[Tagging] How to overcome lack of consensus

Matthijs Melissen info at matthijsmelissen.nl
Wed Sep 18 13:33:32 UTC 2013


Dear all,

Thanks for all the answers. It seems that there is consensus about some
points:

- To know how to tag, it is necessary to look at both actual use, taginfo,
and the wiki.
- It is important to document the way tags are used on the wiki.
- The wiki should be adapted to the outcome of votings and to actual use
(not the other way around).
- If actual use and the outcome of votings agree, but the wiki does not
reflect this, then the wiki should be adapted.

There is no consensus on what to do if votings and actual use disagree. In
particular, there is no clear answer to the following questions:
- If the outcome of votings and actual use disagree, what should mappers
follow?
- If the outcome of votings and actual use disagree, what should the wiki
reflect?
- How should new tagging schemes be introduced?

If the mappers (and the wiki) should always follow the current use, then it
is difficult to implement new tagging schemes.
On the other hand, if mappers (and the wiki) should always follow the
outcome of votings, this might cause problems with people who don't agree
with the outcome of the voting (including people who did not vote at all),
and with existing tagging.

The solution for now seems that if actual use and the outcome of voting
disagree, then both should be included on the wiki. In other words:
- One tagging scheme, two meanings: mention which meaning has been voted
for, and mention the meaning that is actually used as well.
- Two tagging schemes, one meaning: document both the tagging scheme that
has been voted for, and the tagging scheme that is actually used, and let
the pages explaining both tagging schemes link to each other.

There are still different ways how this can be implemented, which reflect
different viewpoints on the taginfor/voting debate. Consider the following
examples based on a hypothetical 'two tagging schemes, one meaning'
controversy between pet=wildebeest and pet=gnu. Assume that pet=gnu is used
more often, but pet=wildebeest has been selected by voting. Which format
would be preferred?

1.
- (No page for pet=wildebeest)
- pet=gnu is used for pet shops focused on gnus.

2.
- pet=wildebeest is discouraged. Use pet=gnu instead, as it is more
commonly used.
- pet=gnu is used for pet shops focused on gnus. Although voting has
selected pet=wildebeest, pet=gnu is used more often.

3.
- pet=wildebeest is used for pet shops focused on selling wildebeests.
Related tags: Consider using pet=gnu instead, which is used more often.
- pet=gnu is used for pet shops focused on selling gnus. Related tags:
Voting has selected pet=wildebeest, but the tag pet=gnu is used more often.

4.
- pet=wildebeest is used for pet shops focused on selling wildebeests.
Related tags: Voting has selected pet=wildebeest, but the tag pet=gnu is
used more often.
- pet=gnu is used for pet shops focused on selling gnus. Related tags:
Voting has selected pet=wildebeest, but the tag pet=gnu is used more often.

5.
- pet=wildebeest is used for pet shops focused on selling wildebeests.
Related tags: The tag pet=gnu is used more often, but voting has selected
pet=wildebeest.
- pet=gnu is used for pet shops focused on selling gnus. Related tags:
Consider using pet=wildebeest instead, which has been selected by voting.

6.
- pet=wildebeest is used for pet shops focused on selling wildebeests.
Related tags: pet=gnu, which is discouraged.
- pet=gnu is discouraged. Use pet=wildebeest instead, as this tag is more
often used.

7.
- pet=wildebeest is used for pet shops focused on selling wildebeests.
- (No page for pet=gnu)

Or or course any other option...

This is a relatively simple example where wildebeests and gnus are exactly
the same animal. Things will of course become more complicated in case of
subtle meaning differences. However, I suppose we should start with easy
examples.

-- Matthijs


On 16 September 2013 16:41, Matthijs Melissen <info at matthijsmelissen.nl>wrote:

> Dear all,
>
> There are some OpenStreetMap features for which there is no consensus on
> how to tag them. It does not seem like consensus will arise soon. The lack
> of consensus does cause problems for the Openstreetmap community, though.
> Therefore, it would be good to have ideas or procedures on how to create
> consensus.
>
> There are currently quite a lot of OpenStreetMap features for which there
> is no consensus on how to tag them. Some examples (but I'm sure there are
> many more):
> - What is the difference between highway=footway and highway=path?
> - What is the right scheme for tagging public transport?
> - Is an unsurfaced residential road a track?
> - Should we use shop=betting or shop=bookmaker?
> - Should we use shop=fishmonger or shop=seafood?
> - Should we use office=estate_agent or shop=estate_agent?
> - Should we use shop=tailor or craft=tailor?
>
> The lack of consensus becomes clear by the fact that there are
> discrepancies between documentation on the wiki, the outcome of a voting,
> actual use (as documented on Taginfo, for example), and what editors and
> renderers support.
>
> The lack of consensus creates several problems. These problems include the
> following.
> - Multiple parallel tagging schemes and unclear documentation creates
> confusion for newcomers.
> - Users are often advised not to follow the documentation on the wiki, and
> to look at Taginfo instead. This makes the wiki useless. It also leads to
> the fact that hardly anybody bothers to edit the wiki anymore.
> - If mappers should follow current use of tags, then it becomes very hard
> to introduce new tagging schemes, such as the office= and craft= keys.
> - Multiple tagging schemes creates problems for the implementation of data
> users (renderers, routers).
>
> It does not look like consensus on these issues will form naturally, as
> there many of them exist for multiple years, and I hardly see any
> initiative to do something about them. Most users seem to accept that there
> are controversies, instead of trying to come to a solution. I think there
> is nothing wrong with a temporary period in which two tagging schemes are
> used in parallel, but permanent lack of consensus creates problems for the
> community.
>
> Rather than trying to solve the individual cases where there is a lack of
> consensus, I think we should first try to agree on a procedure on how to
> solve these issues. For example, we might try to answer the following
> questions.
> - What should be the process to create consensus?
> - Should we recommend (new) users to follow the wiki, actual usage
> (Taginfo), or the votings?
> - Should we recommend data users and editors to follow the wiki, actual
> usage (Taginfo), or the votings?
> - Should the wiki be adapted to actual usage (Taginfo) and/or to votings?
> - Should we allow automatic edits in simple cases (such as bookmaker
> versus betting), or leave the process of standardisation to local
> communities?
>
> I would like to thank anyone responding in advance for his/her feedback.
>
> Best regards,
> Matthijs Melissen
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/tagging/attachments/20130918/7e6c4fa3/attachment.html>


More information about the Tagging mailing list