[Tagging] Feature Proposal - RFC - Discouraging the use of deprecated schemes

stevea steveaOSM at softworkers.com
Tue Mar 23 02:16:35 UTC 2021


On Mon, Mar 22, 2021 at 11:49 AM <mail at marcos-martinez.net> wrote:
> I challenge all of you to provide an short answer: In an ideal world, do you at least agree on the desirability to have a database which is consistent?

Yes, but.  Knowing that there ARE "regional differences" (e.g. how highway classifications at a local level fit into OSM's existing scheme), we must allow for these.  Where this must happen, "consistency on a regional level" is achievable.  Of course, if a person, group or "task force" wishes to really think on a particular problem and propose a solution that can harmonize some tagging scheme worldwide, such efforts are both supported through existing process and highly encouraged as "a problem to solve."  Some problems (seemingly) can't be fully solved, but they can be partially harmonized with "regional differences" (with consistency in that region).  Saying "things change on the other side of that border" is simply being realistic.  Again, "world harmony" might never be achieved (if it can be, it should be), so let's strive for regional harmonies first.

> Would you consider it positive if people voluntarily tagged in an aligned manner?

Yes, but.  The flexibility, growth and let's be honest, creativity that OSM allows with "any tag you like" are hallmarks of OSM.  That isn't about to go away any time soon, if ever — nor should it (imo).  Such growth spurts are both positive now and will continue to be positive when they happen.  There will be circumstances where the project must grow into a new space, and as such "alignment tagging" doesn't already exist, tagging must be "coined afresh."  We can't simply turn that off and let such growth die because one person isn't allowed to be bold enough to create a new tag.

Again, tagging is a language and language has usage:  its "collective habits," not its idealized models of how tagging works or should work in the abstract.  As such, usage defines language, but usage changes over time.  Usage is not an easy topic (going back to Daniel Dafoe in the 17th century) and vacillates between being prescriptive and descriptive:  the same snarl we get mired in about how the wiki does or should operate w.r.t. to tagging.  Like dictionaries, our wiki is much more descriptive than prescriptive (unless it expressly says otherwise, and it does in some isolated cases).

Such "alignment" doesn't happen by fiat.  Well, not very often, anyway.  It happens as usage gets noticed, picked up by others, used more widely and emerges as the accepted standard.  This is always happening in all living languages, tagging in OSM is no exception.  What we see here and now writ large is that people want to seize this power "by fiat" (whether by wiki writing, crafting a proposal, being assertive on a mail-list...) as opposed to "by usage."  People posting taginfo data are an example of demonstrating how usage speaks for itself.  "Active deprecation," while it has existed in OSM, seems an artificial process, as what is old and becomes defunct becomes defunct simply through lack of usage.  That's a natural process and OSM should pay attention to how organic and real it is, rather than replace it with something that not only is more forceful, often dividing and contentious, but artificial and disharmonious with the order of how things will simply fall out of favor anyway.

How do we manage and document all of this?  Well, acknowledging that it is an active process is a good place to start.

Nobody ever said that a worldwide map made by millions of us was going to be easy.


More information about the Tagging mailing list