Re: [OSM-talk] Name:* tags in the local language

[Yuri Astrakhan]

Janko, while I agree that it is possible in theory, I am not sure this is a
big problem. After all, if you are downloading a single region, you are
much more likely to know more about that region, and not need any
defaults.  The problem solved by the default_language is fairly specific to
planet-level maps, where there are too many languages to manage
realistically.   I think that adding and managing this tag on every
possible region (tens of thousands?), just in case someone somewhere might
need it seems like a bad idea. Appending it dynamically on download might
be ok, but again - there has to be someone whom this benefit in a
significant way, and this has to be a simpler solution than alternatives.

One option BTW might be to create a service that produces a map of defaults
- than it would be a relatively small download, allowing users to convert
any geopoint into a language.


On Sat, May 5, 2018 at 2:25 AM Janko Mihelić  wrote:

> sub, 5. svi 2018. u 00:18 Yuri Astrakhan 
> napisao je:
>
>>
>> Tag description:
>> https://wiki.openstreetmap.org/wiki/Key:default_language
>>
>
> I like it overall. I'm not sure about this one:
> *Do not set it on any smaller sub-regions unless their default language is
> different.*
>
> I agree that is cleaner, but what if a data consumer only downloads one US
> state, how will it know the default language? Maybe there are some other
> ways I'm not aware of, like filling in the gaps in data before publishing
> derived maps.
>
> If this is indeed a problem for data consumers, I would set the tag on all
> subregions up until a sensible level. We can see the smallest size of a
> region some applications offer for download.
>
> Janko
>
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Name:* tags in the local language

[Janko Mihelić]

sub, 5. svi 2018. u 00:18 Yuri Astrakhan  napisao
je:

>
> Tag description:
> https://wiki.openstreetmap.org/wiki/Key:default_language
>

I like it overall. I'm not sure about this one:
*Do not set it on any smaller sub-regions unless their default language is
different.*

I agree that is cleaner, but what if a data consumer only downloads one US
state, how will it know the default language? Maybe there are some other
ways I'm not aware of, like filling in the gaps in data before publishing
derived maps.

If this is indeed a problem for data consumers, I would set the tag on all
subregions up until a sensible level. We can see the smallest size of a
region some applications offer for download.

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

Re: [OSM-talk] Name:* tags in the local language

[Yuri Astrakhan]

I have set this value on some of the more obvious cases in N & S Americas.
I have also created a wiki page describing the tag. Any help with this is
greatly appreciated, especially if you have local knowledge about
subregions!

Tag description:
https://wiki.openstreetmap.org/wiki/Key:default_language

P.S. Janko, please take a look at the single vs multiple languages per
region in that wiki page.  Does that make sense?


On Wed, May 2, 2018 at 1:01 PM Janko Mihelić  wrote:

> pon, 30. tra 2018. u 08:28 Yuri Astrakhan 
> napisao je:
>
>>
>> "official_language" is not a good tag name because it does not match the
>> meaning, e.g. the official languages of Canada are both en & fr, but
>> "name"
>> tag is always in English except for Quebec, where it is in French.
>>
>
>  I agree that "official_language" has a much too restrictive meaning. It
> will bury us in bureaucracy of "what is actually official".
> "default_language" is a bit vague, but maybe a better fit to solve this
> problem.
>
> So just look at a region, see in what language >90% of the labels are, and
> add default_language=*. It's not going to be 100% accurate, but infinitely
> better then nothing. Than try to get closer to 100% by adding
> default_language to subregions, and in the end, individual objects.
>
> I like that approach.
>
> Janko
>
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Name:* tags in the local language

[Janko Mihelić]

pon, 30. tra 2018. u 08:28 Yuri Astrakhan  napisao
je:

>
> "official_language" is not a good tag name because it does not match the
> meaning, e.g. the official languages of Canada are both en & fr, but
> "name"
> tag is always in English except for Quebec, where it is in French.
>

 I agree that "official_language" has a much too restrictive meaning. It
will bury us in bureaucracy of "what is actually official".
"default_language" is a bit vague, but maybe a better fit to solve this
problem.

So just look at a region, see in what language >90% of the labels are, and
add default_language=*. It's not going to be 100% accurate, but infinitely
better then nothing. Than try to get closer to 100% by adding
default_language to subregions, and in the end, individual objects.

I like that approach.

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

Re: [OSM-talk] Name:* tags in the local language

[Yuri Astrakhan]

Multiple semicolon-separated values do not solve the main problem -
figuring out the language of the "name" tag. If a region uses one value in
the name tag, "default_language" should be set to just one language.

If the whole region uses "xx - yy" convention in the name tag,
default_language could be set to "xx - yy" -- allowing tools to parse
name tag into two languages (although this would be an error prone method).

"official_language" is not a good tag name because it does not match the
meaning, e.g. the official languages of Canada are both en & fr, but "name"
tag is always in English except for Quebec, where it is in French.

Are there any objections to this (fuzzy) approach? Should the tag be called
something else?

* Use the largest possible admin region to set the "default_language" tag
to a single language code.  "default_language"Z does not mean the official
language of the region. It only specifies the language of the "name" tag.
* A region may contain a sub-region with a different default_language.
* If a region uses mixed languages in all of its name tags, eg. "[name in
en] - [name in zh]", set default_language="en - zh".  Try to keep it to a
somewhat parsable value to help data consumers.
* In some rare cases, additional non-admin regions might be required for
the default_language.  Try to avoid it if possible.


On Thu, Apr 26, 2018 at 5:41 PM Daniel Koć  wrote:

> W dniu 26.04.2018 o 15:36, Martin Koppenhoefer pisze:
>
>
> 2018-04-26 14:16 GMT+02:00 Daniel Koć :
>
>>
>> Isn't it like this:
>>
>> Country Belgium - official_language=de;fr;nl
>> Region Brussels-Capital - official_language=fr;nl
>> City Eupen - official_language=de
>>
>> What would be wrong with this scheme?
>
>
>
> it is only about "official languages" and it would somehow imply we would
> not want names added through ground truth for cases where the language the
> name is in, would not be recognized as an official language.
>
>
> Sure, that's why I suggested common_language=* (common_language=xx +
> name:xx=* is just like saying "name=* is in xx").
>
> Could you explain this problem using some examples?
>
> I also don't know what this would imply for areas without formal
> government / disputed areas. Whose "official" language would we tag?
>
>
> That's interesting case. How do we tag the borders for such areas?
>
> If countries/regions with known official_language=* are overlapping, the
> language would be known for both and you have to choose one or show both
> (the same as official_language=xx;yy).
>
> Another solution would be to use some special values, like "none" or
> "disputed" for this area (unfortunately "no" is a code for Norwegian
> language).
>
> --
> "My method is uncertain/ It's a mess but it's working" [F. Apple]
>
> ___
> 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] Name:* tags in the local language

[Daniel Koć]

W dniu 26.04.2018 o 15:36, Martin Koppenhoefer pisze:
>
> 2018-04-26 14:16 GMT+02:00 Daniel Koć  >:
>
>
> Isn't it like this:
>
> Country Belgium - official_language=de;fr;nl
> Region Brussels-Capital - official_language=fr;nl
> City Eupen - official_language=de
>
> What would be wrong with this scheme?
>
>
>
> it is only about "official languages" and it would somehow imply we
> would not want names added through ground truth for cases where the
> language the name is in, would not be recognized as an official language.

Sure, that's why I suggested common_language=* (common_language=xx +
name:xx=* is just like saying "name=* is in xx").

Could you explain this problem using some examples?

> I also don't know what this would imply for areas without formal
> government / disputed areas. Whose "official" language would we tag?

That's interesting case. How do we tag the borders for such areas?

If countries/regions with known official_language=* are overlapping, the
language would be known for both and you have to choose one or show both
(the same as official_language=xx;yy).

Another solution would be to use some special values, like "none" or
"disputed" for this area (unfortunately "no" is a code for Norwegian
language).

-- 
"My method is uncertain/ It's a mess but it's working" [F. Apple]

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

Re: [OSM-talk] Name:* tags in the local language

[Daniel Koć]

W dniu 26.04.2018 o 14:32, Philip Barnes pisze:
> If a place in England should we assume its name is in English?
>
> Name:en=Llanymynech would be a very odd assumption. As would Hengoed
> or Rhydycroesau.
>
> This cannot be automatic, it needs mappers with local knowledge.

That's pretty sane /general/ assumption, but rules can have /specific/
corner cases, like these. Note that nobody has added name:en=Llanymynech
- it's only name=Llanymynech, see:

https://www.openstreetmap.org/node/29750244

In this case local knowledge is probably to *not add* name:en=*. The
data consumer has no "en" value to render (let's talk about rendering
for example), so she can fall back to just name=* value - or just skip
it, if she wants to show only English names (why not?).

What I propose is to have some general assumptions, but in specific
cases these can be overriden (like official_name=cy for example) or
ommited (if not applicable - for example we don't know the language or
we don't have time to add so specific data and name=* is enough).

-- 
"My method is uncertain/ It's a mess but it's working" [F. Apple]

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

Re: [OSM-talk] Name:* tags in the local language

[Martin Koppenhoefer]

2018-04-26 14:16 GMT+02:00 Daniel Koć :

>
> Isn't it like this:
>
> Country Belgium - official_language=de;fr;nl
> Region Brussels-Capital - official_language=fr;nl
> City Eupen - official_language=de
>
> What would be wrong with this scheme?



it is only about "official languages" and it would somehow imply we would
not want names added through ground truth for cases where the language the
name is in, would not be recognized as an official language.
I also don't know what this would imply for areas without formal government
/ disputed areas. Whose "official" language would we tag?

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

Re: [OSM-talk] Name:* tags in the local language

[Marc Gemis]

Discussing this in parallel on our Riot channel. The decision was made
before I joined OSM (thus before 2011).
Apparently for the renderer, (as a street name with a semi-colon looks
weird to non-insiders I guess).

regards

m

On Thu, Apr 26, 2018 at 3:05 PM, Daniel Koć  wrote:
> W dniu 26.04.2018 o 14:49, Marc Gemis pisze:
>> The name for the country in the name tag is " België / Belgique / Belgien" 
>> (*)
>> The name for any street in Brussels is either " - "
>> or " - " with the majority mapped with French in
>> front.
>>
>> The names never use a semi-colon. Without looking at name:fr / name:nl
>> tags you don't know which part of the name belongs to which language
>> (I think).
>> For Belgium there is no problem, as the names will only contain latin
>> characters, but in other cases it might become difficult, not ?
>
> 1. In database we store values, not typographic conventions, that's why
> semicolon to separate multiple values.
>
> 2. The data consumer can decide what type of separator she wants to use.
> It's data presentation part, not data storage. One can decide to show "
> België - Belgique - Belgien" for example or fall back to name and so on.
> The same with streets: one may always render " - " or
> "/" or whatever else (like one in bold and the second
> one in italics).
>
> --
> "My method is uncertain/ It's a mess but it's working" [F. Apple]
>
>
>
> ___
> 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] Name:* tags in the local language

[Daniel Koć]

W dniu 26.04.2018 o 14:49, Marc Gemis pisze:
> The name for the country in the name tag is " België / Belgique / Belgien" (*)
> The name for any street in Brussels is either " - "
> or " - " with the majority mapped with French in
> front.
>
> The names never use a semi-colon. Without looking at name:fr / name:nl
> tags you don't know which part of the name belongs to which language
> (I think).
> For Belgium there is no problem, as the names will only contain latin
> characters, but in other cases it might become difficult, not ?

1. In database we store values, not typographic conventions, that's why
semicolon to separate multiple values.

2. The data consumer can decide what type of separator she wants to use.
It's data presentation part, not data storage. One can decide to show "
België - Belgique - Belgien" for example or fall back to name and so on.
The same with streets: one may always render " - " or
"/" or whatever else (like one in bold and the second
one in italics).

-- 
"My method is uncertain/ It's a mess but it's working" [F. Apple]



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

Re: [OSM-talk] Name:* tags in the local language

[Marc Gemis]

The name for the country in the name tag is " België / Belgique / Belgien" (*)
The name for any street in Brussels is either " - "
or " - " with the majority mapped with French in
front.

The names never use a semi-colon. Without looking at name:fr / name:nl
tags you don't know which part of the name belongs to which language
(I think).
For Belgium there is no problem, as the names will only contain latin
characters, but in other cases it might become difficult, not ?

m.

(*) I wonder whether the separator was always "/" here.

On Thu, Apr 26, 2018 at 2:16 PM, Daniel Koć  wrote:
> W dniu 26.04.2018 o 13:53, Marc Gemis pisze:
>> Do you now assume that names in region B outside city C have a
>> namehr;nameit in the name tag?
>
> Yes, unless stated otherwise for cities E, F and G.
>
> To be clear: I assume only that if they name:hr=* and name:it=*, both
> are official, but that doesn't mean both need to be tagged (one or more
> name tag may be just missing).
>
>> That's not how we have done it for Belgium or any street in Brussels.
>
> Isn't it like this:
>
> Country Belgium - official_language=de;fr;nl
> Region Brussels-Capital - official_language=fr;nl
> City Eupen - official_language=de
>
> What would be wrong with this scheme?
>
> --
> "My method is uncertain/ It's a mess but it's working" [F. Apple]
>
>
>
> ___
> 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] Name:* tags in the local language

[Philip Barnes]

Placenames are often corruptions of old languages or are in a different 
language to that spoken and not everywhere has an official language.

If a place in England should we assume its name is in English?

Name:en=Llanymynech would be a very odd assumption. As would Hengoed or 
Rhydycroesau.

This cannot be automatic, it needs mappers with local knowledge.

Phil (trigpoint) 

On 26 April 2018 12:35:10 BST, "Daniel Koć"  wrote:
>W dniu 26.04.2018 o 11:51, Mateusz Konieczny pisze:
>> It is not handling
>> - regions with more than one widespread language
>> - features that have name tag in an atypical language
>
>In my opinion there's no single solution for atypical cases, yet it's
>sane to start with defaults. It can be used on many levels and region
>can have more than one language defined, just like Janjko has shown:
>
>Country A - official_language=hr
>Region B - official_language=hr;it
>City C - official_language=it
>
>When a data consumer tries to find what is the official language for
>street D, she finds that city C uses "it", so it's enough to know that
>the street should be also in "it". Speaking of regions this is the
>proper handling in my opinion - in region B you can for example use
>both
>languages or choose the one you prefer, it's up to you.
>
>But we don't have to stop with level 2, 4, or 6. If - for some reason -
>part of the city is different and we have no way to show the borders of
>the area, one can add official_language=* for the streets or objects in
>this particular area. Of course instead of official_language we may use
>default_language=* or common_language=* tag or something similar, the
>same as the users have to choose common name for name=*.
>
>This way we can always assume some language for the (administrative)
>area, but if it's not true locally, we can be more specific, up to the
>single objects.
>
>-- 
>"My method is uncertain/ It's a mess but it's working" [F. Apple]
>
>
>
>___
>talk mailing list
>talk@openstreetmap.org
>https://lists.openstreetmap.org/listinfo/talk

-- 
Sent from my Android device with K-9 Mail. Please excuse my brevity.___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Name:* tags in the local language

[Daniel Koć]

W dniu 26.04.2018 o 13:53, Marc Gemis pisze:
> Do you now assume that names in region B outside city C have a
> namehr;nameit in the name tag?

Yes, unless stated otherwise for cities E, F and G.

To be clear: I assume only that if they name:hr=* and name:it=*, both
are official, but that doesn't mean both need to be tagged (one or more
name tag may be just missing).

> That's not how we have done it for Belgium or any street in Brussels.

Isn't it like this:

Country Belgium - official_language=de;fr;nl
Region Brussels-Capital - official_language=fr;nl
City Eupen - official_language=de

What would be wrong with this scheme?

-- 
"My method is uncertain/ It's a mess but it's working" [F. Apple]



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

Re: [OSM-talk] Name:* tags in the local language

[Daniel Koć]

W dniu 26.04.2018 o 11:51, Mateusz Konieczny pisze:
> It is not handling
> - regions with more than one widespread language
> - features that have name tag in an atypical language

In my opinion there's no single solution for atypical cases, yet it's
sane to start with defaults. It can be used on many levels and region
can have more than one language defined, just like Janjko has shown:

Country A - official_language=hr
Region B - official_language=hr;it
City C - official_language=it

When a data consumer tries to find what is the official language for
street D, she finds that city C uses "it", so it's enough to know that
the street should be also in "it". Speaking of regions this is the
proper handling in my opinion - in region B you can for example use both
languages or choose the one you prefer, it's up to you.

But we don't have to stop with level 2, 4, or 6. If - for some reason -
part of the city is different and we have no way to show the borders of
the area, one can add official_language=* for the streets or objects in
this particular area. Of course instead of official_language we may use
default_language=* or common_language=* tag or something similar, the
same as the users have to choose common name for name=*.

This way we can always assume some language for the (administrative)
area, but if it's not true locally, we can be more specific, up to the
single objects.

-- 
"My method is uncertain/ It's a mess but it's working" [F. Apple]



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

Re: [OSM-talk] Name:* tags in the local language

[Mateusz Konieczny]

It is not handling
- regions with more than one widespread language
- features that have name tag in an atypical language

26. Apr 2018 10:04 by jan...@gmail.com :


>
>
> sri, 25. tra 2018. u 08:51 Darafei "Komяpa" Praliaskouski <> m...@komzpa.net 
> > > napisao je:
>
>> Hi,
>> maps.me >>  took approach similar to Nominatim's: each map 
>> region has "default language" in metadata, and in case the 
>> name: is needed but is missing, just name tag is 
>> taken.
>
> I think this is the best solution with the least amount of work for mappers. 
> But we put default languages in regions instead of an external database, as 
> someone before me already mentioned.
>
> Tag the default language on the whole country (official_language=hr). If a 
> region of a country has two official languages and the labels are, for 
> example, Rovinj/Rovigno, then put a official_language=hr/it. And on the safe 
> side, tag all the other subregions with official_language=hr, if the data 
> renderer chooses to download only one region.
>
> After that, no other changes are needed.
>
> Janko Mihelić
>
>___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Name:* tags in the local language

[Janko Mihelić]

sri, 25. tra 2018. u 08:51 Darafei "Komяpa" Praliaskouski 
napisao je:

> Hi,
>
> maps.me took approach similar to Nominatim's: each map region has
> "default language" in metadata, and in case the
> name: is needed but is missing, just name tag is
> taken.
>

I think this is the best solution with the least amount of work for
mappers. But we put default languages in regions instead of an external
database, as someone before me already mentioned.

Tag the default language on the whole country (official_language=hr). If a
region of a country has two official languages and the labels are, for
example, Rovinj/Rovigno, then put a official_language=hr/it. And on the
safe side, tag all the other subregions with official_language=hr, if the
data renderer chooses to download only one region.

After that, no other changes are needed.

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

Re: [OSM-talk] Name:* tags in the local language

[Christoph Hormann]

On Wednesday 25 April 2018, Jo wrote:
>
> There is this proposal:
>
> https://wiki.openstreetmap.org/wiki/Proposed_features/
> Language_information_for_name

That was a very different idea to what Frederik mentioned - this was 
meant to specify the language of the name=* tag.  This was intended to 
solve a very specific problem but was mostly rejected because it would 
add a lot of complexity to only solve that problem but not affect most 
of the other language related issues in mapping.

Lukas had made another proposal ultimately rejected that is somewhat 
more generic but is still doing things 'the other way round', i.e. by 
providing language information for the name tag:

https://wiki.openstreetmap.org/wiki/Proposed_features/Language_information


I don't think the idea of specifying the local name by reference to the 
local language has so far been elaborated on in the OSM wiki.

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

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

Re: [OSM-talk] Name:* tags in the local language

[Jo]

2018-04-25 8:21 GMT+02:00 Mateusz Konieczny :

>
> 24. Apr 2018 21:29 by rich...@systemed.net:
>
> If I read Frederik's proposal right, the language=en tag would be placed on
> the object with the name tag
>
>
> Interesting idea, I like it. Is there already a page on the OSM wiki
> describing this proposal?
>
Hi Mateusz,

There is this proposal:

https://wiki.openstreetmap.org/wiki/Proposed_features/
Language_information_for_name

Apparently it failed.

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

Re: [OSM-talk] Name:* tags in the local language

[Komяpa]

Hi,

maps.me took approach similar to Nominatim's: each map region has "default
language" in metadata, and in case the name: is
needed but is missing, just name tag is taken.

Similar approach was used when making a map for Wargaming's World of Tanks,
looking at the country's iso code. A list used there is available on github,
https://github.com/wgnet/globalmap/blob/master/data/country_languages.csv

(Translations that OSM lacks were performed by professional translators and
also published under ODbL,
https://github.com/wgnet/globalmap/blob/master/data/wgnl_localizations.csv,
and for non-critical map labels we've trained anything-to-latin and
anything-to-cyrillic neural networks, results
https://github.com/wgnet/globalmap/tree/master/data/rnn).

ср, 25 апр. 2018 г. в 9:23, Mateusz Konieczny :

>
> 24. Apr 2018 21:29 by rich...@systemed.net:
>
> If I read Frederik's proposal right, the language=en tag would be placed on
> the object with the name tag
>
>
> Interesting idea, I like it. Is there already a page on the OSM wiki
> describing this proposal?
> ___
> 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] Name:* tags in the local language

[Mateusz Konieczny]

24. Apr 2018 21:29 by rich...@systemed.net :

> If I read Frederik's proposal right, the language=en tag would be placed on
> the object with the name tag
>




Interesting idea, I like it. Is there already a page on the OSM wiki describing 
this proposal? 

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

Re: [OSM-talk] Name:* tags in the local language

[Yuri Astrakhan]

Adding a language=xx to each feature seems excessive, and will be forgotten
most of the time, unless there is some extensive tool support for it.
Adding it to admin regions seems like a better approach.  Some utility
could then calculate a clean translation map, using admin_level number as
the deciding factor in case of multiple values.

On a side note, a validation tool can than be used to check if "name" has
the same value as "name:xx", where xx is taken from that calculated map.
(I'm sure in many cases it won't match because some places use "lang:local
- lang:en" style of naming for the name tag.

On Tue, Apr 24, 2018 at 10:48 PM, Jo  wrote:

> I thought we were already indicating which language name is in, with the
> name:language=:iso tag?
>
> Hmm, apparently not:
>
> https://wiki.openstreetmap.org/wiki/Proposed_features/
> Language_information_for_name
>
> Polyglot
>
> 2018-04-24 21:29 GMT+02:00 Richard Fairhurst :
>
>> Paul Norman wrote:
>> > If there's agreement that there is a problem here, I could look
>> > at preparing a mechanical edit or MapRoulette challenge to add
>> > name:* tags, e.g. adding name:en to objects in the US with
>> > other name:* tags, and adding name:zh in China. As an
>> > estimate, this would be 115k changes in China, touching 28%
>> > of roads there.
>>
>> This is pretty fragile too, though. Two minutes after the mechanical
>> edit, a
>> newbie will come along and change the name= tag on a random American road
>> from MLK Boulevard to Martin Luther King Boulevard, without knowing they
>> now
>> have to change the name:en= tag as well. Bang, inconsistent data. Fast
>> forward two years and a bunch of history-losing way splits, and it's no
>> longer clear which is the accurate street name and which is the original,
>> mistaken TIGER-imported one.
>>
>> In theory you could bake support for this into editing software (at the
>> expense of complicating the interface), but even if JOSM, iD, Vespucci and
>> P2 all add support, the name= tag is probably the most likely to be
>> changed
>> by minority editors (e.g. mobile or 'quick fix' apps) and it's unlikely
>> they'll all add the same logic.
>>
>> Mateusz Konieczny wrote:
>> > This language=en tag would be placed on a administrative
>> > relation, right?
>>
>> If I read Frederik's proposal right, the language=en tag would be placed
>> on
>> the object with the name tag, though putting it on admin relations is an
>> interesting idea.
>>
>> Richard
>>
>>
>>
>> --
>> Sent from: http://gis.19327.n8.nabble.com/General-Discussion-f5171242.
>> html
>>
>> ___
>> talk mailing list
>> talk@openstreetmap.org
>> https://lists.openstreetmap.org/listinfo/talk
>>
>
>
> ___
> 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] Name:* tags in the local language

[Jo]

I thought we were already indicating which language name is in, with the
name:language=:iso tag?

Hmm, apparently not:

https://wiki.openstreetmap.org/wiki/Proposed_features/Language_information_for_name

Polyglot

2018-04-24 21:29 GMT+02:00 Richard Fairhurst :

> Paul Norman wrote:
> > If there's agreement that there is a problem here, I could look
> > at preparing a mechanical edit or MapRoulette challenge to add
> > name:* tags, e.g. adding name:en to objects in the US with
> > other name:* tags, and adding name:zh in China. As an
> > estimate, this would be 115k changes in China, touching 28%
> > of roads there.
>
> This is pretty fragile too, though. Two minutes after the mechanical edit,
> a
> newbie will come along and change the name= tag on a random American road
> from MLK Boulevard to Martin Luther King Boulevard, without knowing they
> now
> have to change the name:en= tag as well. Bang, inconsistent data. Fast
> forward two years and a bunch of history-losing way splits, and it's no
> longer clear which is the accurate street name and which is the original,
> mistaken TIGER-imported one.
>
> In theory you could bake support for this into editing software (at the
> expense of complicating the interface), but even if JOSM, iD, Vespucci and
> P2 all add support, the name= tag is probably the most likely to be changed
> by minority editors (e.g. mobile or 'quick fix' apps) and it's unlikely
> they'll all add the same logic.
>
> Mateusz Konieczny wrote:
> > This language=en tag would be placed on a administrative
> > relation, right?
>
> If I read Frederik's proposal right, the language=en tag would be placed on
> the object with the name tag, though putting it on admin relations is an
> interesting idea.
>
> Richard
>
>
>
> --
> Sent from: http://gis.19327.n8.nabble.com/General-Discussion-f5171242.html
>
> ___
> 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] Name:* tags in the local language

[Christoph Hormann]

On Tuesday 24 April 2018, Frederik Ramm wrote:
>
> Someone once suggested to have a special tag that indicates which
> name tag should be used by default. I.e. we'd have tons of "name:xx"
> tags plus one tag called e.g. "language=en", that would then mean:
> The default name to use is the name:en name.

I think that someone was me. :-)

Details on:

http://blog.imagico.de/you-name-it-on-representing-geographic-diversity-in-names/

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

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

Re: [OSM-talk] Name:* tags in the local language

[Richard Fairhurst]

Paul Norman wrote:
> If there's agreement that there is a problem here, I could look 
> at preparing a mechanical edit or MapRoulette challenge to add 
> name:* tags, e.g. adding name:en to objects in the US with 
> other name:* tags, and adding name:zh in China. As an 
> estimate, this would be 115k changes in China, touching 28% 
> of roads there.

This is pretty fragile too, though. Two minutes after the mechanical edit, a
newbie will come along and change the name= tag on a random American road
from MLK Boulevard to Martin Luther King Boulevard, without knowing they now
have to change the name:en= tag as well. Bang, inconsistent data. Fast
forward two years and a bunch of history-losing way splits, and it's no
longer clear which is the accurate street name and which is the original,
mistaken TIGER-imported one.

In theory you could bake support for this into editing software (at the
expense of complicating the interface), but even if JOSM, iD, Vespucci and
P2 all add support, the name= tag is probably the most likely to be changed
by minority editors (e.g. mobile or 'quick fix' apps) and it's unlikely
they'll all add the same logic.

Mateusz Konieczny wrote:
> This language=en tag would be placed on a administrative 
> relation, right? 

If I read Frederik's proposal right, the language=en tag would be placed on
the object with the name tag, though putting it on admin relations is an
interesting idea.

Richard



--
Sent from: http://gis.19327.n8.nabble.com/General-Discussion-f5171242.html

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

Re: [OSM-talk] Name:* tags in the local language

[Mateusz Konieczny]

24. Apr 2018 20:03 by md...@xs4all.nl :


> I think you're opening another can of worms here. How would this work out for 
> me (in Dutch) where I don't want to see places in latin script translated? I 
> do not want to see Londen or Berlijn or Brunswijk, I also don't want to see 
> English translations like Cologne or Munich, but I do want to see readable 
> (i.e. non-Cyrillic, Kanji, Hangul, etc) names.
>




Simple fallback would not work here but something like that may work

if name tag has only latin characters use name tag

otherwise fallback through following name tags:

name:nl, name:de, name:en...

With such scheme missing name:nl would not cause problems here (except
cases where dutch name has characters beyond latin script, I have no
idea whatever it may happen).

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

Re: [OSM-talk] Name:* tags in the local language

[Sarah Hoffmann]

On Tue, Apr 24, 2018 at 08:16:08PM +0200, Frederik Ramm wrote:
> Hi,
> 
> On 04/24/2018 07:23 PM, Paul Norman wrote:
> > If there's agreement that there is a problem here, I could look at
> > preparing a mechanical edit or MapRoulette challenge to add name:* tags,
> > e.g. adding name:en to objects in the US with other name:* tags, and
> > adding name:zh in China. As an estimate, this would be 115k changes in
> > China, touching 28% of roads there.
> 
> Even if there should be agreement that there is a problem here, there
> are other potential solutions.
> 
> Someone once suggested to have a special tag that indicates which name
> tag should be used by default. I.e. we'd have tons of "name:xx" tags
> plus one tag called e.g. "language=en", that would then mean: The
> default name to use is the name:en name.
> 
> I think this would be more elegant than the duplication that you are
> suggesting.

That still doesn't scale. You will have a hard time to convince mappers
to repeatedly add something to objects for which there is an obvious
default. This really should be solved at least partially in software.

I'd like to point to https://wiki.openstreetmap.org/wiki/Nominatim/Country_Codes
again. This list is a good place to start when you want to guess the
language a name tag is in and solves the case for monolingual coutries.
Multilingual countries tend to be more sensitive to the language issue
so that coverage with name:* tags tends to be better.

Also don't mix up the issue of languages and scripts. name:zh is
a point in case where knowing the language doesn't help you if you
want to present the users the map in their favourite script. name:zh
will normally contain Pidgin but may also have traditional Chinese
from time to time.

I strongly recommend to have a look into Sven's work on the German
mapnik style, which has been trying to address a lot of issues with
localized maps: https://github.com/giggls/mapnik-german-l10n

Sarah

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

Re: [OSM-talk] Name:* tags in the local language

[Mateusz Konieczny]

On Tue, 24 Apr 2018 20:16:08 +0200
Frederik Ramm  wrote:

> I.e. we'd have tons of "name:xx" tags
> plus one tag called e.g. "language=en", that would then mean: The
> default name to use is the name:en name.
> 
> I think this would be more elegant than the duplication that you are
> suggesting.

This language=en tag would be placed on a administrative relation,
right?

Note that it would work only in places with single language, and even
in very homogeneous countries, with single culture there are some
correctly tagged objects with name tag in a different languages.

So it would be anyway necessary to supply name:en to names that are in
English to confirm that this names are really in English.

This method would also not work at all in places with more complex
language situation (multiple communities, each using local language).

On top of that it is a bit complicated to use as a data consumer.

Overall I am not convinced that it is a good replacement.

PS My comments in future will be from a matkoni...@tutanota.com email,
that is still me - not some impostor.

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

Re: [OSM-talk] Name:* tags in the local language

[Tobias Knerr]

On 24.04.2018 19:23, Paul Norman wrote:
> It is sometimes recommended that when you add a name in another language
> you also indicate the name in the local language by adding a suitable
> name:* tag at the same time. For example, if adding "name:fr=Londres" to
> London, you would also add "name:en=London" if it weren't present.
> 
> This practice is not widely followed.

I have to say I'm not a fan of this practice, at least for
straightforward cases.

Yes, there are parts of the world where explicit language information is
necessary. But in many others, there is an obvious default, and it feels
wrong to require mappers to manually repeat this on millions of
individual elements. Don't add extra tags to German names in Germany,
for example – tag the exceptions from the rule instead.

As a less important secondary point, I also consider the specific
tagging (duplicating the name with a different key) counter-intuitive
and would prefer a "language of the name" key, with the language given
as the value. The purpose of that tag would be much more
self-documenting than with the recommendation described above.

That being said, displaying labels in the user's language is an exciting
use case, and I hope we can find a solution that allows your project to
succeed. Could you elaborate a bit more on the technical limitations
regarding regional processing?

> Fixing this in software doesn't work
> well because it requires regional processing of what are increasingly
> small regions as you get to less used languages.

You mention increasingly small regions as the obstacle. Would that still
allow for the pragmatic compromise of setting country-level defaults,
and using explicit tagging only for smaller-scale language regions?

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

[OSM-talk] Name:* tags in the local language

[Frederik Ramm]

Hi,

On 04/24/2018 07:23 PM, Paul Norman wrote:
> If there's agreement that there is a problem here, I could look at
> preparing a mechanical edit or MapRoulette challenge to add name:* tags,
> e.g. adding name:en to objects in the US with other name:* tags, and
> adding name:zh in China. As an estimate, this would be 115k changes in
> China, touching 28% of roads there.

Even if there should be agreement that there is a problem here, there
are other potential solutions.

Someone once suggested to have a special tag that indicates which name
tag should be used by default. I.e. we'd have tons of "name:xx" tags
plus one tag called e.g. "language=en", that would then mean: The
default name to use is the name:en name.

I think this would be more elegant than the duplication that you are
suggesting.

Bye
Frederik

-- 
Frederik Ramm  ##  eMail frede...@remote.org  ##  N49°00'09" E008°23'33"

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

Re: [OSM-talk] Name:* tags in the local language

[Maarten Deen]

On 2018-04-24 19:40, Mateusz Konieczny wrote:

I fully agree that it is a problem. I encountered nasty issues after
implementing name rendering fallback with following language order
(order is simplified for this example):

name:pl, name:en, name:de, name

Intention was to render English name in China/Korea rather than
unreadable (for typical person from Poland) local name.

The result was rendering German names in Poland, because name:pl was
missing and name:de was present.


I think you're opening another can of worms here. How would this work 
out for me (in Dutch) where I don't want to see places in latin script 
translated? I do not want to see Londen or Berlijn or Brunswijk, I also 
don't want to see English translations like Cologne or Munich, but I do 
want to see readable (i.e. non-Cyrillic, Kanji, Hangul, etc) names.


Maarten

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

Re: [OSM-talk] Name:* tags in the local language

[Mateusz Konieczny]

I fully agree that it is a problem. I encountered nasty issues after
implementing name rendering fallback with following language order
(order is simplified for this example):

name:pl, name:en, name:de, name

Intention was to render English name in China/Korea rather than
unreadable (for typical person from Poland) local name.

The result was rendering German names in Poland, because name:pl was
missing and name:de was present.

Somebody else also encountered this issue and places in Poland are
now partially fixed.

On Tue, 24 Apr 2018 10:23:19 -0700
Paul Norman  wrote:

> preparing a mechanical edit or MapRoulette challenge

I would consider also making a quest in StreetComplete (it is on my
TODO list since a long time).

In my experience StreetComplete is working much better that
MapRoulette, though I am not sure whatever this task fits well "local
survey" model.

In case of mechanical edit - can you consider sharing script
(hopefully one working without complex environment)? I would happily run
such mechanical edit in Poland (including obtaining permission from
local community, checking whatever edits are correct and potential
cleanup of wrong edits).

Making mechanical edit is also on my TODO list since a long time, but
it never went further than vague plan and half-written explanation why
it is useful.

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

[OSM-talk] Name:* tags in the local language

[Paul Norman]

As part of my Wikimedia Foundation work, I'm working on labeling in
multiple languages using OSM data. We've run into an issue, and it's not
clear how to best solve it.

It is sometimes recommended that when you add a name in another language
you also indicate the name in the local language by adding a suitable
name:* tag at the same time. For example, if adding "name:fr=Londres" to
London, you would also add "name:en=London" if it weren't present.

This practice is not widely followed. For example, 3% of the roads with
names in multiple languages in the western US have a name:en tag.[1]
Cities are better, but still only 30% of cities and towns with names in
multiple languages have a name:en tag. For the same criteria in China
and with Chinese, it's 30% of roads and 75% of cities.[2]

This is a problem for displaying labels. Proper display of labels is
more than just showing the name in the language the user requested and
falling back to the name=* tag. It requires falling back through
multiple languages. For example, if someone requests a map in
Luxembourgish, you would show German labels where places and objects
have no Luxembourgish names. There are similar situations for French
Creoles and many regional and minority languages.

We'd like to be able to do more. Specifically, we'd like to always show
labels in the user's script if possible. For example if a French user is
viewing a map of India, they're more likely to be able to read an
English name than one in a Dravidian language, Hindi, or Bengali. The
common lack of a name:* tag for the primary language while having other
name:* tags makes this impossible.

There aren't any great solutions. Fixing this in software doesn't work
well because it requires regional processing of what are increasingly
small regions as you get to less used languages. Language detection from
the tag value is fragile and introducing magic logic. This really needs
addressing on the OSM data side.

If there's agreement that there is a problem here, I could look at
preparing a mechanical edit or MapRoulette challenge to add name:* tags,
e.g. adding name:en to objects in the US with other name:* tags, and
adding name:zh in China. As an estimate, this would be 115k changes in
China, touching 28% of roads there.

[1]: https://phabricator.wikimedia.org/T192662#4151714
[2]: https://phabricator.wikimedia.org/T192662#4147291


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