[Imports] Ongoing Canadian building import needs to be stopped, possibly reverted
bike756 at gmail.com
Sat Jan 19 04:13:41 UTC 2019
Thanks for marking this as on-hold in the tasking manager. I know I came
in like a wrecking ball and I really appreciate y'all holding things up
while we discuss.
I'd be happy to validate data and help import the rest of central
Toronto once we're up and running again! I use the data in this area a
lot in my work... so I have a vested interest in keeping it at it's best :-)
As to the conflation issue, one of the things we're doing in the other
import I'm working on is that we've essentially split it into two parts.
First we're importing buildings that don't conflict with OSM at all -
this is the easy part - and only later will we go in a bit more
surgically and try to add tags to existing ways and replace geometries
with better data. We haven't started that part yet, though I imagine it
will be a real slog. IMO, it seems like a lot to ask that editors do
both things at once as these are really very different tasks, especially
given the size of the tasks here.
I wonder if you'd have any interest in a similar separation of tasks for
this import? I think one of the benefits is that less experienced
mappers can get their hands dirty on the easier new-data-import part,
without having to be expert on which geometry is better, how to preserve
way histories and tags, etc. Like I said, we haven't started this part
yet in the other import, but even I find the prospect a little daunting!
Jack of all trades, Master of Geography, PhD candidate in Urban Planning
On 1/18/19 2:15 PM, Yaro Shkvorets wrote:
> I'll change the project name to reflect that the import is on hold. As
> a local mapper, if you want to take a lead on the Toronto import
> that'd be great.
> I did review some of DannyMcD's edits last night
> (Mississauga-Brampton-Vaughan) and to be honest was rather
> disappointed with the quality. It appears Danny chose to import only
> new buildings (i.e. residential homes mostly), leaving most of the
> existing hand-traced non-residential building outlines in OSM
> untouched. That's unfortunate, the dataset offers some really good
> data and leaving half of it behind makes it more difficult to revisit
> in the future.
> In my edits (Markham-Scarborough-East York) I was aiming to replace as
> many existing geometries with outlines from the import as possible. I
> think that's what we should be trying to do going forward.
> Looking forward to your comments and discussion.
> On Fri, Jan 18, 2019 at 1:07 PM Nate Wessel <bike756 at gmail.com
> <mailto:bike756 at gmail.com>> wrote:
> Hi all,
> I've just joined the talk-ca list, so please accept my apologies
> for not addressing this list earlier. I'm happy to take this
> thread off the imports list for now and onto talk-ca until things
> are ready to begin again. The next person to reply can please feel
> free to remove that email if they agree.
> I've just made a note on the draft import plan wiki page noting
> that the import has been stopped:
> I would really appreciate it if the person with admin access to
> the tasking manager projects could please take those offline for
> the moment, or perhaps place them in a validation-only mode if
> that's possible.
> Like I said in my last email, which perhaps didn't make it to the
> talk-ca list
> I'm now proposing that we leave the data that has already been
> imported and enter a phase of thorough validation on that data.
> My plan, over the next several days, is to do a general survey of
> the quality of the data that has been imported so far and make a
> list of systematic issues I see that should be addressed before we
> can consider moving forward again. I'll add those comments to the
> conversation in talk-ca and on the wiki page (link above), as I
> feel is appropriate. As I said before, I'm of the mind that this
> import did not get adequate review or approval and did not follow
> all the import guidelines. I think therefore we need to take
> stock, cross the t's, dot the i's, and move this thing back toward
> where it needs to be. Step one is a thoroughly documented wiki
> page outlining the proposal and responding to everything required
> in the import guidelines.
> I know there are people excited about this import, and people who
> are eager to get back to work bringing buildings in, but I think
> everyone will be happier in the end if we take the time to do this
> right. We don't need to stop forever - we just need to stop until
> we get things right. I sincerely respect the good intentions of
> everyone involved in this and I hope we can all work together to
> make OSM a map known for it's coverage AND it's quality.
> Nate Wessel
> Jack of all trades, Master of Geography, PhD candidate in Urban
> NateWessel.com <http://natewessel.com>
> On 1/17/19 9:05 PM, OSM Volunteer stevea wrote:
>> The thread link is:https://lists.openstreetmap.org/pipermail/imports/2019-January/005878.html
> Best Regards,
> Yaro Shkvorets
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Imports