the typical oxbow laybys that you get next to main roads that have been straightened. So the presence of a historic:highway=* tag and the absense of a highway=* tag would just be an indication that the current highway type has yet to be determined.

52950822 about 8 years ago

Are you sure the postcode is correct on way/532684269 and way/532685134 ? According to Code-Point Open there are no postcodes starting "TD11 5".

52831398 about 8 years ago

Are you sure the postcodes on node/5161224522 and node/5161224521 are correct? According to Code-Point Open, EH4 8AJ is located on Strathalmond Park about over 5km away.

52807943 about 8 years ago

I think you made a mistake editing way/520664349 in this changeset, as you've left it with addr:postcode=Eastbourne. Unfortunaetly, the neighbouring houses on each side have different postcodes, so you can't tell easliy which one this house should have.

52592020 about 8 years ago

Typo fixed thanks. That one is local to me, but I haven't tested it yet.

52620509 about 8 years ago

Yes, just a typo. Now fixed, thanks.

52140467 about 8 years ago

Where did you get the postcode for node/2444679163 from and are you sure it's correct?

According to Code-Point Open, "E17 7JH" is about 7km away in Walthamstow. This location is in the EN3 postal district.

48666498 over 8 years ago

There are actually quite a few other ways tagged with highway=no in the UK: http://taginfo.openstreetmap.org.uk/tags/highway=no

If the route available on the ground is significantly different (to the point where it would be clearly incorrect to gat either one as the other), then I think both should be recorded. If there's a route that's the definitive line, but there's no physical highway and the route is not used as a highway, then arguably you cannot tag it as highway=path or any of the other highway values.

Then the question is how to tag it. I would argue that we need a way to distinguish between a way that may or may not be a highway and hasn't yet been checked, and a way that has been checked and definitely isn't a (physical or used) highway. I think highway=no fits the bill perfectly here.

50698997 over 8 years ago

Some more suspect postcodes in this changeset I think (presumably the same software bug you mentioned before). According to Code-Point Open "NG16 2WX" is right for the Awsworth properties, but not for the ones in Swingate on Babbington Lane. But both sets have been marked: https://overpass-turbo.eu/?w=(https://wiki.openstreetmap.org/wiki/Tag:%22addr:postcode%22=%22NG16%202WX%22)+in+UK&R

Code-Point Open suggests that the houses along Babbington Lane, should probably be "NG16 2PT", but I wouldn't want to add that to them all without some verification or local knowledge.

52029393 over 8 years ago

Yes, good spot, thanks. Now fixed.

51346382 over 8 years ago

I wondered that too, but NG16 2JP is in Kimberley, not Ashworth. So it's presumably not just a missing 6.

51620647 over 8 years ago

Are you sure about the postcode "BH21 2RP" that you've given to this row of Ashbury Cottages? According to Code-Point Open, this postcode is located about 5km away.

(Discrepancy found using http://robert.mathmos.net/osm/postcodes/location-errors.cgi )