"Cable car (closed)" seems a strange name for a fully operational cable car. It should not even be used for a non-operational one.

This should probably be tagged using "disused" lifecycle tags, and should probably not have a name.

The cooling plant is not going to be a social facilty; that is used for places like old people's homes! There may well be a need for a new main tag for this sort of feature.

Also, the name looks like it may be a description, rather than the name. Even if it is a name, I wouldn't expect infrastructure to be named only in English in Dubai.

Even if "Cooling Station" appears in the name, this is significant enough that it really requires its own tags.

Although I can't identify exactly what has been changed back here, I'm concerned that the reversion is mapping for the renderer. "It doesn't render right" is not generally a valid reason to change something, and I would not expect to have a cycle path mapped in two places.

This change breaks a basic rule for bridges that the layers must always be greater for the bridge than what it crosses.

There is another basic rule that one does not map for the renderer. If this renders incorrectly, raise a bug report on the renderer.

In my view, the original was correct and the change is wrong.

The other issue here is that names should not be used to carry meta-data.

The entrance to the St George's centre should have been mapped on the building outline (correcting that as necessary).

This is not a 90 degree bend. It is at least a 450 degree spiral, and I think it may even be an 820 or more one. The adjacent service road is actually part of this ramp.

I would still say that it should not be called Vacant Shop; it should be un-named.

I don't believe there is a hairdresser's here called Vacant Shop; I think I would have noticed the name if there was. This looks like misusing the name as meta-data.

There are at least two ways of indicating void shops; one is shop=vacant, but there is also a lifecycle prefix way of doing it.

I thought there was a general principle that we did not map information about private individuals. If nothing else, it creates Data Protection Act responsibilities - full redaction may be needed if there is a a complaint.

Also mapping internet_access=wlan seems to me to imply at least customer level, if not public access; I presume this is really private access to the household. Whilst this might be a BT one, with BT customer access on the side, mapping those will be impossible to maintain.

This changeset is being discussed on http://forum.openstreetmap.org/viewtopic.php?id=55463

You tagged it as one way in the wrong direction. Unfortunately, although now corrected, that may have lost OSM an important user, see the discussion on osm.org/changeset/37144552

Using an unacceptable source won't delay when the change goes live, which is immediately, for the database, within minutes for the standard rendering, and an unknown amount of time for third party renderings and third party navigation applications.

However, using a source that breaches copyright puts you at risk of having the change completely removed. Google maps is a competing product. It is a specific breach of the terms of use of both Google Maps and Google Street View to use them as the basis of other maps.

In choosing to use OpenStreetmap, you have to be aware that it is reliant on voluntary effort so the quality of mapping will vary greatly from place to place. You also need to understand that it can only use information that has very liberal copyright licences, so the fact that another source obviously contradicts only gives a hint that someone needs to actually go there and check for themselves.

On the other hand, in the western world the level of detail and accuracy tends to be surpass that in any of the more commercial maps, when taken as a whole.

It actually looks like it was only marked as one way five days ago, so whatever you are using updated within five days last time. Whilst it might update on a cycle, and you might have been lucky that it was near the end of the cycle, it at least suggest that the change will reach the tool you are using within a week.

Note, that I was going to comment about the reference to Google, even before I saw the changeset discussion.

In this case, in http://forum.openstreetmap.org/viewtopic.php?id=53632 you say that you got the information from your employee. That would have been sufficient to claim local knowledge as the source.

You should not put meta data, like "(Under Construction)" in the name field, see osm.wiki/Names#Name_is_the_name_only

You should use landuse=construction then construction=<eventual main tag>. You should not include tags, which, in the absence of the construction tag would result in a valid main tag, as renderers may not act on the construction tags

In this case, the next level should probably be construction=aerodrome, although I think you can also do construction=aeroway; construction:aeroway=aerodrome.

"New" in the name should also only be used if the construction site has a name and it appears in that name.

This comment was the result of finding osm.org/node/1042054418/history.

According to osm.wiki/WikiProject_Pakistan I think this should actually be highway=unclassified

Unfortunately, that page doesn't have any guidelines on language, but I would then argue that the default policy applies, which is that you map in the language that actually appears on signs.

Where you get a lot of mapping in English, in Asia, it is generally either because the mapping has been done by non-native mappers, or because it has been done by people in an intellectual elite that is familiar with English. OSM should be for all the people.

A road in this location is unlikely to be primary. One would only expect this to be highway=tertiary osm.wiki/Tag:highway%3Dtertiary although it might even be highway=unclassified.

Is the name really as given, and in English, or is this just a description? If it is a description, destination=* may be more appropriate osm.wiki/Key:destination If this is a name, but it is signed in Urdu, it should be mapped in Urdu.

You need a source tag for the name, as that cannot be derived from the Bing imagery.

The changeset comment doesn't match the actual change, which is adding the road. Also, if it really were part of one change, it is better to try and do all the changes in one changeset.

You haven't removed the access=not specified yet. This tag should be completely removed. The "not specified" will be an artefact of the way that iD tabulates possible access tags. It is not a value that should actually appear in the map.

In my view, "correcting access" is not a sufficient change set comment. Correcting it to what, why and on what?

Also, the only effective source you are quoting is Bing, but Bing imagery cannot tell you what the access is. If your editor can't set a source tag, include that information in the comment as well.

Is the official name of the cemetery really "Cemetery". That seems unlikely. If there is no verifiable name, it should not have a name at all.

Is the official name for this area of land really "Strawberry and Vegetable Farms"? It sounds like a description and names should never be used as a description. You should use proper meta-data for that. Remember that the map is the data; it is not the standard rendering, and you should never map to obtain a particular rendering.

See osm.wiki/Names#Name_is_the_name_only

I think the name in version 13 is more plausible, but still the rule is to follow what is signposted on the ground. If this is an anonymous collection of farms owned by the University, it should have no name, but could have an operator tag.

Even if there is a sign saying "to the strawberry and vegetable farms", you should decide whether that is just a description, rather than a name.