[Talk-ca] Some feedback on import quality in Toronto

[OSM Volunteer stevea]

Mmm, careful with your language, John.  The data "have a license which is compatible with OSM's ODbL" (is an accurate way to say it).  I believe that took about eight years and was a difficult slog, a lot of hard work by many, lessons learned from Ottawa, a determination by OSM's LWG, but it is done.  (I am grateful for that, it is an important milestone).

A different issue is whether the data and their concomitant quality "are acceptable" to the OSM community.  The license being ODbL compatible is not that; these are different issues.  We now discuss data quality (and what to do to improve them, if anything) here in talk-ca and in the mildly-being-updated Import Plan, with experiences of what Yaro, Danny and Nate have done in Toronto.

It is not the case, as John says, that "the data (are) licensed to be acceptable to (OSM)."  The concept of "acceptability" is not related to the data being licensed compatible with ODbL.

What often/usually happens is an Import Plan gets wide vetting and acceptance by the OSM community before data become imported, including suitability/acceptability of the quality of the data.  What happened here is the Import Plan was attempted to be widely vetted, but this seems to have been largely ignored or little-paid-attention-to.  While the Plan's initial shortcomings were pointed out, yet not remedied, the Import began, then the community began to react (with some complaint and some "what Yaro does seems OK").  Yaro (and Nate, I believe) then updated the Import Plan with Yaro's specific technical steps.

Still, complaints and/or disagreements about simplification, squaring and potentially other issues continue.  (As two asides, I'll say that MANY buildings are not necessarily "square" — like buildings with bay windows — and that truly square/rectangular buildings should express this with a tagged way/closed polygon made up of exactly four nodes).

As these discussions continue, eventually what will emerge is "how do we (algorithmically, manually...) change the data, whether pre-import or during-import, so that they achieve wide acceptance as to their quality by the community."  We're getting closer, it seems to me, but I don't think we're there yet.

Nobody seems to be arguing there is or isn't a "desire to bring in the buildings by many" or to "use them for many purposes."  That point seems "decided."  The questions remain:  "how, with the existing data?"  Once those are determined (and documented in the Import Plan), over time, the data will (or will not) be imported.  I wish to offer encouragement to this process, it does appear to continue here and can likely bear fruit in the near future.  Keep going!  Consensus is ahead!

SteveA

On Feb 3, 2019, at 10:55 AM, john whelan <jwhelan0112 at gmail.com> wrote:
> So I suggest that you name yourself as the coordinator on the wiki page for Toronto that allows the local mappers in Toronto to import at the rate and to the standard you suggest.
> 
> For the rest of the country the data is licensed to be acceptable to OpenStreetMap thus anyone can set up their own import plan and import it even if this import is abandoned.  I'd like to see this voiced as the general desire though on talk-ca before it happens as it was a talk-ca decision to proceed.
> 
> My reading of the posts on talk-ca is that there is a desire to bring in the buildings by many.  There is also a desire by many to use them for many purposes.
> 
> Cheerio John
> 
> On Sun, Feb 3, 2019, 1:42 PM Nate Wessel <bike756 at gmail.com wrote:
> John, 
> You seem to be mostly addressing topics which have been brought up elsewhere. My email was meant to address specific data quality issues in Toronto, so I'm not sure how to respond to all of this. 
> To your broader question though, my position is that we *do* have the volunteers and skills necessary to make this a good import. Supposing that we didn't though, then I would have to say that the import should wait until we have the right people working on it. A bad import is worse than no import.
> 
> Cheers,
> Nate Wessel
> Jack of all trades, Master of Geography, PhD candidate in Urban Planning
> NateWessel.com 
<remainder redacted for brevity>