[Talk-ca] FW: Re: Importing buildings in Canada

[Nate Wessel]

Howdy all,

a) I would also be happy to host data - I plan to preprocess/host data 
for Toronto/GTA at the least.

b) I'd suggest sticking to no more than ~200 buildings or ~1,500 nodes 
for an import task. If that seems too small, you can always follow it up 
with another task right after, without taking more than a few extra 
seconds to click through the task manager interface. I find that the 
amount of time needed varies widely. Personally, I wouldn't want to 
spend an entire hour on a task - 20 minutes max. It's easier to select 
another task than to decide what to do with a half finished one. Better 
to go too small than too large!

c) That totally depends on the quality of the data to be checked. The 
data here in Toronto was pretty much dumped as is. That means that most 
of the work which should have been done during the import has to be done 
now, retroactively. It is definitely not faster than actively importing. 
I find myself cleaning up a lot of other old/low-quality data along the 
way too which can take up as much time as you let it.

d) Keep in mind that the "main activity center" won't account for 
bimodal distributions. My center is still somewhere in Ohio, though I've 
been active in Toronto for years. The "past six months" option puts me 
in the right place. Personally, I wouldn't bother users with <100 
changesets, but that might just be a Toronto perspective since we have a 
lot of very active mappers here. Smaller cities will need to adjust 
that. I might suggest trying to contact the top 10 mappers in an area. 
They would likely know who else to contact if anyone.

Best,

Nate Wessel, PhD
Planner, Cartographer, Transport Nerd
NateWessel.com <https://www.natewessel.com>

On 2020-01-15 8:33 a.m., Daniel @jfd553 wrote:
>
> Bonjour Groupe,
>
> Concerning the proposal (ODB import), there are questions that remain 
> before moving forward; here are some of them…
>
> a)Who on this list host the ODB data? If pre-processing is required, 
> how should we proceed to have the results made available (assuming I 
> ran the pre-process)?
>
> b)Nate mentioned that the working units (shapes) proposed by the task 
> manager are customizable according to data density. So, how many 
> buildings can be imported properly (i.e. following the procedure) in 
> an hour or so? Would it be a good size to proceed?
>
> c)There are areas where the data has already been imported. What would 
> be the size of an import check task? I expect that a much larger 
> number of buildings can be checked in an hour or so, but by how much 
> (10x)?
>
> d)Who should be contacted when trying to get the local mappers buy-in? 
> IMHO I would contact only those found by Pascal Neis’ tool [1], which 
> would have contributed more than 10 changesets and for which the main 
> activity centre is within a the concerned municipality (tick/untick 
> the display option boxes [1]).
>
> Proposals or comments you which to share?
>
> Daniel
>
> [1] Overview of OpenStreetMap Contributors aka Who’s around me? 
> http://resultmaps.neis-one.org/oooc?
>
> *From:*Daniel @jfd553 [mailto:jfd553 at hotmail.com]
> *Sent:* Saturday, January 11, 2020 15:41
> *To:* talk-ca at openstreetmap.org
> *Subject:* [Talk-ca] FW: Re: Importing buildings in Canada
>
> By the way, have a look at
>
> https://wiki.openstreetmap.org/wiki/Canada_Building_Import
>
> https://wiki.openstreetmap.org/wiki/The_Open_Database_of_Buildings
>
> Cheers
>
>
> _______________________________________________
> Talk-ca mailing list
> Talk-ca at openstreetmap.org
> https://lists.openstreetmap.org/listinfo/talk-ca
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/talk-ca/attachments/20200115/dbae82aa/attachment.htm>