[Imports-us] Union County Georgia Building Footprints

Tue Dec 3 14:27:14 UTC 2013

Good Morning Frederik - I guess I owe Carol a beer.

-----------------
Randal Hale, GISP
North River Geographic Systems, Inc
http://www.northrivergeographic.com
423.653.3611 rjhale at northrivergeographic.com
<mailto:rjhale at northrivergeographic.com>
twitter:rjhale
http://about.me/rjhale

On 12/03/2013 02:38 AM, Frederik Ramm wrote:
> Hi,
>
>     couple general comments:
>
> On 11/30/13 18:23, Randal Hale wrote:
>> Union County Georgia made their buildings (as well as roads) accessible
>> through a public domain license - they've asked myself (rjhale1971) and
>> Carol (geocak) to upload them into OpenStreetMap.
> What's your assessment of the OSM community in Union County Georgia? Do
> you know the people who are active there, have you been in touch?
My goal is to build a community there - Right now there isn't  one to 
speak of as you point out throughout your email - but since I've become 
known as the guy who constantly says "OSM is a good thing" - I've gotten 
some attention and a reputation to uphold. The GIS department approached 
me a year ago to import their roads which are infinitely better than the 
current tiger roads - but I had a theory on any upload we would attempt 
past the Athens dust up and buildings were the lowest hanging fruit. 
Plus the road data (while being much better geometry wise) doesn't have 
any attribution past a addr:street so I've got a lot of work to do to 
get that into any presentable shape. It almost going to be easier to fix 
the existing roads.
>
> Just to avoid any misunderstandings, we are talking of this right:
>
> http://www.openstreetmap.org/relation/532213#map=11/34.8316/-83.9919
Absolutely wonderful area - Not terribly far from our own slice of 
Germany. Heh. Come over one day and we will horrify you with Helen 
Georgia. I owe you a beer anyway.
>
> Looking just at the name of the user who have last edited any object in
> this county, only the following people have more than 100 edits in the area:
>
>    71555 jumbanho
>    50903 woodpeck_fixbot
>    20677 Liber
>     6589 sadam-AT
>     2274 bot-mode
>     1671 EdLoach
>     1175 DaveHansenTiger
>      789 Le Fou
>      530 rh3662
>      508 ksamples
>      299 iandees
>      226 BeauMan
>      140 egore911
>      137 42429
>
> Zooming in a bit, it becomes immediately clear that there seems to be a
> problem with fixing bad imports in Union County Georgia - e.g.
Yes - I would love to rip all the landcover out of Georgia. Probably the 
NHD and do it right. It is sad that a bot (are they actual robots or SQL 
scripts? Ha. ) made it into your list. Not unexpected though. Once we 
have a well planned out edit day in Athens I am inviting the guys from 
Union county down. Hopefully we hold an OSM meetup there also.
>
> http://www.openstreetmap.org/relation/532213#map=15/34.9172/-84.0614
>
> shows clearly overlapping water and wood areas from USGS and NHD - only
> one can be right (and Bing imagery claims a third way in some areas I
> have looked at).
Bing Imagery is hit or miss since it's a combination of data sources. 
Since I don't know the vintage or quality of Bing (my general feeling is 
only use it if it's the only thing available) but my gut feeling is 
don't use it until it's investigated more - which is why I stopped at 
two edits. If they have imagery I could ask that they share it - it is 
always going to be better than bing. My worry as a GIS person is the 
ready acceptance of Bing imagery in OSM.
>
> In addition to that, there's TIGER data that is totally unrelated to
> things on the ground.
>
> These three imports seem to have been there for years, unloved and
> unfixed. Do you have any ideas/action plans to remedy that, to get more
> people interested in doing the work that needs to be done to make the
> map good in Union County Georgia? You say you'd like to get the GIS dept
> to get involved more but they're unlikely to be able to fix this without
> help from the population at large.
My current argument is that OSM is a great compliment to a GIS 
department. The general population can be more places than the GIS 
department. I was able to at least keep momentum here with no import 
(well - I did a NHD import back when that was cool)  - 
http://www.openstreetmap.org/search?query=Chattanooga%20Tn#map=16/35.0560/-85.3041 
. I personally believe it will take an import with people going "Hey 
where is my house" to get this area kickstarted. I could be wrong - I'm 
guessing you would weigh in on the wrong side. I would gleefully go in 
the other direction. I want to go up and kill a saturday making edits. I 
want the people in the county to get excited and start making edits. 
This is my first attempt. Which will (is currently) be documented for 
the benefit of others.
>
> If you import building footprints here, they are
>
> * (almost) guaranteed to conflict with old imported never-fixed USGS
> data (buildings in forests etc.)
My plan is to look at the data and decide based on it's merit. If it's 
existing data and looks good - it stays. If it's existing data and is 
not good - it goes. I expect more deletion than addition.
>
> * (almost) guaranteed to conflict with old imported never-fixed NHD data
> (buildings in lakes and rivers etc.)
Yeah - I've been trying to decide how to fix the NHD (continue using my 
grid perhaps). My guess is it's going to make the building import slow. 
Landcover is going to make everything slow. If only OSM were more like 
GIS and I could clip and remove the landcover easier than I think I 
could. Hopefully someone will tell me how that can be done.
>
> * (almost) guaranteed to conflict with old imported never-fixed TIGER
> roads (buildings sitting on top of streets etc.)
Yeah - Tiger data. The bane of GIS people - that's why we hardly ever 
use it until the last version was released.
>
> Just to be clear, when you say
>
>>        o This isn't a scripted import. Each grid will be checked in JOSM
>>          and validated before it's uploaded.
> does that mean that you will fix the existing data to not be in conflict
> with your buildings, are you planning to only check for errors within
> your building data set?
Well - I will no doubt end up fixing more than that - but that will be 
my main concern. I would absolutely love to clip and cut all the landuse 
out of union as a first attempt at this. As you can see someone has 
removed (and I would love to take the credit) 80% of the landcover in 
Athens-Clarke County.
>
> You write that you and Carol have been asked to do the import. I note
> that you have 2 (two) edits in that county and Carol has zero. Would you
> consider yourselves members of the OSM community in Union County
> Georgia? If so, have you never noticed the above broken import and felt
> the urge to fix them, at least one tiny bit?
Thanks for pointing that out. I hope you at least fixed some things 
while you were generating these statistics? My goal is to stir up 
interest. I think my best way to get that done is to start here. With 
any hope I continue teaching the URISA/OSM class that myself and Carol 
help author and we start holding regular meetings in the area. My goal 
is to involve the general population to make edits on the ground - it's 
how I've gotten 99% of any OSM task I've started. With the exception of 
the High School where I regularly taught OSM - their edits were all by 
armchair. It's going to be a task with no clear end as these things 
alway are. Given our involvement with the local GIS community we have at 
least a path to beginning that. Carol is excited about the community in 
Athens - my goal will be to push in other areas. If I get one more 
mapping that wasn't - that's one more than we had.  The general feel 
here is that OSM is too cryptic - so I'm attempting to remove that 
impression and show the benefit to my local area.
>
> Just to clarify, when you say you have been "asked to do the import",
> does that mean that you or your company has been contracted to do the
> import, i.e. will you be receiving any benefits from this other than
> seeing the map improved, or is this a hobby activity for you?
Does it matter? For the record the answer is no. If it were yes - does 
that mean the import gets blocked? Is it a hobby or a business for you? 
We see this as a community effort. That's why Carol has been hammering 
on everyone to define community. My thought are that you can't/won't - 
because it is too bloblike with no definition and defining it sets up 
more questions that no one wants to answer. My personal feelings is that 
it is small and localized. I think at one time I was rated somewhere 
towards the top of some list for edits - which would mean that the total 
community is rather small (given there are well over 1,000,000 
registered accounts). So small communities tend to move in an unhealthy 
direction (IMO). I hope this action starts new people coming in.....I 
hope when they make mistakes that they are helped.

So how does GeoFabrik make money off OSM? I've used your downloads quite 
a bit. I would love to incorporate OSM into some sort of business case 
as mapbox/Geofabrik has done.

and I guess I should say - thank for volunteering to help in Union 
County? Please start deleting landcover - it's terrible.

Or

You could have proven your point in a lot fewer sentences. So please 
help fix the data? or does Union County extend outside your community?
>
> Bye
> Frederik
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/imports-us/attachments/20131203/a2c8b5d0/attachment.html>