Re: [OSM-talk] source=Google

2013-05-24 Thread Steve Bennett
On Tue, May 21, 2013 at 9:26 PM, Martin Koppenhoefer
dieterdre...@gmail.com wrote:
 What do you propose to do with source tags found on an object when you
 modify this object based on a different source?

Speaking for myself, I either replace it (if I'm replacing virtually
all the geometry) or supplement it: source=gps;Bing. If I'm tweaking
something that had no source before, I sometimes use
source=unknown;Bing.

Steve

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


Re: [OSM-talk] source=Google

2013-05-21 Thread Dave F.

On 18/05/2013 16:56, Martin Koppenhoefer wrote:

Point is it doesn't really belong in the data, because it is metadata.


This is a false argument. There's nothing wrong with metadata in the 
database if it's *useful*.


When I go for a walk I survey. I use a gps. When mapping the data I use 
Bing imagery to confirm where I went  add details I couldn't fully see 
like hedges etc. I then use the internet to search for names of shops  
places that I forgot. All different sources within one changeset edit.


Dave F.

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


Re: [OSM-talk] source=Google

2013-05-21 Thread Martin Koppenhoefer
2013/5/21 Dave F. dave...@madasafish.com

 On 18/05/2013 16:56, Martin Koppenhoefer wrote:

 Point is it doesn't really belong in the data, because it is metadata.


 This is a false argument. There's nothing wrong with metadata in the
 database if it's *useful*.

 When I go for a walk I survey. I use a gps. When mapping the data I use
 Bing imagery to confirm where I went  add details I couldn't fully see
 like hedges etc. I then use the internet to search for names of shops 
 places that I forgot. All different sources within one changeset edit.




Yes, you could either split your edits into similar edits (useful IMHO, as
it introduces some logical system into the way you perform uploads) or
combine the sources into something like survey and tracing from bing
aerial imagery and websearch

IMHO it is useful to
a) know from when the aerial imagery was you traced from
b) know whether the mapper knew the area or only traced from aerial imagery
c) know whether the mapping was based on a recent survey
d) know in particular cases like import from supposedly precise sources
(e.g. cadastre) where the data was taken from
e) know the reason when an object gets modified or deleted (e.g. the pub
was closed)
f) know the source of hard to survey real world objects (e.g. undersea
cables, underground stuff)
g) store the source of data that requires some kind of attribution

But in the end I think this whole source thing is completely overestimated.
In the end the following mappers will compare what is on the map with what
they know or believe to be there in reality, and in case of discrepancies
will probably modify the map based on their findings, regardless of any
source tag.

What do you propose to do with source tags found on an object when you
modify this object based on a different source?

cheers,
Martin
___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk


Re: [OSM-talk] source=Google

2013-05-21 Thread Richard Fairhurst
Martin Koppenhoefer wrote:
 But in the end I think this whole source thing is completely 
 overestimated.

Yup.

 What do you propose to do with source tags found on an object when 
 you modify this object based on a different source?

OSM has full object history. :)

cheers
Richard





--
View this message in context: 
http://gis.19327.n5.nabble.com/source-Google-tp5761629p5762025.html
Sent from the General Discussion mailing list archive at Nabble.com.

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


Re: [OSM-talk] source=Google

2013-05-21 Thread Marc Gemis
On Tue, May 21, 2013 at 1:26 PM, Martin Koppenhoefer dieterdre...@gmail.com
 wrote:


 But in the end I think this whole source thing is completely
 overestimated. In the end the following mappers will compare what is on the
 map with what they know or believe to be there in reality, and in case of
 discrepancies will probably modify the map based on their findings,
 regardless of any source tag.



+1

if during a survey, you notice something different than what is in OSM, you
change it, no matter what's the source.
___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk


Re: [OSM-talk] source=Google

2013-05-21 Thread Henning Scholland

Hi,
maybe a little bit. For remote-mappers it's a good hint, if objects 
doesn't fit to imagery.


Henning


Am 21.05.2013 13:26, schrieb Martin Koppenhoefer:
But in the end I think this whole source thing is completely 
overestimated.



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


Re: [OSM-talk] source=Google

2013-05-21 Thread Peter Wendorff
Am 21.05.2013 13:30, schrieb Richard Fairhurst:
 Martin Koppenhoefer wrote:
 What do you propose to do with source tags found on an object when 
 you modify this object based on a different source?
 
 OSM has full object history. :)

...which is an argument that the source may fit well on the changeset,
too (while that's only true for changesets that rely on one source only).

regards
Peter

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


[OSM-talk] source=Google

2013-05-18 Thread malenki
There are abot 33.000 objects in OSM which have google in the one
way or another in their source tag:
http://malenki.ch/d/2013-05-18_142122_scr_source_google.png
Just type google in the value-field:
http://taginfo.openstreetmap.org/keys/source#values

Any thoughts about that?



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


Re: [OSM-talk] source=Google

2013-05-18 Thread pec...@gmail.com
http://wiki.openstreetmap.org/wiki/WikiProject_Haiti/Imagery_and_data_sources#Google_Imagery

This is one explanation. Also source=Google really says nothing that source
is Google Maps.

P.


2013/5/18 malenki o...@malenki.ch

 There are abot 33.000 objects in OSM which have google in the one
 way or another in their source tag:
 http://malenki.ch/d/2013-05-18_142122_scr_source_google.png
 Just type google in the value-field:
 http://taginfo.openstreetmap.org/keys/source#values

 Any thoughts about that?



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




-- 
mortigi tempo
Pēteris Krišjānis
___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk


Re: [OSM-talk] source=Google

2013-05-18 Thread Martin Koppenhoefer
2013/5/18 malenki o...@malenki.ch

 Any thoughts about that?




don't put source tags on objects but on the changeset?
If you are interested what the intended meaning is, you should ask the
contributors who added this, I guess different mappers used this for
different things, google is quite generic.

cheers,
Martin
___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk


Re: [OSM-talk] source=Google

2013-05-18 Thread Chris Hill

On 18/05/13 13:31, malenki wrote:

There are abot 33.000 objects in OSM which have google in the one
way or another in their source tag:
http://malenki.ch/d/2013-05-18_142122_scr_source_google.png
Just type google in the value-field:
http://taginfo.openstreetmap.org/keys/source#values

Any thoughts about that?



Recently I asked a new contributor to the area I routinely check about 
the fact that he had tagged source=google on a changeset for some paths 
he had added. It turned out that he had legitimately traced Bing 
imagery, but he had assumed it was Google - the only imagery he was 
familiar with. This is now burned into the history, forever giving the 
wrong impression.


My take on this is that _all_ of these dubious sources need 
investigating with the original editors. Any real transgressions should 
be redacted, but expect some genuine errors as I found. These need 
marking or recording to show they are not worth investigating again in 
the future.


--
Cheers, Chris
user: chillly


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


Re: [OSM-talk] source=Google

2013-05-18 Thread malenki
Martin Koppenhoefer wrote:

2013/5/18 malenki o...@malenki.ch

 Any thoughts about that?

don't put source tags on objects but on the changeset?

I see no difference in using source= on objects or changesets except
the visibility for the source


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


Re: [OSM-talk] source=Google

2013-05-18 Thread malenki
pec...@gmail.com wrote:

http://wiki.openstreetmap.org/wiki/WikiProject_Haiti/Imagery_and_data_sources#Google_Imagery

This is one explanation. 

As it seems for the value parts with the matching dates on them. 

Also source=Google really says nothing that source is Google Maps.

hopefully


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


Re: [OSM-talk] source=Google

2013-05-18 Thread Dave F.

On 18/05/2013 13:31, malenki wrote:

There are abot 33.000 objects in OSM which have google in the one
way or another in their source tag:
http://malenki.ch/d/2013-05-18_142122_scr_source_google.png
Just type google in the value-field:
http://taginfo.openstreetmap.org/keys/source#values

Any thoughts about that?


I use Google daily to map in OSM. I search their database for names  
websites of schools, restaurants etc. Nothing wrong in that. I don't tag 
the source as google, but others might.


IMO Source should be on the object, not on the changeset.

Dave F.

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


Re: [OSM-talk] source=Google

2013-05-18 Thread malenki
Dave F. wrote:

I use Google daily to map in OSM. I search their database for names  
websites of schools, restaurants etc. Nothing wrong in that. I don't
tag the source as google, but others might.

Since Google links to websites of the schools, restaurants etc I'd
consider it wrong to say google is the source. One could also say
source=internet or source=brain. :)

IMO Source should be on the object, not on the changeset.

+1 (except if there is one changeset for one object (; )


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


Re: [OSM-talk] source=Google

2013-05-18 Thread Yohan Boniface

On 05/18/2013 03:55 PM, malenki wrote:

Dave F. wrote:

IMO Source should be on the object, not on the changeset.


+1 (except if there is one changeset for one object (; )


This is not my opinion. Let's take a simple example: a school.

Some first user maps it from imagery, and so just draws the building.
So we have something like:
building=yes
source=Bing

Now someone who went on the place add the fact that this building is a 
school. So we have something like:

building=yes
amenity=school
name=Ecole de la Vie
source=Bing;survey

Now, the public administration release an open data file with many 
informations on every schools. Someone takes this info and add it to our 
school. So we have now, for example:

building=yes
amenity=school
name=Ecole de la vie
capacity=1800
wheelchair=yes
addr:street=Rue de la Vie
addr:housenumber=42
source=?

Now, we have the opportunity to fly a drone on the area, and so we have 
a very accurate imagery, which is great. So someone goes on our building 
way and move it a little bit, change a little bit the shape to better 
follow the reality, etc.


So the question is: now, what should be the source tag value? And which 
of every attributes or geo informations is covered by this value?


My personal answer is that the source should be on the changeset, which 
is where the data comes from.



Yohan



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


Re: [OSM-talk] source=Google

2013-05-18 Thread malenki
Yohan Boniface wrote:

On 05/18/2013 03:55 PM, malenki wrote:
 Dave F. wrote:
 IMO Source should be on the object, not on the changeset.

 +1 (except if there is one changeset for one object (; )

This is not my opinion. Let's take a simple example: a school.

Some first user maps it from imagery, and so just draws the building.
So we have something like:
building=yes
source=Bing

there is source:outline=* 
http://taginfo.osm.org/keys/source%3Aoutline
(used 10150 times so far)

Now someone who went on the place add the fact that this building is a 
school. So we have something like:
building=yes
amenity=school
name=Ecole de la Vie
source=Bing;survey

source=survey - since it has been surveyed

Now, the public administration release an open data file with many 
informations on every schools. Someone takes this info and add it to
our school. So we have now, for example:
building=yes
amenity=school
name=Ecole de la vie
capacity=1800
wheelchair=yes
addr:street=Rue de la Vie
addr:housenumber=42
source=?

there is source:addr=
http://taginfo.osm.org/keys/source%3Aaddr
(used 2015605 times so far)

Now, we have the opportunity to fly a drone on the area, and so we
have a very accurate imagery, which is great. So someone goes on our
building way and move it a little bit, change a little bit the shape
to better follow the reality, etc.

change source:outline

So the question is: now, what should be the source tag value? And
which of every attributes or geo informations is covered by this value?

My personal answer is that the source should be on the changeset,
which is where the data comes from.

Mine you see above.


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


Re: [OSM-talk] source=Google

2013-05-18 Thread Martin Koppenhoefer
2013/5/18 malenki o...@malenki.ch

 Martin Koppenhoefer wrote:

 2013/5/18 malenki o...@malenki.ch
 
  Any thoughts about that?
 
 don't put source tags on objects but on the changeset?

 I see no difference in using source= on objects or changesets except
 the visibility for the source




Point is it doesn't really belong in the data, because it is metadata. What
would you say if people tagged reason=why I did this edit on the objects?
Similar thing but maybe more obvious. The source is only refering to the
edit when the tag is added. The next mapper changing some property with a
different source (e.g. knowledge) has then to decide how to deal with this
source tag, and there really is no satisfying option to do so.

cheers,
Martin
___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk


Re: [OSM-talk] source=Google

2013-05-18 Thread Cartinus
Anybody else who noticed we already had this discussion last year ;)

http://lists.openstreetmap.org/pipermail/talk/2012-November/065034.html

On 05/18/2013 02:31 PM, malenki wrote:
 There are abot 33.000 objects in OSM which have google in the one
 way or another in their source tag:
 http://malenki.ch/d/2013-05-18_142122_scr_source_google.png
 Just type google in the value-field:
 http://taginfo.openstreetmap.org/keys/source#values
 
 Any thoughts about that?

---
m.v.g.,
Cartinus

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


Re: [OSM-talk] source=Google

2013-05-18 Thread Christoph Hormann
On Saturday 18 May 2013, Yohan Boniface wrote:
 On 05/18/2013 03:55 PM, malenki wrote:
  Dave F. wrote:
  IMO Source should be on the object, not on the changeset.
 
  +1 (except if there is one changeset for one object (; )

 This is not my opinion. Let's take a simple example: a school.

 [...]

The problem is currently neither changeset nor object tags are really a 
good solution for true metadata (that is information characterizing the 
data and not the real world object).

Changeset tags have mainly two problems:

- they always apply to the whole changeset so everything you map 
together needs to have the same metadata.  This might seem to be a 
problem primarily for imports but it can also be troublesome in manual 
mapping - imagine mapping something based on satellite images and you 
need to use different images for various parts due to clouds or even 
the common case of supplementing survey data with Bing images.

- many large objects are included in a lot of changesets without 
actually being substantially modified (like moving a single node in a 
500 node way etc.)  This means finding the actual changeset a certain 
geometry originates from to get the metadata information is not so 
easy.

The solution in my opinion would be to have separate metadata tags which 
are reset everytime a substantial change is made to the data they refer 
to unless the user explicitly sets them (either individually or for the 
whole changeset).  Geometry metadata tags for example would be reset 
if:

- in case of a node the node is moved
- in case of a way more than X percent of the nodes are changed (X being 
something like 30)
- in case of a multipolygon more than X percent of the ways are 
added/removed or substantially modified

This would not be fool proof of course (small changes could accumulate 
to a substantial change without being noticed).

Greetings,

-- 
Christoph Hormann
http://www.imagico.de/

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


Re: [OSM-talk] source=Google

2013-05-18 Thread Toby Murray
I will point out here that iD has invented a new changeset tag which I find
useful. It automatically records what imagery layers you use while editing
and throws them into an imagery_used=* tag. This removes the need for users
to manually tag source information if they are just tracing imagery. I
wouldn't mind seeing other editors adopt this convention.

Toby



On Sat, May 18, 2013 at 11:12 AM, Christoph Hormann chris_horm...@gmx.dewrote:

 On Saturday 18 May 2013, Yohan Boniface wrote:
  On 05/18/2013 03:55 PM, malenki wrote:
   Dave F. wrote:
   IMO Source should be on the object, not on the changeset.
  
   +1 (except if there is one changeset for one object (; )
 
  This is not my opinion. Let's take a simple example: a school.
 
  [...]

 The problem is currently neither changeset nor object tags are really a
 good solution for true metadata (that is information characterizing the
 data and not the real world object).

 Changeset tags have mainly two problems:

 - they always apply to the whole changeset so everything you map
 together needs to have the same metadata.  This might seem to be a
 problem primarily for imports but it can also be troublesome in manual
 mapping - imagine mapping something based on satellite images and you
 need to use different images for various parts due to clouds or even
 the common case of supplementing survey data with Bing images.

 - many large objects are included in a lot of changesets without
 actually being substantially modified (like moving a single node in a
 500 node way etc.)  This means finding the actual changeset a certain
 geometry originates from to get the metadata information is not so
 easy.

 The solution in my opinion would be to have separate metadata tags which
 are reset everytime a substantial change is made to the data they refer
 to unless the user explicitly sets them (either individually or for the
 whole changeset).  Geometry metadata tags for example would be reset
 if:

 - in case of a node the node is moved
 - in case of a way more than X percent of the nodes are changed (X being
 something like 30)
 - in case of a multipolygon more than X percent of the ways are
 added/removed or substantially modified

 This would not be fool proof of course (small changes could accumulate
 to a substantial change without being noticed).

 Greetings,

 --
 Christoph Hormann
 http://www.imagico.de/

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

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


Re: [OSM-talk] source=Google

2013-05-18 Thread Martin Koppenhoefer
2013/5/18 Christoph Hormann chris_horm...@gmx.de

 - imagine mapping something based on satellite images and you
 need to use different images for various parts due to clouds or even
 the common case of supplementing survey data with Bing images.



yes, this is very common, at least in regions with alternative high
resolution imagery, but it is similar for object and for changeset source
tags. For instance bing alone doesn't tell you which zoom level you used,
but different (high) zoom levels in Bing are based on different aerial
imagery (in my area there are years if not a decade in between z20 and z21
and they also have different offsets). I almost always use at least 2
aerial imagery providers (pcn and bing), drawing in one and positioning in
the other, in the comment I simply write ...and tracing from PCN2008 and
bing. Usually with aerial imagery from webmaps you also don't see from
when they are, at least almost nobody stores this information in the source
tag, but it is much more relevant (IMHO) to know traced from aerial
imagery from 2007 than traced from bing aerials
In the end it is almost pointless to compare different aerial imagery
layers with each other, what matters is reality and the best way to find
out is leave your desk and go out mapping ;-)



 - many large objects are included in a lot of changesets without
 actually being substantially modified (like moving a single node in a
 500 node way etc.)  This means finding the actual changeset a certain
 geometry originates from to get the metadata information is not so
 easy.



you will see from the changeset that only one node was moved (added or
deleted) and this will supposedly be based on your specified source,
generally moving a node will not create a new way version, but I get what
you intended (e.g. add or delete a node). IMHO this is a point for
associating the source to the changeset, as modifying it on the object
after adding this one node will tell the wrong story but not modifying the
source tag is neither desirable.


The solution in my opinion would be to have separate metadata tags which
 are reset everytime a substantial change is made to the data they refer
 to unless the user explicitly sets them (either individually or for the
 whole changeset).  Geometry metadata tags for example would be reset
 if:



Your mention of geometry metadata reminds me of another point: a simple
source is not enough, you'd need a distinct source tag for all properties
not one source for the whole object.

cheers,
Martin
___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk


Re: [OSM-talk] source=Google

2013-05-18 Thread Christoph Hormann
On Saturday 18 May 2013, Martin Koppenhoefer wrote:
 [...] Usually with aerial
 imagery from webmaps you also don't see from when they are, at least
 almost nobody stores this information in the source tag, but it is
 much more relevant (IMHO) to know traced from aerial imagery from
 2007 than traced from bing aerials

Much agreed, date of the images is much more relevant than the provider.  
The lack of this information in Bing etc. combined with the fact that 
first hand survey information is usually entered shortly after the 
survey has lead to a lack of practical need for entering this 
information.

 Your mention of geometry metadata reminds me of another point: a
 simple source is not enough, you'd need a distinct source tag for
 all properties not one source for the whole object.

I already had that in mind - i just explained it for the geometry only.  
Each tag could have its own metadata and this would be reset everytime 
the tag value is changed.

And of course in addition to 'source' one could think of various other 
metadata types.  The date is obvious but accuracy as well as 
reliability information could also be useful metadata.

-- 
Christoph Hormann
http://www.imagico.de/

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