Re: [Talk-us] Proposed cleanup: NHD rivers

2011-03-23 Thread Alan Millar
On Mar 22, 2011, at 9:17 PM, Paul Norman wrote:
 This is now complete for the area west of Portland Oregon as a test.
 
 http://www.paulnorman.ca/blog/?attachment_id=96 shows the difference.

Nice improvement!  I like it.

 About 99.8% of the data was untouched since it was imported. I checked the
 other dozen or so ways by hand.

I did that import, and did a lot of cleaning up as I was importing it.  So I 
think that probably reduced the amount of cleanup that anyone needed to do 
later.

But I definitely felt at the time that the source data and value translations 
were a little vague.  I really like the current proposal and example results.  
Looks great!

- Alan


___
Talk-us mailing list
Talk-us@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk-us


Re: [Talk-us] Proposed cleanup: NHD rivers

2011-03-22 Thread Paul Norman
If there are no objections to the retagging part I'll proceed with retagging
FCodes 46003 and 46006 as documented on
http://wiki.openstreetmap.org/wiki/User:Pnorman/NHDCleanup

I will not be joining waterways at this time.

 -Original Message-
 From: Paul Norman [mailto:penor...@mac.com]
 Sent: Sunday, March 20, 2011 6:52 PM
 To: 'Richard Welty'; talk-us@openstreetmap.org
 Subject: Re: [Talk-us] Proposed cleanup: NHD rivers
 
 This is the view I subscribe to too. An example of two ways I would want
 to join would be http://www.openstreetmap.org/browse/way/68711710 and
 http://www.openstreetmap.org/browse/way/68710322
 These differ only in nhd:com_id and they're both really short ways.
 
 In any case, I'd like to make it clear that there are two separate
 parts.
 The retagging of the ways, and the joining of them. The first one is a
 serious issue, visible out to z8 in the rendering and hopefully
 uncontroversial to change. The second one is a less important issue that
 it seems more debate is required on.
 
 For the retagging, I've done up a table at
 http://wiki.openstreetmap.org/wiki/User:Pnorman/NHDCleanup which
 explains the changes I'm proposing. I should of thought of a table
 earlier
 
 Also, I need to empathize that any data edited by users since the
 imports won't be touched without manual review.
 
  -Original Message-
  From: Richard Welty [mailto:rwe...@averillpark.net]
  Sent: Sunday, March 20, 2011 5:38 PM
  To: talk-us@openstreetmap.org
  Subject: Re: [Talk-us] Proposed cleanup: NHD rivers
 
  On 3/20/11 8:16 PM, Nathan Edgars II wrote:
   On 3/20/2011 8:13 PM, Richard Welty wrote:
   d suggest using relations to group ways that are parts of named
   rivers rather than trying to combine the ways.
  
   If the only difference between the ways is that NHD assigns a
   different ID number to them, not combining them seems silly.
  if the ids are consistent from one release to the next and there is
  any notion of doing an update later, then combining them destroys
  useful information.
 
  richard
 
 
  ___
  Talk-us mailing list
  Talk-us@openstreetmap.org
  http://lists.openstreetmap.org/listinfo/talk-us
 
 
 ___
 Talk-us mailing list
 Talk-us@openstreetmap.org
 http://lists.openstreetmap.org/listinfo/talk-us


___
Talk-us mailing list
Talk-us@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk-us


Re: [Talk-us] Proposed cleanup: NHD rivers

2011-03-22 Thread Paul Norman
This is now complete for the area west of Portland Oregon as a test.

http://www.paulnorman.ca/blog/?attachment_id=96 shows the difference.

About 99.8% of the data was untouched since it was imported. I checked the
other dozen or so ways by hand.

 -Original Message-
 From: John Chambers [mailto:jcha...@gmail.com]
 Sent: Tuesday, March 22, 2011 3:48 PM
 To: talk-us@openstreetmap.org
 Subject: Re: [Talk-us] Proposed cleanup: NHD rivers
 
 I know of least one 46006 that I would consider a river (Tussahaw
 creek) , but doesn't have river in the name, but as bad as other NHD
 data I've seen is, this little problem will be small.
 
 upstream
 
 On Tue, Mar 22, 2011 at 4:50 PM, Paul Norman penor...@mac.com wrote:
  If there are no objections to the retagging part I'll proceed with
  retagging FCodes 46003 and 46006 as documented on
  http://wiki.openstreetmap.org/wiki/User:Pnorman/NHDCleanup
 
  I will not be joining waterways at this time.
 
  -Original Message-
  From: Paul Norman [mailto:penor...@mac.com]
  Sent: Sunday, March 20, 2011 6:52 PM
  To: 'Richard Welty'; talk-us@openstreetmap.org
  Subject: Re: [Talk-us] Proposed cleanup: NHD rivers
 
  This is the view I subscribe to too. An example of two ways I would
  want to join would be
  http://www.openstreetmap.org/browse/way/68711710 and
  http://www.openstreetmap.org/browse/way/68710322
  These differ only in nhd:com_id and they're both really short ways.
 
  In any case, I'd like to make it clear that there are two separate
  parts.
  The retagging of the ways, and the joining of them. The first one is
  a serious issue, visible out to z8 in the rendering and hopefully
  uncontroversial to change. The second one is a less important issue
  that it seems more debate is required on.
 
  For the retagging, I've done up a table at
  http://wiki.openstreetmap.org/wiki/User:Pnorman/NHDCleanup which
  explains the changes I'm proposing. I should of thought of a table
  earlier
 
  Also, I need to empathize that any data edited by users since the
  imports won't be touched without manual review.
 
   -Original Message-
   From: Richard Welty [mailto:rwe...@averillpark.net]
   Sent: Sunday, March 20, 2011 5:38 PM
   To: talk-us@openstreetmap.org
   Subject: Re: [Talk-us] Proposed cleanup: NHD rivers
  
   On 3/20/11 8:16 PM, Nathan Edgars II wrote:
On 3/20/2011 8:13 PM, Richard Welty wrote:
d suggest using relations to group ways that are parts of named
rivers rather than trying to combine the ways.
   
If the only difference between the ways is that NHD assigns a
different ID number to them, not combining them seems silly.
   if the ids are consistent from one release to the next and there is
   any notion of doing an update later, then combining them destroys
   useful information.
  
   richard
  
  
   ___
   Talk-us mailing list
   Talk-us@openstreetmap.org
   http://lists.openstreetmap.org/listinfo/talk-us
 
 
  ___
  Talk-us mailing list
  Talk-us@openstreetmap.org
  http://lists.openstreetmap.org/listinfo/talk-us
 
 
  ___
  Talk-us mailing list
  Talk-us@openstreetmap.org
  http://lists.openstreetmap.org/listinfo/talk-us
 
 
 ___
 Talk-us mailing list
 Talk-us@openstreetmap.org
 http://lists.openstreetmap.org/listinfo/talk-us


___
Talk-us mailing list
Talk-us@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk-us


[Talk-us] Proposed cleanup: NHD rivers

2011-03-20 Thread Paul Norman
A mapnik rendering change has revealed a problem in some areas with NHD
imported waterways. An example of the problem is at
http://www.openstreetmap.org/?lat=45.3lon=-123.3zoom=9layers=M

Essentially, all the streams are tagged as waterway=river, with
waterway=stream being used for what appear to be intermittent streams.

I propose doing the following changes. These changes would *only* be done to
ways that have not been modified since import. I have experience with this
type change from cleanup on Canadian NHN data.
1. Adding intermittent=yes to NHD streams.
2. Downgrading waterway=river to waterway=stream for non-rivers. 
3. Joining rivers into a single way

Steps 1 and 2 would be done in one set of imports while joining rivers would
be done in a second pass. 

Spot checks in the area linked indicate this would cause no problems. If
verification with imagery was necessary I'd use MapQuest's Open Aerial Map
as it seems to be the highest quality in these remote areas. 


___
Talk-us mailing list
Talk-us@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk-us


Re: [Talk-us] Proposed cleanup: NHD rivers

2011-03-20 Thread john
Has anyone determined for sure that the streams you plan to tag as intermittent 
are so, in fact?  This would require either getting confirmation from the 
organization that made the original survey, or at least checking with folks 
with local knowledge that a large. enough sample of the streams were, in fact, 
all intermittent.

---Original Email---
Subject :[Talk-us] Proposed cleanup: NHD rivers
From  :mailto:penor...@mac.com
Date  :Sun Mar 20 16:29:54 America/Chicago 2011


A mapnik rendering change has revealed a problem in some areas with NHD
imported waterways. An example of the problem is at
http://www.openstreetmap.org/?lat=45.3lon=-123.3zoom=9layers=M

Essentially, all the streams are tagged as waterway=river, with
waterway=stream being used for what appear to be intermittent streams.

I propose doing the following changes. These changes would *only* be done to
ways that have not been modified since import. I have experience with this
type change from cleanup on Canadian NHN data.
1. Adding intermittent=yes to NHD streams.
2. Downgrading waterway=river to waterway=stream for non-rivers. 
3. Joining rivers into a single way

Steps 1 and 2 would be done in one set of imports while joining rivers would
be done in a second pass. 

Spot checks in the area linked indicate this would cause no problems. If
verification with imagery was necessary I'd use MapQuest's Open Aerial Map
as it seems to be the highest quality in these remote areas. 


___
Talk-us mailing list
Talk-us@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk-us

-- 
John F. Eldredge -- j...@jfeldredge.com
Reserve your right to think, for even to think wrongly is better than not to 
think at all. -- Hypatia of Alexandria
___
Talk-us mailing list
Talk-us@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk-us


Re: [Talk-us] Proposed cleanup: NHD rivers

2011-03-20 Thread Paul Norman
I checked with http://wiki.openstreetmap.org/wiki/NHD#Mapping and
StreamRiver 46003 was mapped to waterway=stream, and the description is
intermittent streams. As far as I can tell, nothing else was mapped to
waterway=stream.

NAIP imagery seems to verify this. 

Seeing that the FCode was imported, I'm thinking I'll use that to identify
which are rivers and which are streams. Essentially, this will be changing
46003 to intermittent streams and 46006 to streams, with exceptions for
46006 where the name indicates it's a river.

 -Original Message-
 From: j...@jfeldredge.com [mailto:j...@jfeldredge.com]
 Sent: Sunday, March 20, 2011 2:42 PM
 To: OpenStreetMap talk-us list
 Subject: Re: [Talk-us] Proposed cleanup: NHD rivers
 
 Has anyone determined for sure that the streams you plan to tag as
 intermittent are so, in fact?  This would require either getting
 confirmation from the organization that made the original survey, or at
 least checking with folks with local knowledge that a large. enough
 sample of the streams were, in fact, all intermittent.
 
 ---Original Email---
 Subject :[Talk-us] Proposed cleanup: NHD rivers
 From  :mailto:penor...@mac.com
 Date  :Sun Mar 20 16:29:54 America/Chicago 2011
 
 
 A mapnik rendering change has revealed a problem in some areas with NHD
 imported waterways. An example of the problem is at
 http://www.openstreetmap.org/?lat=45.3lon=-123.3zoom=9layers=M
 
 Essentially, all the streams are tagged as waterway=river, with
 waterway=stream being used for what appear to be intermittent streams.
 
 I propose doing the following changes. These changes would *only* be
 done to ways that have not been modified since import. I have experience
 with this type change from cleanup on Canadian NHN data.
 1. Adding intermittent=yes to NHD streams.
 2. Downgrading waterway=river to waterway=stream for non-rivers.
 3. Joining rivers into a single way
 
 Steps 1 and 2 would be done in one set of imports while joining rivers
 would be done in a second pass.
 
 Spot checks in the area linked indicate this would cause no problems. If
 verification with imagery was necessary I'd use MapQuest's Open Aerial
 Map as it seems to be the highest quality in these remote areas.
 
 
 ___
 Talk-us mailing list
 Talk-us@openstreetmap.org
 http://lists.openstreetmap.org/listinfo/talk-us
 
 --
 John F. Eldredge -- j...@jfeldredge.com
 Reserve your right to think, for even to think wrongly is better than
 not to think at all. -- Hypatia of Alexandria
 ___
 Talk-us mailing list
 Talk-us@openstreetmap.org
 http://lists.openstreetmap.org/listinfo/talk-us


___
Talk-us mailing list
Talk-us@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk-us


Re: [Talk-us] Proposed cleanup: NHD rivers

2011-03-20 Thread James U
1 and 2 make sense to me.  What criteria would you use for 2?  I have 
done a fair bit of NHD imports and simply used the name, i.e.  river, 
to classify rivers.  Some parts of the country have different naming 
traditions that others.

What is the rationale for 3?  




On Sunday, March 20, 2011 05:29:54 pm Paul Norman wrote:
 A mapnik rendering change has revealed a problem in some areas with 
NHD
 imported waterways. An example of the problem is at
 
http://www.openstreetmap.org/?lat=45.3lon=-123.3zoom=9layers=M
 
 Essentially, all the streams are tagged as waterway=river, with
 waterway=stream being used for what appear to be intermittent 
streams.
 
 I propose doing the following changes. These changes would *only* be 
done
 to ways that have not been modified since import. I have experience 
with
 this type change from cleanup on Canadian NHN data.
 1. Adding intermittent=yes to NHD streams.
 2. Downgrading waterway=river to waterway=stream for non-rivers.
 3. Joining rivers into a single way
 
 Steps 1 and 2 would be done in one set of imports while joining rivers
 would be done in a second pass.
 
 Spot checks in the area linked indicate this would cause no problems. 
If
 verification with imagery was necessary I'd use MapQuest's Open Aerial 
Map
 as it seems to be the highest quality in these remote areas.
 
 
 ___
 Talk-us mailing list
 Talk-us@openstreetmap.org
 http://lists.openstreetmap.org/listinfo/talk-us

___
Talk-us mailing list
Talk-us@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk-us


Re: [Talk-us] Proposed cleanup: NHD rivers

2011-03-20 Thread Paul Norman
Name for 2. It might miss some rivers, but the data source doesn't
differentiate between streams and rivers in any of the metadata.

3 is about making the rivers into single ways, more like a mapper would do
by hand. I'm not really set on this step and if done it would be after steps
1 and 2 have been done everywhere. Looking at nhd:com_id it might cause
problems with updating, so I'm thinking I'll drop this step for now.

 -Original Message-
 From: James U [mailto:jumba...@gmail.com]
 Sent: Sunday, March 20, 2011 4:42 PM
 To: talk-us@openstreetmap.org
 Subject: Re: [Talk-us] Proposed cleanup: NHD rivers
 
 1 and 2 make sense to me.  What criteria would you use for 2?  I have
 done a fair bit of NHD imports and simply used the name, i.e. 
 river, to classify rivers.  Some parts of the country have different
 naming traditions that others.
 
 What is the rationale for 3?
 
 
 
 
 On Sunday, March 20, 2011 05:29:54 pm Paul Norman wrote:
  A mapnik rendering change has revealed a problem in some areas with
 NHD
  imported waterways. An example of the problem is at
 
 http://www.openstreetmap.org/?lat=45.3lon=-123.3zoom=9layers=M
 
  Essentially, all the streams are tagged as waterway=river, with
  waterway=stream being used for what appear to be intermittent
 streams.
 
  I propose doing the following changes. These changes would *only* be
 done
  to ways that have not been modified since import. I have experience
 with
  this type change from cleanup on Canadian NHN data.
  1. Adding intermittent=yes to NHD streams.
  2. Downgrading waterway=river to waterway=stream for non-rivers.
  3. Joining rivers into a single way
 
  Steps 1 and 2 would be done in one set of imports while joining rivers
  would be done in a second pass.
 
  Spot checks in the area linked indicate this would cause no problems.
 If
  verification with imagery was necessary I'd use MapQuest's Open Aerial
 Map
  as it seems to be the highest quality in these remote areas.
 
 
  ___
  Talk-us mailing list
  Talk-us@openstreetmap.org
  http://lists.openstreetmap.org/listinfo/talk-us
 
 ___
 Talk-us mailing list
 Talk-us@openstreetmap.org
 http://lists.openstreetmap.org/listinfo/talk-us


___
Talk-us mailing list
Talk-us@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk-us


Re: [Talk-us] Proposed cleanup: NHD rivers

2011-03-20 Thread Richard Welty

On 3/20/11 7:50 PM, Paul Norman wrote:

3 is about making the rivers into single ways, more like a mapper would do
by hand. I'm not really set on this step and if done it would be after steps
1 and 2 have been done everywhere. Looking at nhd:com_id it might cause
problems with updating, so I'm thinking I'll drop this step for now.

i'd suggest using relations to group ways that are parts of named rivers 
rather

than trying to combine the ways.

richard


___
Talk-us mailing list
Talk-us@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk-us


Re: [Talk-us] Proposed cleanup: NHD rivers

2011-03-20 Thread Nathan Edgars II

On 3/20/2011 8:13 PM, Richard Welty wrote:

On 3/20/11 7:50 PM, Paul Norman wrote:

3 is about making the rivers into single ways, more like a mapper
would do
by hand. I'm not really set on this step and if done it would be after
steps
1 and 2 have been done everywhere. Looking at nhd:com_id it might cause
problems with updating, so I'm thinking I'll drop this step for now.


i'd suggest using relations to group ways that are parts of named rivers
rather
than trying to combine the ways.


If the only difference between the ways is that NHD assigns a different 
ID number to them, not combining them seems silly.


___
Talk-us mailing list
Talk-us@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk-us


Re: [Talk-us] Proposed cleanup: NHD rivers

2011-03-20 Thread Ian Dees
On Sun, Mar 20, 2011 at 7:18 PM, Ian Dees ian.d...@gmail.com wrote:

 On Sun, Mar 20, 2011 at 7:13 PM, Richard Welty rwe...@averillpark.netwrote:

 On 3/20/11 7:50 PM, Paul Norman wrote:

 3 is about making the rivers into single ways, more like a mapper would
 do
 by hand. I'm not really set on this step and if done it would be after
 steps
 1 and 2 have been done everywhere. Looking at nhd:com_id it might cause
 problems with updating, so I'm thinking I'll drop this step for now.

  i'd suggest using relations to group ways that are parts of named rivers
 rather
 than trying to combine the ways.


 There were several clients that did not attempt to connect the ways of each
 NHD linestring so there are duplicated nodes and ways that touch but aren't
 joined. Hopefully fixing this is what Paul was talking about.


... and the reach code is what should be joined on, probably not the
com_id.
___
Talk-us mailing list
Talk-us@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk-us


Re: [Talk-us] Proposed cleanup: NHD rivers

2011-03-20 Thread Mike N

On 3/20/2011 8:18 PM, Ian Dees wrote:

There were several clients that did not attempt to connect the ways of
each NHD linestring so there are duplicated nodes and ways that touch
but aren't joined. Hopefully fixing this is what Paul was talking about.


  Agreed - this is a safe case of duplicate node removal: when all 
items touching duplicate nodes are of type hydro - river, stream, lake, 
reservoir, etc (I forget the exact tag names).


   Just combining segments within the same reach code will frequently 
produce ways containing   500 nodes, which makes editing more 
complicated, and does not make any improvement to the usability of the 
map data.



___
Talk-us mailing list
Talk-us@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk-us


Re: [Talk-us] Proposed cleanup: NHD rivers

2011-03-20 Thread Ian Dees
On Sun, Mar 20, 2011 at 7:30 PM, Mike N nice...@att.net wrote:

 On 3/20/2011 8:18 PM, Ian Dees wrote:

 There were several clients that did not attempt to connect the ways of
 each NHD linestring so there are duplicated nodes and ways that touch
 but aren't joined. Hopefully fixing this is what Paul was talking about.


  Agreed - this is a safe case of duplicate node removal: when all items
 touching duplicate nodes are of type hydro - river, stream, lake, reservoir,
 etc (I forget the exact tag names).

   Just combining segments within the same reach code will frequently
 produce ways containing   500 nodes, which makes editing more complicated,
 and does not make any improvement to the usability of the map data.

 The ways don't need to be *combined*, just *connected*.
___
Talk-us mailing list
Talk-us@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk-us


Re: [Talk-us] Proposed cleanup: NHD rivers

2011-03-20 Thread Nathan Edgars II

On 3/20/2011 9:12 PM, Richard Welty wrote:

if combining them meaningfully improves the map, by all means do it.


Or improves editing.

___
Talk-us mailing list
Talk-us@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk-us


Re: [Talk-us] Proposed cleanup: NHD rivers

2011-03-20 Thread Paul Norman
This is the view I subscribe to too. An example of two ways I would want to
join would be http://www.openstreetmap.org/browse/way/68711710 and
http://www.openstreetmap.org/browse/way/68710322 
These differ only in nhd:com_id and they're both really short ways.

In any case, I'd like to make it clear that there are two separate parts.
The retagging of the ways, and the joining of them. The first one is a
serious issue, visible out to z8 in the rendering and hopefully
uncontroversial to change. The second one is a less important issue that it
seems more debate is required on.

For the retagging, I've done up a table at
http://wiki.openstreetmap.org/wiki/User:Pnorman/NHDCleanup which explains
the changes I'm proposing. I should of thought of a table earlier

Also, I need to empathize that any data edited by users since the imports
won't be touched without manual review.

 -Original Message-
 From: Richard Welty [mailto:rwe...@averillpark.net]
 Sent: Sunday, March 20, 2011 5:38 PM
 To: talk-us@openstreetmap.org
 Subject: Re: [Talk-us] Proposed cleanup: NHD rivers
 
 On 3/20/11 8:16 PM, Nathan Edgars II wrote:
  On 3/20/2011 8:13 PM, Richard Welty wrote:
  d suggest using relations to group ways that are parts of named
  rivers rather than trying to combine the ways.
 
  If the only difference between the ways is that NHD assigns a
  different ID number to them, not combining them seems silly.
 if the ids are consistent from one release to the next and there is any
 notion of doing an update later, then combining them destroys useful
 information.
 
 richard
 
 
 ___
 Talk-us mailing list
 Talk-us@openstreetmap.org
 http://lists.openstreetmap.org/listinfo/talk-us


___
Talk-us mailing list
Talk-us@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk-us