[Tagging] Rules (was: Feature proposal - Approved - deprecate embassy=embassy)

Fri Feb 4 15:03:10 UTC 2022

I value the discussions on this mailing list, and the resulting outcome.
The voting system provides timing and focus, and mainly indicates that an
international discussion has been held. Mappers, data users and tool
builders can choose to trust this process, or read the arguments, or ignore
the lot.

As a rule, I tend to follow the OSM mapper's community in my country,
rather than blindly applying approvals. This goes for new tagging, adapted
tagging, adapted interpretation of tagging, and deprecations.

I also tend to ignore most tagging warnings in JOSM and other editors, and
I seldom use presets or auto-corrections.

As for the wiki documentation, that's a total mess^H^Hlting pot of
languages, conflicting information, old and new, gems and rubbish,
description and prescription, and competing schemes. Again, the local
community gives me much better answers and examples.

Which makes this discussion, about how compelling the rules are, rather
pointless. Doesn't matter which dog wins, because the bones are eaten
elsewhere.
Still, a 50-0 approval after an international discussion does weigh in.
Even more so, if it comes with an implementation plan, including prior
commitment by data users and tool builders. It will probably, though not
always, gain support from the local community.

Fr Gr Peter Elderson

Op do 3 feb. 2022 om 14:14 schreef Andy Townsend <ajt1047 at gmail.com>

>
> On 03/02/2022 11:22, mail at marcos-martinez.net wrote:
>
> Where does it say people can't vote for a deprecation and that their vote
> is useless and without consequence? There have been 33 people who are in
> favor. I see you and Simon are opposing. Are you both the representatives
> of the community above others? If not and you are referring to some noble
> and basic OSM rule that needs to be followed, I kindly request you to
> direct me to a place where the rule is stated and who approved it/on which
> consensus it was elaborated?
>
> Firstly, it's perhaps worth mentioning that its silly having an argument
> over amenity=embassy because there's only one of those left
> https://overpass-turbo.eu/s/1fKw , and that appears to be mistagged.
>
> However, the general point is important, and it's important to separate
> OSMF from OSM here.  OSMF's mission statement is here:
>
> https://wiki.osmfoundation.org/wiki/Mission_Statement
>
> It has as its first sentence "The OpenStreetMap Foundation is an
> international, not-for-profit, democratic organisation with the tasks of *supporting
> the OSM project*, running and protecting the OSM database, and making it
> available to all. " (my emphasis).
>
> The OSMF is an organisation that you can join and has a democratically
> elected board that you can stand for.  It has a number of appointed working
> groups who report to the board.  If you believe that the OSMF's approach
> should be more "hands on" than the current "supporting" role then please
> stand for the board and make that case.
>
> There are a bunch of other policies -
> https://wiki.osmfoundation.org/wiki/Policies_and_other_Documents - but
> you'll notice that there's no "tagging policy" among those.
>
> How people map things and which tags they use is handled by a general
> community consensus and a bit of common sense on all sides.  Over the years
> an aversion to deprecation has grown up due to the surprises that it can
> cause for people who work with OSM data - not just editor developers, but
> data consumers of all stripes.  There are times where deprecation makes
> sense (e.g. "highway=ford" on ways doesn't let you say what sort of highway
> it is, beyond being a ford) but many times when it does not.  If a new form
> of tagging makes sense mappers (guided by the editors that they use) will
> adopt it and as it appears in the OSM database, data consumers will support
> it.
>
> You say that 33 people voted in favour of this "deprecation" (and there
> are around 660 people who will see this message directly via this mailing
> list), but those are tiny "even less than a rounding error" numbers
> compared to
> https://wiki.openstreetmap.org/wiki/Stats#Contributor_statistics_reports .
>
> It's important that we discuss how we tag things as a project, and the
> tagging mailing list and the wiki both form part of that.  What we don't
> currently have (at API, board or working group level) is a way of enforcing
> the views of 33 people over the 300k (active contributors last year).  What
> you appear to be wanting is a way of not only telling lots of other people
> that they're "doing it wrong" but actively enforcing your views over them.
> Historically OSMF members have tended to vote for board members who were in
> favour of a "supporting" role rather than a "telling people how to map" one
> - if you read through the candidate manifestos from last and previous years
> you'll see a number of people suggesting a more "hands on" role for the
> foundation, but OSMF members have tended not to vote for them.
>
> It's perhaps also worth remembering that OSM, with its hands-off "any tags
> you like" approach has succeeded where other initiatives from around the
> same time (e.g. Wikimapia) have been less successful.  Elsewhere, Wikipedia
> succeeded where Nupedia failed; in the world of computing the "Expert
> Systems" approach (all the rage when I started working on systems doing
> what would now be called "Machine Learning" at the back end of the 1980s)
> has similarly been eclipsed.  If you want to change one of the core
> principles of OSM (consensus rather top-down imposed tagging) you're going
> to need to make some really good arguments to persuade people.  So far
> you've tried this (at least) on Telegram and tried this here, but have
> received significant pushback in both places.
>
> It's been said before, but bears repeating, that OpenStreetMap is _not_ a
> computer science project.  Someone emerging from college will surely look
> at the OSM tagging that has evolved over the years and suggest how it ought
> to have been designed back in the 2000s.  However, unless they have access
> to a time machine they don't have to opportunity to go back and make that
> change.  Any changes now need to consider the "installed base" of data,
> mappers, data consumers etc.  Many or most of these people are volunteers.
> As a data consumer, exactly why should I spend my own free time playing
> catch up with the latest wiki vote until a new way of tagging is broadly
> supported by OSM mappers?
>
> Best Regards,
>
> Andy (a member of the Data Working Group, but written in a personal
> capacity)
>
> PS: I've resisted replying until now, as I suspect that many other people
> have, because unfortunately https://en.wikipedia.org/wiki/Brandolini's_law
> applies - it's possible for people to make the same arguments against the
> status quo again and again in slightly different ways without actually
> addressing reasons advanced why something might be a bad idea.  To answer
> _every one_ of these would be a waste of time that could be better spent
> doing something useful, like mapping.  Some people, unfortunately, seem to
> be "only here for the argument".  As the saying goes "Never wrestle with a
> pig. you both get dirty and the pig likes it".
>
>
>
>
> _______________________________________________
> Tagging mailing list
> Tagging at openstreetmap.org
> https://lists.openstreetmap.org/listinfo/tagging
>
-- 
Vr gr Peter Elderson
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/tagging/attachments/20220204/89036761/attachment-0001.htm>