This reverted changeset/189793098 and changeset/189848112

These are the current explicit default values (plus time-based conditions):
motorcar=permit
motorhome=permit
bicycle=yes
motorcycle=yes
goods=yes
hgv=yes
moped=yes
emergency=yes
bus=yes

I added a brief note explaining.
Routers not being able to deal with it is not our problem. Some information on router support for conditional tags is available here: osm.wiki/Conditional_restrictions#Software_support

Reverted here: changeset/189865647

Reverted here: changeset/189865647

There is nothing in motorcar=permit, motorhome=permit, affecting bicycles, vans, buses and lorries. If that is the problem you have encountered, we can add explicit tags for those modes (*=yes).
If a van driver uses a router set for motorcars, that is not our problem. Our aim is to make the underlying data as accurate as possible and as useful for the general user.

Your edit is not the solution. Further comments here: changeset/189793098

I will revert your changes, and add more specific tags to help with other modes of transport.

* Sorry for typo in the first comment. Read: > which does NOT seem correct to me.

The two cases of Marston Ferry Rd and Hollow Way are more dubious. The traffic filter is active for a shorter time, so it is arguably not the norm, but a special condition. However, I thought that setting permit as the default was better at avoiding people fines than the opposite. I'm fine with a change, if people think otherwise.

For the other regular traffic filters, however, I am strongly in favour of the previous tagging. It is more consistent with reality and more useful for correct routing avoiding the filters.

Hi. Can you explain this change, and point to those previous discussions, please?

The daytime status should be the default access value. I think that is clear from all documentation.
It is exactly right that a naive router (one that doesn't account for hours or permits) avoids the traffic filter. That is exactly their point.

With your change, you are defaulting to the nighttime value of access=yes, which does seem correct to me.

Oops, thanks. However, I changed it again -- I think it relates to the direction of travel, so left in both cases, right?

No worries, thanks again, and welcome to OSM.

I have restored the previous names, but added alt_name=Beanwood, so that both are searchable.
Thanks for your contribution, and welcome to OSM!

Hi. Thanks for this edit! I think you have broken Bus route relations.

Hi. What's your source for this? Both OS and Wytham Woods maps use Bean Wood, and Cowleaze as a subpart. You have removed information which I think was correct.
Beanwood, if also used, can be added as an alt_name.

Why would it be an error? The two tags are semantically different and serve different purposes. There can't be an automatic exclusion. As you said, you don't remove every case, just because the value is the same. At most, it can be redundant, but removal can result in loss of detail and information, for little gain.

- For houses, take the case of a historic building, listed status, wiki pages, etc.. That can deserve its own name tag. Whether it has an address or not, and whether the name is part of it or not, is irrelevant.
- For commercial buildings, and those inside hospital grounds, schools, universities, etc. It is likely that name is the correct tag, not addr:housename, if they cannot be addressed individually. However, even it they can, it doesn't mean that the building name should be removed.
The more likely case is that automated tools encourage users to fill addresses for all buildings, even when not appropriate, so less experienced users end up putting the name into addr:housename.
- For flats, it is also complicated, since the name might or might not be part of the address. Depending on your interpretation of the address tags (cf. osm.wiki/Addresses_in_the_United_Kingdom), it might not even go into the housename tag, but the substreet one.

If you are confident with your edits and individual checks, I trust you -- I will change back all of those in Oxford that I think are mistaken, and others can do the same elsewhere.

Thanks.

Hi.

I don't think this is the correct approach. It looks like an automated edit, but every single case should be checked individually.

There is nothing preventing the name of the building (name)=* from also being used in the postal address (add:housename)=*, and the address tag does not automatically render name=* a duplicate.

It is debatable in the case of single residential houses. Some names are indeed purely residential addresses; others, however, could be name tags (e.g. if the building is historical).
The change is certainly wrong in other cases of buildings -- for example most of these in Oxford that I know.

I think the changeset needs to be reverted, and every case judged with local knowledge.

Thanks.

Hi. Some of these edits to pharmacies, changed to amenity=hospital, don't seem correct.

I have reverted these changes (changeset/184903167).

Please do add additional information with sidewalk tags where missing, if you want.

Best wishes.

There is also a verge=* tag, which can be helpful in some cases.

I think that a highway=trunk with sidewalk=no already implies that there is no safe space for pedestrian walking.
Other tags like sidewalk:left=separate or sidewalk:right=separate mean that a pavement is present, and mapped separately.

If the sidewalk=* tag is missing on some segments, that can be added, to provide the detail you are looking for, and to allow routers to better avoid these roads.

The access tag changes will have to be reverted though.

Best wishes.

Hi. Welcome and thanks for your edit.

I think the established use of access tags like "foot=*" is to represent _legal_ access, not practicality or safety. If pedestrian have the legal right to walk on these trunk roads (as I believe is the case, but correct me if I'm wrong), even if not advisable, then your "foot=no" is wrong.
It is up to any router to de-prioritise use of trunk roads and direct users on other paths.

You have also added an "access=no", which applies to all forms of access.

Actually, probably on the path on the inner side of the pond. way/1356162891

Is the current tagging correct? Sending bikes on the outer ring on what looks more like an informal grassy path seems strange.