I think statuslink should be stated mandatory for proposed status as well. Any feature can't actually get this status if a related proposal isn't available, can you? Fanfouer (talk) 17:14, 4 August 2019 (UTC)
- Some proposals were made directly in the standard wiki pages rather than using the proposal format, so these may still be a proposal but do not have a separate proposal page. Also, the text currently states that statuslink "should be included for all tags and keys which had a proposal", so I think it's clear enough? --Jeisenbe (talk) 12:17, 5 August 2019 (UTC)
- Sometimes people document in use tags that were used just 5 times. Which status - if not "proposed" - would fit? Mateusz Konieczny (talk) 12:07, 15 January 2021 (UTC)
New status: "import"?
I would like to add a new status: "import" for tags that were used in a large import, but are no longer commonly being added. These tags often have high numbers in the database but have not been accepted by the community and rarely are being used by mappers (in some cases they have never been added, except by the import), so they should not have the status "in use", and they were never proposed in most cases.
Using "import" as the status would make it clear that these tags are not "de facto", "in use" or "approved", and would be more specific and clear than "unspecified". --Jeisenbe (talk) 06:42, 17 March 2020 (UTC)
- I like the idea, maybe there could even be room for 2 new values: tags introduced by an import and not or rarely adopted for manual mapping, and tags that had already been in use (maybe scarcely) but significant (most?) usage stems from imports. Or maybe a flag that states: significant amount of usage is from imports. —Dieterdreist (talk) 08:54, 17 March 2020 (UTC)
- I was not intending this status to be a "mark of shame", but rather informative. Certainly this information should also be mentioned in the main text of each Tag: or Key: wiki page, but it's nice to have a quick and simple way to show how a tag got into the database. Some imported tags like gnis:feature_id=* are useful to keep the Openstreetmap database object directly linked to an object in an external database. That's why I am not suggesting the use of "deprecated" or "obsolete". The main reason to mark them is so that mappers and database users will understand where the tag came from.
I was also thinking about tags like "object:postcode=*" - this tag is only used in Germany on highway=street_lamp features which appear to have been imported mostly in 2015: https://taghistory.raifer.tech/#***/object:postcode/ . Though the usage numbers are moderately high, it is helpful to know that this tags are not often being used, except for the imported features.the tag was added by a mechanical edit, not imported: see below
- Neither did I intend this as mark of shame. I believe it is useful information for mappers to know whether a tag is used in high numbers because everybody uses it, or whether it is because a fee people introduced it through an import, and eventually the numbers are misleading and the tag is neither accepted nor used by mappers. —Dieterdreist (talk) 08:45, 18 March 2020 (UTC)
Import status vs. external id
If mappers manually add a particular tag by hand, such as with openbenches:id=*, I feel the tag shouldn't be labelled with the "import" status. This makes the classification less useful by mixing very different situations (manually added links to an external database, automatically imported links to an external database, other automatically imported tags). The fact that it's an external ID can, at least in cases like this, already be understood by the presence of an URL pattern parameter. --Tordanik 17:18, 25 March 2020 (UTC)
- Re: openbenches:id - my understanding is that you find this by searching in the openbenches dataset, right? "Use the ID number (from the URL slug) on the bench's page at https://openbenches.org as an OSM-Tag." So the value of this key is always imported from the database shown at opensbenches.org
- Most imports require some manual input, and many are done entirely "by hand". But if the information is only available by referencing an external database (as is almost always the case with an External Reference or External ID), then it is an imported tag: it cannot be added or verified by surveying the feature. While it's true that the URL pattern of the value would suggest that it was imported from an external data set, some URLs are actually listed on signs at a feature: the value of a key like website=* would not always be imported, but could be confirmed by visiting a shop or business in person and seeing the website on the window or door sign or business cards.
- As mentioned on the mailing list, this status is not meant to be a "mark of shame", but rather informative, so that database users will know that this information is linked with or available in an external data set, and so that mappers will know that these tags can be added by referencing an external data set, not by survey. --Jeisenbe (talk) 00:58, 26 March 2020 (UTC)
- Manually browsing an external database to look for an ID is the closest equivalent of a survey for this kind of tag. And it's a substantially different process from importing external IDs. After all, "import" has a relatively established meaning in OSM that does not include adding a link to another site. For example, adding Wikipedia links does not require you to follow the rules for imports (which it would if doing so was considered an import).
- I understand that it's not meant as a "mark of shame", but I disagree that using the status generously makes the wiki more informative. Surely, tags referencing external databases can be proposed, approved, in use etc.? Where do we store this information now? --Tordanik 17:23, 31 March 2020 (UTC)
- Re: tags referencing external databases can be proposed, approved... - Do you have any examples of these? I agree that Wikipedia/Wikidata links are probably accepted by community consensus. But are there any other external reference ids which are even "proposed" status? Please provide an example. If they exist, I would be happy to mark the status as "draft/proposed/rejected/approved" as appropriate. The status "import" is intended for tags that did not go through the proposal process. --Jeisenbe (talk) 11:18, 1 April 2020 (UTC)
unspecified, undefined, unknown - is it ever valid?
- You don't provide such value manually - it's only when nobody has given any other value. maro21 17:12, 15 January 2021 (UTC)
- Yes, but why you would use it instead of leaving status unfilled? Mateusz Konieczny (talk) 17:34, 15 January 2021 (UTC)
Translation into other languages
the difference between "de facto" and "in use"
- What is the difference between "de facto" and "in use"?
- How many uses must a tag have in order to be considered de facto? What is the minimum?
- Is an OSM Carto icon enough to change status to de facto?
- What is "widespread use"?
- Should the "de facto" tag be used all over the world? What if a tag is used one million times but only in one country?
- If a tag has an OSM Carto rendering, preset in iD and JOSM, can it be marked as "de facto"? maro21 22:00, 21 March 2021 (UTC)
- "If a tag has an OSM Carto rendering, preset in iD and JOSM, can it be marked as "de facto"" - definitely.
- "Should the "de facto" tag be used all over the world? What if a tag is used one million times but only in one country?" - depends a bit on tag. Is it something present only in that country? Is it mapped elsewhere?
- "Is an OSM Carto icon enough to change status to de facto?" - not always, right now shop=dhhdjasjhdasuhudusdahusdahuas tag would be still displayed with a generic dot icon.
- "How many uses must a tag have in order to be considered de facto? What is the minimum?" - sadly it depends. Compare threshold that would be required for capital=yes and building=house/house=detached. And it depends on other factors. 1 million manually added objects vs 1 million imported objects, competing tagging schemes
- So overall answer is sadly "it depends"
- Mateusz Konieczny (talk) 08:53, 19 April 2021 (UTC)