Re: [OSM-talk] Mozambique's living streets

[john whelan]

In Lichinga, Niassa, Mozambique

http://www.openstreetmap.org/node/335126704#map=15/-13.3219/35.2649

practically every street is labeled highway=living_street including many
that look quite wide to me.  In other parts of the city the same type of
street when viewed in Bing is labeled residential.

I'm not talking narrow alleyways here.

Cheerio John

On 11 December 2016 at 11:29, Martin Koppenhoefer 
wrote:

>
> 2016-12-11 11:55 GMT+01:00 Tobias Zwick :
>
>> These kinds of streets are not only common in old town centers in Europe
>> but can be found in any place in the world that hadn't been built with
>> cars in mind, in shanty towns and also in modern (non-shanty) towns -
>> especially in Asia.
>>
>
>
> +1
>
>
>
>>
>> Now, the alley tag mentioned is problematic because there is no clear
>> definition what may still count as an alley and what should be a
>> residential road already.
>>
>
>
>
> For me it's basically an alley if it's too narrow to be a residential
> street. I'd expect the osm residential streets to be passable with a
> "normal" car plus a pedestrian, otherwise it would be an alley  or a path,
> or (if motor vehicles are not permitted), likely a footway or bikepath.
> Also if the turns are too sharp (corners) for a car it would not be a
> residential street. Typically the residential street will have width for 2
> cars to pass, plus pedestrians on each side.
>
> Cheers,
> Martin
>
> ___
> talk mailing list
> talk@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/talk
>
>
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Mozambique's living streets

[Martin Koppenhoefer]

2016-12-11 11:55 GMT+01:00 Tobias Zwick :

> These kinds of streets are not only common in old town centers in Europe
> but can be found in any place in the world that hadn't been built with
> cars in mind, in shanty towns and also in modern (non-shanty) towns -
> especially in Asia.
>


+1



>
> Now, the alley tag mentioned is problematic because there is no clear
> definition what may still count as an alley and what should be a
> residential road already.
>



For me it's basically an alley if it's too narrow to be a residential
street. I'd expect the osm residential streets to be passable with a
"normal" car plus a pedestrian, otherwise it would be an alley  or a path,
or (if motor vehicles are not permitted), likely a footway or bikepath.
Also if the turns are too sharp (corners) for a car it would not be a
residential street. Typically the residential street will have width for 2
cars to pass, plus pedestrians on each side.

Cheers,
Martin
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Mozambique's living streets

[john whelan]

Which icon is it?  I was never very good at icons.

Thanks John

On 11 December 2016 at 07:58, Matthijs Melissen 
wrote:

> On 11 December 2016 at 01:15, Jean-Marc Liotier  wrote:
> > But participants
> > in the thread were mostly experienced with French-speaking Africa - so I
> > cannot rule out the existence of a living street legal classification in
> > other locales.
>
> At least South Africa seems to have legal living_streets:
> https://en.wikipedia.org/wiki/Road_signs_in_South_Africa#/
> media/File:SADC_road_sign_R403.svg
>
> -- Matthijs
>
> ___
> talk mailing list
> talk@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/talk
>
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Mozambique's living streets

[Matthijs Melissen]

On 11 December 2016 at 01:15, Jean-Marc Liotier  wrote:
> But participants
> in the thread were mostly experienced with French-speaking Africa - so I
> cannot rule out the existence of a living street legal classification in
> other locales.

At least South Africa seems to have legal living_streets:
https://en.wikipedia.org/wiki/Road_signs_in_South_Africa#/media/File:SADC_road_sign_R403.svg

-- Matthijs

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

Re: [OSM-talk] Mozambique's living streets

[Martin Koppenhoefer]

sent from a phone

> On 11 Dec 2016, at 01:15, Jean-Marc Liotier  wrote:
> 
> The resulting consensus was that, while unpaved residential streets in Africa 
> are living streets in practice


which aspects of living streets? That you may only park in marked parking lots? 
That you'll always have to give way when coming from such a street? That 20km/h 
are more than legally allowed?

Cheers,
Martin 
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Mozambique's living streets

[Tobias Zwick]

> from minor ones... Some bright spark took initiative by tagging a large
> part of Dakar's residential streets as highway=service - that has been
> universally perceived as a very bad idea.
Well, there is this line in the wiki for highway=service and service=alley:

"In some medieval European settlements alleys may be the very narrow
streets which run in-between buildings, providing public through-access.
" ( http://wiki.openstreetmap.org/wiki/Tag:service%3Dalley )

These kinds of streets are not only common in old town centers in Europe
but can be found in any place in the world that hadn't been built with
cars in mind, in shanty towns and also in modern (non-shanty) towns -
especially in Asia.

Now, the alley tag mentioned is problematic because there is no clear
definition what may still count as an alley and what should be a
residential road already.

Actually I raised the same issue 3 years ago in the forum
https://forum.openstreetmap.org/viewtopic.php?id=21942 (with links to
example imagery).
The outcome, more or less was:
* do not use living_street for non-living-streets
* use path/footway for anything that does not fit a car anymore (i.e.
only a motorcycle/rickshaw)
* use highway=service + service=alley for residential alleys that barely
fit a car if convenient
* otherwise residential for everything else, use width/lanes to give a
hint how wide it is

(Not saying that I am absolutely content about it, it is still ambiguous)

The existence of these discussions about how to tag really small
residential roads popping up and the living_street tagging being popular
where there is no such (legal) thing as a living street again and again
is a sign that there is an unalleviated uncertainty and thus room for
conflict and disagreement about this, since it is not clearly defined.

Now, talking about this is good. But can we solve and clear this up here
in the interest of all mappers and correct the ambiguities in the wiki
definition once and for all?

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

Re: [OSM-talk] Mozambique's living streets

[Oleksiy Muzalyev]

On 11.12.16 01:15, Jean-Marc Liotier wrote:

On 12/10/2016 08:47 PM, john whelan wrote:
I just did a search on part of Mozambique and came across more than 
500 highway=living_street.


I always understood them to be a European concept highway with signs 
on the street and a very low max speed.  I wouldn't have expected to 
see so many clustered together in Mozambique.


Idle curiosity are they a legal entity in Mozambique and other parts 
of Africa?


A while ago 
(https://lists.openstreetmap.org/pipermail/talk-fr/2013-August/061580.html) 
I started a thread about usage of highway=living_street in Africa. The 
resulting consensus was that, while unpaved residential streets in 
Africa are living streets in practice, they are not legally classified 
that way - and therefore the correct tagging is highway=residential... 
But participants in the thread were mostly experienced with 
French-speaking Africa - so I cannot rule out the existence of a 
living street legal classification in other locales.




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


Another possibility is a combination:

highway=service plus service=alley

https://wiki.openstreetmap.org/wiki/Tag:service%3Dalley

for example here: 
http://www.openstreetmap.org/way/230413004#map=14/46.6890/31.1548


These are not really residential streets like in a city, which are a 
shown on the map with strong lines.



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

Re: [OSM-talk] Mozambique's living streets

[john whelan]

>Doesn't that feel like Africa already ?   dunno never been there.  It's
-13c tonight and -25c coming up shortly.  There is white stuff on the
ground.

> width and lanes keys could be used to provide info about smaller streets.
Pierre you're talking sense again.

I think we have highlighted the issue and that's really all we can do.
There is a difference between a local mapper with a tape measure and an
armchair mapper mapping from imagery for the first time without reading any
instructions but I can't put my finger on the critical difference.

I think we need boots on the ground to identify alleyways and very narrow
highways and Pierre's suggestion of them using width and lane tags makes a
lot of sense.

Thank you both for your thoughts.  If its currently warmer than -25c over
there I'm very tempted to pick up my tape measure and pop over to measure a
few street widths.

Cheerio John

On 10 December 2016 at 20:17, Pierre Béland <pierz...@yahoo.fr> wrote:

> width and lanes keys could be used to provide info about smaller streets.
>
>
> Pierre
>
>
> --
> *De :* Jean-Marc Liotier <j...@liotier.org>
> *À :* john whelan <jwhelan0...@gmail.com>
> *Cc :* OpenStreetMap talk mailing list <talk@openstreetmap.org>
> *Envoyé le :* samedi 10 décembre 2016 20h10
> *Objet :* Re: [OSM-talk] Mozambique's living streets
>
> Legal definition vs."duck typing" - there is a bit of both in
> Openstreetmap tagging... But because "living street" has a precise legal
> definition in Europe, involved mappers chose not to confuse the issue by
> using a different one.
>
> Then there was the question of differentiating large residential streets
> from minor ones... Some bright spark took initiative by tagging a large
> part of Dakar's residential streets as highway=service - that has been
> universally perceived as a very bad idea.
>
> I mostly just highway=residential with a surface=* tag but now that you
> remind me of those discussion, I'm thinking that a key differentiator is
> the presence of a sidewalk. What about:
>
> highway=residential
> surface=ground
> sidewalk=no
>
> Doesn't that feel like Africa already ?
>
>
> ___
> talk mailing list
> talk@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/talk
>
>
>
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Mozambique's living streets

[Pierre Béland]

width and lanes keys could be used to provide info about smaller streets.
  
Pierre 


  De : Jean-Marc Liotier <j...@liotier.org>
 À : john whelan <jwhelan0...@gmail.com> 
Cc : OpenStreetMap talk mailing list <talk@openstreetmap.org>
 Envoyé le : samedi 10 décembre 2016 20h10
 Objet : Re: [OSM-talk] Mozambique's living streets
   
Legal definition vs."duck typing" - there is a bit of both in 
Openstreetmap tagging... But because "living street" has a precise legal 
definition in Europe, involved mappers chose not to confuse the issue by 
using a different one.

Then there was the question of differentiating large residential streets 
from minor ones... Some bright spark took initiative by tagging a large 
part of Dakar's residential streets as highway=service - that has been 
universally perceived as a very bad idea.

I mostly just highway=residential with a surface=* tag but now that you 
remind me of those discussion, I'm thinking that a key differentiator is 
the presence of a sidewalk. What about:

highway=residential
surface=ground
sidewalk=no

Doesn't that feel like Africa already ?

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


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

Re: [OSM-talk] Mozambique's living streets

[Jean-Marc Liotier]

Legal definition vs."duck typing" - there is a bit of both in 
Openstreetmap tagging... But because "living street" has a precise legal 
definition in Europe, involved mappers chose not to confuse the issue by 
using a different one.


Then there was the question of differentiating large residential streets 
from minor ones... Some bright spark took initiative by tagging a large 
part of Dakar's residential streets as highway=service - that has been 
universally perceived as a very bad idea.


I mostly just highway=residential with a surface=* tag but now that you 
remind me of those discussion, I'm thinking that a key differentiator is 
the presence of a sidewalk. What about:


highway=residential
surface=ground
sidewalk=no

Doesn't that feel like Africa already ?

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

Re: [OSM-talk] Mozambique's living streets

[john whelan]

My gut feel is these have been mapped before the African highway wiki
http://wiki.openstreetmap.org/wiki/Highway_Tag_Africa guidelines were
written.  One batch was mapped in 2013.

If they were mapped on a HOT project that I was validating I'd probably
switch them to residential unless it was an experienced local mapper in
which case we might have a discussion.

I'm not sure what the implications of having them tagged living_street is.
I know highway=road is sometimes omitted by routing software as it is a
temporary type and you don't know if its a motorway or a footway with
steps.  Would routing software ignore them unless you were on foot?

I think its one of those things that needs one or two local mappers
gathered together for each country to check or make a decision on but
locating them and getting them to take ownership is difficult.  Certainly
sitting in Canada looking at imagery I wouldn't want to make the call when
it might be a local mapper and they might be correct.

The other approach is if there are no legally defined living_streets in
Africa or even a list of countries where they are not legally defined then
switch them to residential. Many residential streets have football games
etc played in the middle if the traffic isn't heavy in many parts of the
world.  I certainly remember telling a copper who thought we shouldn't be
playing rounders in the street. Why not its our street.  He took himself
off.

Perhaps the African highway wiki needs to be extended to cover them?

Thoughts?

Cheerio John

On 10 December 2016 at 19:15, Jean-Marc Liotier  wrote:

> On 12/10/2016 08:47 PM, john whelan wrote:
>
> I just did a search on part of Mozambique and came across more than 500
> highway=living_street.
>
> I always understood them to be a European concept highway with signs on
> the street and a very low max speed.  I wouldn't have expected to see so
> many clustered together in Mozambique.
>
> Idle curiosity are they a legal entity in Mozambique and other parts of
> Africa?
>
>
> A while ago (https://lists.openstreetmap.org/pipermail/talk-fr/2013-
> August/061580.html) I started a thread about usage of
> highway=living_street in Africa. The resulting consensus was that, while
> unpaved residential streets in Africa are living streets in practice, they
> are not legally classified that way - and therefore the correct tagging is
> highway=residential... But participants in the thread were mostly
> experienced with French-speaking Africa - so I cannot rule out the
> existence of a living street legal classification in other locales.
>
>
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Mozambique's living streets

[Jean-Marc Liotier]

On 12/10/2016 08:47 PM, john whelan wrote:
I just did a search on part of Mozambique and came across more than 
500 highway=living_street.


I always understood them to be a European concept highway with signs 
on the street and a very low max speed.  I wouldn't have expected to 
see so many clustered together in Mozambique.


Idle curiosity are they a legal entity in Mozambique and other parts 
of Africa?


A while ago 
(https://lists.openstreetmap.org/pipermail/talk-fr/2013-August/061580.html) 
I started a thread about usage of highway=living_street in Africa. The 
resulting consensus was that, while unpaved residential streets in 
Africa are living streets in practice, they are not legally classified 
that way - and therefore the correct tagging is highway=residential... 
But participants in the thread were mostly experienced with 
French-speaking Africa - so I cannot rule out the existence of a living 
street legal classification in other locales.


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

[OSM-talk] Mozambique's living streets

[john whelan]

I just did a search on part of Mozambique and came across more than 500
highway=living_street.

I always understood them to be a European concept highway with signs on the
street and a very low max speed.  I wouldn't have expected to see so many
clustered together in Mozambique.

Idle curiosity are they a legal entity in Mozambique and other parts of
Africa?

Does the routing software commonly exclude them from routes? as I would
expect.

Thanks John
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk