What's the source of the new names -- most of the roads aren't signposted. Ditto the roundabout "centres"?

It's usually just a "code" BSxx yyy on the front paper sheet with collection times -- you can see matched post boxes at https://osm.mathmos.net/postboxes/progress/BS/BS34/#13/51.5201/-2.5648

Has it got a ref?

Is there a road sign for this?

The changeset comment seems a little misleading?

psv=designated / bus=designated is a perfectly valid value why did it need changing?

No worries -- just seemed odd that an office block was marked as a generator.

highway = f ?

What were the routing errors?

Surely only the panel area should be mapped -- not the whole building

"name" should really be what's on the signage.
Moved the names to description.

No worries -- thanks for reviewing

It's already mapped as an area :-)

If you want a discussion start a topic on talk-gb as suggested previously.

If you want less of a blunderbuss approach then make the individual change(s) yourself.

Think this will need to be reverted: there's a clear cycle path -- not a footpath https://binged.it/2LDjuKd

Nowhere near the docks :-)

I've squared up (orthogonalised) some of your edits.

ESRI clarity is the best source for imagery -- unless Maxar is newer

I note one of your recent changes has been completely reverted (not by me!).

It would be better to explain your ideas/motivations on talk-gb mailing list -- then you will get a general range of opinions. E.g. using the duck test, etc.

Well it could have been number, rather than name?

Is osm.org/way/337967806 really called "57"

Maybe you could try adjusting your imagery offset to match what's already mapped.

This area has been carefully mapped -- and you've just drawn a driveway straight through a building!

I think it would have been better to merge the footpath with the start of the service road, rather than having 2 highways on top of each other.