[Imports] Ongoing Canadian building import needs to be stopped, possibly reverted

OSM Volunteer stevea steveaOSM at softworkers.com
Fri Jan 18 02:03:54 UTC 2019

This is redirected (by request of its author) from a thread on the (talk-) imports mailing list at .
On Jan 17, 2019, at 4:55 PM, John Whelan <jwhelan0112 at gmail.com> wrote:
> The import was discussed on talk-ca and in my opinion there was a consensus of opinion it should go ahead. The data comes from the municipalities of which there are some 37,000 separate ones in Canada.  The idea of a single import plan was suggested on talk-ca by someone not involved rather than have 37,000 different import plans.  Many municipalities are very small.

There was a serious dearth of reply, and nothing even approaching "consensus of opinion," indicating (to me and likely others) that a nationwide import did not have the wide, national consensus it must have to continue.  John, we're simply going to disagree about that, it seems.  Especially in light of the events in this desire/wiki/project going back to 2017, MUCH more consensus ought to have been built.  I kept my mouth largely shut at the reboot two months ago, yet here we are.

> If you look at Canada there are locations with poor imagery and lots of buildings and my suspicion is some were being imported without the licenses really meeting the OSM licensing requirements.

Really?  Then, fix this.  That's a bit blunt of me, yes, yet, there it is.  Stop.  Fix this.  You need coordination, via built-in tools like wiki, this list, and a real dialog made up of real conversation for this to continue.  There appears to be a desire to do this, but there does not appear to be anybody willing to take responsibility to do it right, except perhaps (presently) Yaro, who John apparently not only has never met, but in all likelihood, has never communicated with.  Hence, "let your wiki communicate for you."  It is very good at this, I speak from personal experience on nationwide efforts, though I will tell you it takes work, dedication, time and writing skills.  Dismissing wiki as "too detailed" or "ineffective" is not a green flag to begin a race.  It means "stop, fix the wiki so anybody who comes along will know what to do."  So:  "bzzzt."  Not there yet, though, it seems Yaro was able to glean enough to make progress.  Grow his kernel of expertise and expand what works about it.

> Many locations in Canada do not have a group of OpenStreetMap local mappers.  Many locations do.  Locating local groups is difficult.  My understanding is it has always been it is the local community who make the decision to do an import.  We did use talk-ca and if you wish to continue the conversation then I would suggest that is the place to hold the conversation.  I think the problem here is defining the ideal local community.  It is probably somewhere between the whole of Canada and each individual municipality and contains groups of local mappers.

Really?  Then, fix this.  A national project (nearly single-handedly) like this into OSM is a very, very tall order.  It seems quite disingenuous to me to initiate these data and "hope for the best."  It's not exactly the same as giving matches, gasoline, dynamite, liquor and car keys to teenagers and being surprised at a tragic outcome, but it does stray in that direction of irresponsibility.  This time, I don't think you are allowed to be surprised.

> The intent is certainly to involve local mappers, building outlines by themselves are especially interesting but building outlines with added tags are much more interesting.  We need boots on the ground, we need local mappers onside. 

You (John, others, the poorly-focused impetus to import these data...) DID get local mappers involved.  Today, I believe John implied or said he hasn't even communicated with them.  (Or barely has).  Yet while some work seems patient and better quality, enough bad data alarmed Nate (and others, myself included) enough to say something.  This simply shouldn't happen.

Please right your ship.  I don't pretend to represent all of OSM as I ask this, but I do stand tall in this project and politely ask you to do so.  Nate wrote some powerfully-potent directions for this building importation to go in with his later post.  Those are a good start, a "productive way forward."  Talk-ca (this channel) and wiki (including its Talk page, Discussion tab) are wide-area, very open, vetted, scrutinizable methodologies to communicate about this.  Let's keep the way(s) forward OPEN for discussion so this doesn't happen any longer.  Please!

We don't need 37,000 import plans and nobody is suggesting we do.  We do need good project management to import Canada's buildings.  There, I said it.  (Again).  This starts with good, open communication.  Nothing less will do, or you consign your project to a whispered hope upon the wind, and that won't do it.


More information about the Imports mailing list