[Talk-ca] What do I poutine the name tag of a road with a suffix?

[stevea]

An enlightening thread, indeed.

I think it would be freaking amazing if TTS (or other kinds of software-driven, "AI"(ish)-driven... use cases for OSM data) EVER became as accurate as "pronouncing" (rendering, routing, speaking...) the data in OSM as humans are at pronouncing [the same, similar] data.  Such as on a sign, when they see it, according to local convention, and determining that it IS an abbreviation, and if so HOW to (render, route, speak) that abbreviation.  But I don't think we'll ever arrive:  like Zeno's paradox [1], (Achilles and the Tortoise, I think) we'll get closer and closer, without ever truly achieving arrival / perfection.  Something like that.

I like the tenet that OSM Contributors put into the map what "is" (locally, and we seem to globally or almost globally say "eschew abbreviations"), rather than abbreviate.  If this is too long, or too abbreviated, or "inconvenient" for the "parser" (renderer, router, TTS...) let the (renderer, router, TTS...) deal with it.  If that means special casing, or "in Alberta, follow these special rules..." well, that's up to the (renderer, router, TTS...) to deal with.  So, deal with it.

This allows (continues to allow) mappers, as data entry volunteers, to follow simple rules, like "don't abbreviate" or "if the sign says Blvd. enter Boulevard."  Does Alberta have something odd going on?  Well, it can either be "untangled," or if it can't/won't, let the special-casing (in the parser) happen.

These things will "chase each other's tails" (up and down OSM's "data stack") and incrementally, things get better in the long-run.  In the short- and medium-run (timeframes), yes, it's likely there will be "mispronunciations" and "odd renderings" and "hey, you shouldn't have routed me along there."

To some this may be obvious.  To some, maybe you are reading about this for the first time.  It's a little chewy and a little thorny (at times, for some), but OSM has been on the right track for a long time.  Data, in an ideal world, "shouldn't" have to "catch up," rather, software should.  But sometimes, data DO have to catch up, such as when "the rules" (tagging tastes, standards and schemes) change.  This is why we have "don't tag for the renderer," and "don't abbreviate" and "ugh, ANOTHER re-working of trunk/expressway, AGAIN?!"  It gets better, as long as the problems to solve "get smaller and easier."  OSM has such problems, sure, and they do seem to get smaller, so I think we're doing OK.  However, we keep discovering "new realms" into which we pour our knowledge into the data of our map (database), and so "the space continues to get LARGER."

Happy mapping, everybody.

[1] https://en.wikipedia.org/wiki/Zeno's_paradoxes