[Talk-ca] Coastline Clean Up 2

David Groom reviews at pacific-rim.net
Sun Jan 29 13:43:28 GMT 2012


----- Original Message ----- 
From: "Paul Norman" <penorman at mac.com>
To: "'James A. Treacy'" <treacy at debian.org>; "'Andrew Allison'" 
<andrew.allison at teksavvy.com>
Cc: "'talk-ca'" <talk-ca at openstreetmap.org>
Sent: Sunday, January 29, 2012 3:55 AM
Subject: Re: [Talk-ca] Clean up - natural coastline


>> From: James A. Treacy [mailto:treacy at debian.org]
>> Subject: Re: [Talk-ca] Clean up - natural coastline
>>
>> On Sat, Jan 28, 2012 at 10:33:09AM -0500, Andrew Allison wrote:
>> > Hello:
>> > I'm in the process of recreating non-ct data.
>>
>> What areas need to be replaced?
>>
>
> I know that some PGS coastline imports are not CT-clean. Importing a
> coastline from CanVec to replace the PGS data is on my list to-do, but I
> don't see it happening before CanVec 9.0 since the current coastline model
> is broken on the west coast and it takes about an hour a tile when there's
> no other data to replace.
>

As I mentioned in my earlier email, over the last few months I've been going 
round clearing up a lot of coastline "issues" remaining from the original 
PGS import.

These include
 - retagging coastline ways rivers as multipolygons with waterway = 
riverbank
 - joining short sections of coastline ways together
 - dealing with the small coastline artefacts, either deleting them or 
combining them with the main coastline way

What I probably didn't make clear is that I have been doing this 
systematically on a worldwide basis.

Although I have done some changes in Canada, so far I have concentrated on 
Europe, South & Central America, & Africa. Its not that I have anything 
against Canada, its just that I had to start somewhere :)

However last week it occurred to me that if I were to continue my cleanup 
into Canada I would be wasting my time as Canada data could be better 
recreated from Cavnvec.

I then decided to look at some Canvec data and see what was involved.  I 
deliberately choose an area up in the arctic, so the only real data likely 
to be there was water body based.  I was surprised at how time consuming it 
was to correct the Canvec data and get it into a state ready to replace the 
OSM data.   This lead me to believe that a wholesale replacement of OSM 
Canada coastline with Canvec data was unlikely to happen any time soon.

The question therefore is, should I continue my cleanup of coastline data 
into Canada?

The downside is that eventually all my cleanup work should be rendered 
obsolete due to Canvec imports.

The upside is that it's likely I could complete the cleanup a lot sooner 
than Canvec is imported, and that by tidying up the coastline sooner, it may 
make it easier to merge the Canvec data into OSM without breaking coastlines 
and causing "flooding".

What are your thoughts?

David 





More information about the Talk-ca mailing list