Waterway length

classic Classic list List threaded Threaded
29 messages Options
12
Reply | Threaded
Open this post in threaded view
|

Waterway length

Eugene Podshivalov
Hi all,
The relation:waterway wiki page recommends using "distance" tag for "the total length of river in km". Was there any discussion of this choice?
It seems a bit incorrect and confusing, because "distance" is more suitable for routes as discribed on its proper page. The existing "length" tag would fit better, woudn't it?

Cheers,
Eugene

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Waterway length

Eugene Podshivalov
André, that's correct but do you happen to know why "distance" was selected for route and waterway length then?

Cheers,
Eugene

вт, 29 янв. 2019 г. в 22:41, André Pirard <[hidden email]>:
On 2019-01-29 16:37, Eugene Podshivalov wrote:
Hi all,
The relation:waterway wiki page recommends using "distance" tag for "the total length of river in km". Was there any discussion of this choice?
It seems a bit incorrect and confusing, because "distance" is more suitable for routes as discribed on its proper page. The existing "length" tag would fit better, woudn't it?

Cheers,
Eugene
A distance is the length of the straight line between two points or, for that matter, the two ends of any line.
They can be far or near.
The length of a line is the number of meters that your odometer measures if you follow every bend of it.
Mathematically, it is the sum of the lengths of tiny slices of the line that can be taken as straight (and the shorter the straighter, that's "infinitesimal analysis" but don't tell anybody).
That is what is called "traveling a long road".
So, to go from one point to another, you may follow two routes one of which is longer than the other and, yet, you won't have moved any farther (no more distantly if I may say).

So, it is a mistake to talk of the distance of a route, it is a length.
And be it macadam, water, rail or cycling or footing etc., it doesn't change anything to that.

All the best,

André.


_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Waterway length

Graeme Fitzpatrick

On Sat, 16 Feb 2019 at 01:38, Eugene Podshivalov <[hidden email]> wrote:
André, that's correct but do you happen to know why "distance" was selected for route and waterway length then?

No idea why, but rivers should certainly be shown with a length.

EG The Nile rises in Lake Victoria & travels to the Mediterranean with a length of ~6695km (depending on reference used), but the distance between Kampala, on the north shore of Lave Victoria & Cairo is only 3300km.

OSM shows it fairly accurately as 6853km but as a distance, when it should clearly be a length - may be a re-write required?

Thanks

Graeme

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Waterway length

Eugene Podshivalov
The use of "distance" for river length distracts me as well.
But I'm trying to find its origin on this wiki page
https://wiki.openstreetmap.org/wiki/Relation:waterway
distance* (optional)Total length of river in km
Cheers,
Eugene

сб, 16 февр. 2019 г. в 02:33, Graeme Fitzpatrick <[hidden email]>:

On Sat, 16 Feb 2019 at 01:38, Eugene Podshivalov <[hidden email]> wrote:
André, that's correct but do you happen to know why "distance" was selected for route and waterway length then?

No idea why, but rivers should certainly be shown with a length.

EG The Nile rises in Lake Victoria & travels to the Mediterranean with a length of ~6695km (depending on reference used), but the distance between Kampala, on the north shore of Lave Victoria & Cairo is only 3300km.

OSM shows it fairly accurately as 6853km but as a distance, when it should clearly be a length - may be a re-write required?

Thanks

Graeme
_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Waterway length

André Pirard-2
On 2019-02-16 00:41, Eugene Podshivalov wrote:
The use of "distance" for river length distracts me as well.
But I'm trying to find its origin on this wiki page
https://wiki.openstreetmap.org/wiki/Relation:waterway
distance * (optional) Total length of river in km
не знаю, почему это так
это как, it's just like
speed
* (optional) height of the  building
it should be changed as well (re-written)
and other distances for lengths if there is
Cheers,
Eugene

сб, 16 февр. 2019 г. в 02:33, Graeme Fitzpatrick <[hidden email]>:

On Sat, 16 Feb 2019 at 01:38, Eugene Podshivalov <[hidden email]> wrote:
André, that's correct but do you happen to know why "distance" was selected for route and waterway length then?

No idea why, but rivers should certainly be shown with a length.

EG The Nile rises in Lake Victoria & travels to the Mediterranean with a length of ~6695km (depending on reference used), but the distance between Kampala, on the north shore of Lave Victoria & Cairo is only 3300km.

OSM shows it fairly accurately as 6853km but as a distance, when it should clearly be a length - may be a re-write required?

Thanks

Graeme
_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging


_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging


_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Waterway length

Andy Mabbett
In reply to this post by Graeme Fitzpatrick
On Fri, 15 Feb 2019 at 23:31, Graeme Fitzpatrick <[hidden email]> wrote:

> The Nile rises in Lake Victoria & travels to the Mediterranean with
> a length of ~6695km (depending on reference used), but the distance
> between Kampala, on the north shore of Lave Victoria & Cairo is only
> 3300km.

Why would we tag either, when software can calculate them?

--
Andy Mabbett
@pigsonthewing
http://pigsonthewing.org.uk

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Waterway length

Eugene Podshivalov
сб, 16 февр. 2019 г. в 16:30, Andy Mabbett <[hidden email]>: 
Why would we tag either, when software can calculate them? 
Calculated value may differ from the official one and is error-prone, e.g. when a segment is deleted by mistake or when a segment role is changed between main and side stream.

Cheers,
Eugene

сб, 16 февр. 2019 г. в 16:30, Andy Mabbett <[hidden email]>:
On Fri, 15 Feb 2019 at 23:31, Graeme Fitzpatrick <[hidden email]> wrote:

> The Nile rises in Lake Victoria & travels to the Mediterranean with
> a length of ~6695km (depending on reference used), but the distance
> between Kampala, on the north shore of Lave Victoria & Cairo is only
> 3300km.

Why would we tag either, when software can calculate them?

--
Andy Mabbett
@pigsonthewing
http://pigsonthewing.org.uk

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Waterway length

Sergio Manzi
On 2019-02-16 14:46, Eugene Podshivalov wrote:
Calculated value may differ from the official one ...

Official according to whom?

From https://en.wikipedia.org/wiki/Nile#cite_note-length-1 :

"The length of the Nile is usually said to be about 6,650 km (4,130 mi), but reported values lie anywhere between 5,499 km (3,417 mi) and 7,088 km (4,404 mi). The length measurements of many rivers are only approximations and differ from each other because there are many factors that determine the calculated river length, such as the position of the geographical source and the mouth, the scale of measurement, and the length measuring techniques"

From https://en.wikipedia.org/wiki/Amazon_River :

"The Amazon River ... in South America is the largest river by discharge volume of water in the world, and by some definitions it is the longest." (my emphasis)

Encyclopedias are the preferred source for this kind of information, not maps (or GIS databases).

Sergio


_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging

smime.p7s (4K) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: Waterway length

Daniel Koć
W dniu 16.02.2019 o 15:00, Sergio Manzi pisze:
> On 2019-02-16 14:46, Eugene Podshivalov wrote:
>> Calculated value may differ from the official one ...
>
> Official according to whom?
>
Good question - who can we trust?

The question is also how to calculate them? If a river has a fork (or
even multiple forks in its delta), what should we do - add all their
length (the simplest code would do it using database entries), choose
one (which and how?), make an average or maybe something else?...

Which also means we have to know exactly which parts carry the same
name, which is also not always easy to determine, since most of the
waters on Earth are connected and people just name some of them, but not
in a precise way and the name is not visible on the surface.

So adding official length tag makes sense for me, just like the official
population or borders, however source is needed then.


--
"I see dead people" [Sixth Sense]



_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Waterway length

Andy Mabbett
In reply to this post by Eugene Podshivalov
On Sat, 16 Feb 2019 at 13:46, Eugene Podshivalov <[hidden email]> wrote:
>
> сб, 16 февр. 2019 г. в 16:30, Andy Mabbett <[hidden email]>:

>> Why would we tag either, when software can calculate them?

> Calculated value may differ from the official one and is error-prone

I would suggest that values entered by human mappers are more likely
to be "error prone"; and that we should be more concerned with
on-the-ground reality than "offical" figures.

--
Andy Mabbett
@Pigsonthewing http://pigsonthewing.org.uk

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Waterway length

Markus-5
In reply to this post by Eugene Podshivalov
On Sat, 16 Feb 2019, 00:43 Eugene Podshivalov <[hidden email] wrote:
The use of "distance" for river length distracts me as well.
But I'm trying to find its origin on this wiki page
https://wiki.openstreetmap.org/wiki/Relation:waterway
distance* (optional)Total length of river in km

This has been added more than four years ago, see:


Regards

Markus



_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Waterway length

Markus-5
In reply to this post by Andy Mabbett
On Sat, 16 Feb 2019, 15:26 Andy Mabbett <[hidden email] wrote:
I would suggest that values entered by human mappers are more likely
to be "error prone"; and that we should be more concerned with
on-the-ground reality than "offical" figures.

I agree. Besides, official figures may not be compatible with OSM's licence.

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Waterway length

Eugene Podshivalov
In reply to this post by Markus-5
What is the best way to correct this, so that all other langauge pages got the correction as well?

Cheers,
Eugene

сб, 16 февр. 2019 г. в 21:57, Markus <[hidden email]>:
On Sat, 16 Feb 2019, 00:43 Eugene Podshivalov <[hidden email] wrote:
The use of "distance" for river length distracts me as well.
But I'm trying to find its origin on this wiki page
https://wiki.openstreetmap.org/wiki/Relation:waterway
distance* (optional)Total length of river in km

This has been added more than four years ago, see:


Regards

Markus


_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Waterway length

André Pirard-2
In reply to this post by Markus-5
On Sat, 16 Feb 2019, 15:26 Andy Mabbett <[hidden email] wrote:
I would suggest that values entered by human mappers are more likely
to be "error prone"; and that we should be more concerned with
on-the-ground reality than "offical" figures.

It's easy to make a script to total up all the segments of a waterway or any way.
It would be already done if the routers added "by boat" to my suggested "by plane" ("as the crow flies").
I suggested that real routes can be made of multiple segments of those various kinds.
Like a car travel followed by a hike including a canoe trip.
Like the Tour de France (some parts by plane).
But I seem like not well understood or maybe seen as eccentric.

All the best,

André.


_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Waterway length

Sergio Manzi

TBH, I'm all with you (and maybe I'm seen as an eccentric too...) and I see the tagging of waterways length as egregiously useless.

Beside, I smell a lack of verifiability [1] in this waterways property: I'm not a geographer, by far, but in the years I made up my mind that this is one of those things that experts debate in their congresses and can be object of accademic thesis.

On the other hand a tool to compute the total length of waterways from their source (once you have pinpointed it) to their mouth (again, once you have pinpointed it), could be really interesting.

A static value for a river length in OSM, without any information about its source (pun not intended, but valid anyway...), is the kind of information I will usually route to /dev/null and for which I'll seek more autortiative sources (again...)

Sergio


[1] https://wiki.openstreetmap.org/wiki/Verifiability

On 2019-02-16 22:10, André Pirard wrote:
It's easy to make a script to total up all the segments of a waterway or any way.
...
But I seem like not well understood or maybe seen as eccentric.

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging

smime.p7s (4K) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: Waterway length

Markus-5
In reply to this post by Eugene Podshivalov
On Sat, 16 Feb 2019, 20:06 Eugene Podshivalov <[hidden email] wrote:
What is the best way to correct this, so that all other langauge pages got the correction as well?

I'm not aware of any other way than correcting it on each page. I've just done this and also added a note that this tag lacks verifiability.

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Waterway length

Eugene Podshivalov
Thank you, Markus.

Cheers,
Eugene

вс, 17 февр. 2019 г. в 01:01, Markus <[hidden email]>:
On Sat, 16 Feb 2019, 20:06 Eugene Podshivalov <[hidden email] wrote:
What is the best way to correct this, so that all other langauge pages got the correction as well?

I'm not aware of any other way than correcting it on each page. I've just done this and also added a note that this tag lacks verifiability.
_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Waterway length

marc marc
In reply to this post by Sergio Manzi
Le 16.02.19 à 22:32, Sergio Manzi a écrit :
> A static value for a river length in OSM, without any information about
> its source

every tag you add into osm have a changeset with a source tag, isn't it?
so adding the lenght should/must also have a source (extrapolation (sum
of all way of a relation) of osm data is a source)

a few month ago, I have checked the length of Rhône [1]
the french wikipedia list 2 sources for the lenght... both are very fair
away of the lenght found after some work on osm data.
which one to choose? osm without hesitation. maybe it is not fair but at
least it is verifiable (everyone can load the relationship, see the
result and correct errors if necessary) while the other 2 sources
(including the official French source) are totally unverifiable.

unfortunately I did not send in osm the result of the cleaning because
it concerned partly errors in osm (mainly roles in the relationship)
but I started by purging everything that didn't interest me in the
relationship before fixing. it will have to be done again

[1] https://en.wikipedia.org/wiki/Rh%C3%B4ne
_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Waterway length

Sergio Manzi
Then why don't you submit a paper to the CNFG (http://www.cnfg.fr/) and correct the Wikipadia articles?

Sergio


On 2019-02-16 23:07, marc marc wrote:

> Le 16.02.19 à 22:32, Sergio Manzi a écrit :
>> A static value for a river length in OSM, without any information about
>> its source
> every tag you add into osm have a changeset with a source tag, isn't it?
> so adding the lenght should/must also have a source (extrapolation (sum
> of all way of a relation) of osm data is a source)
>
> a few month ago, I have checked the length of Rhône [1]
> the french wikipedia list 2 sources for the lenght... both are very fair
> away of the lenght found after some work on osm data.
> which one to choose? osm without hesitation. maybe it is not fair but at
> least it is verifiable (everyone can load the relationship, see the
> result and correct errors if necessary) while the other 2 sources
> (including the official French source) are totally unverifiable.
>
> unfortunately I did not send in osm the result of the cleaning because
> it concerned partly errors in osm (mainly roles in the relationship)
> but I started by purging everything that didn't interest me in the
> relationship before fixing. it will have to be done again
>
> [1] https://en.wikipedia.org/wiki/Rh%C3%B4ne
> _______________________________________________
> Tagging mailing list
> [hidden email]
> https://lists.openstreetmap.org/listinfo/tagging

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging

smime.p7s (4K) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: Waterway length

Sergio Manzi

Sorry for the typo: of course Wikipadia was meant to be Wikipedia!

On 2019-02-16 23:15, Sergio Manzi wrote:
Then why don't you submit a paper to the CNFG (http://www.cnfg.fr/) and correct the Wikipadia articles?

Sergio


On 2019-02-16 23:07, marc marc wrote:
Le 16.02.19 à 22:32, Sergio Manzi a écrit :
A static value for a river length in OSM, without any information about 
its source
every tag you add into osm have a changeset with a source tag, isn't it?
so adding the lenght should/must also have a source (extrapolation (sum 
of all way of a relation) of osm data is a source)

a few month ago, I have checked the length of Rhône [1]
the french wikipedia list 2 sources for the lenght... both are very fair 
away of the lenght found after some work on osm data.
which one to choose? osm without hesitation. maybe it is not fair but at 
least it is verifiable (everyone can load the relationship, see the 
result and correct errors if necessary) while the other 2 sources 
(including the official French source) are totally unverifiable.

unfortunately I did not send in osm the result of the cleaning because 
it concerned partly errors in osm (mainly roles in the relationship)
but I started by purging everything that didn't interest me in the 
relationship before fixing. it will have to be done again

[1] https://en.wikipedia.org/wiki/Rh%C3%B4ne
_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging

    

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging

smime.p7s (4K) Download Attachment
12