Re: [OSM-talk] Adding wikidata tags to the remaining objects with only wikipedia tag

[dkiselev]

You can compareanything (title, coordinates), in any direction with someapproximation if needed etc.  That's the root of an evil. That comparison have to be done manually. I really don't understand why wikidata needs to be added. It needs to be added to avoid manual fixes of wikipedia links,because wikipedia's articles names aren't constant. That looks like inventing sphisticated comparison procedurewhich has to be done manualy only not to have wikidata tags. What's the point in not having wikidata?  02.10.2017, 05:15, "john whelan" :Rather than fill OSM up with automated edits that have not even been discussed with the local community can we think more about functionality? Since an OSM object has  lat and long value and it appears that wiki whatever also has one the entries can be linked. "This gives you a very simple table with: lat/lonThat/page_title.  No parsing or anything else involved.  You then take data from OSM - lat/lon/wikipedia.tag  So you have two tables of same structure. Voila. You can compareanything (title, coordinates), in any direction with someapproximation if needed etc. No OSM wikidata involved at all." I really don't understand why wikidata needs to be added.  Note the word need, I'm missing the requirement somehow that overides following normal OSM practices. Cheerio John On 1 Oct 2017 7:19 pm,  wrote:Hi everybody.We already accepted wikipedia links keep in mind that wiki article isn't the same abstraction as OSM object.And the way we make a reference on wiki articles differs over time.It was a link, it was an article name, it was a name with language prefix.Wikidata id is a way to make a refererence onto wikipedia articlewhich wiki community states to be the right and consistent one.Why don't we just accept the recomended way of referencing wiki articles?That doesn't make wikipedia articles more or less verifable on the ground.But that makes the link persistent agains the changes in wikipedia.01.10.2017, 16:38, "Christoph Hormann" :> On Sunday 01 October 2017, Marc Gemis wrote:>>  If I noticed an inception date on a information sign next to a>>  building, is this original research or a secondary source ?>> The date on the sign is verifiable as a signed date, not necessarily as> a date connected to the building. Historic information like dates from> before the timeframe witnessed by today's mappers is something that is> problematic in OSM in general.>> As any historian will be able to tell you history as a science is not> about analyzing events of the past in the same way you analyze present> day observations in empiric science disciplines but about analyzing and> interpreting sources - in this way it resembles the approach of> Wikipedia more than that of OSM.>>>  If a Wikidata contributor/Wikipedian researches websites like Orbis,>>  Onroerend erfgoed, etc and finds different inception dates and>>  records them all, is it bad that s/he uses secondary sources instead>>  of the information sign on the ground ?>> I don't want to judge the different approaches of Wikipedia and OSM -> both have their pros and cons. As a contributor i am more comfortable> with the OSM approach but i would never say that collecting information> from secondary sources is inherently less valuable than collecting> empiric data.>>>  I wonder whether OSM is really always based on original research, or>>  whether we sometimes just want to believe this.>> Well - human perception is of course always connected to our past> experience and beliefs, we cannot observe our environment in a truly> neutral way. And of course with armchair mapping much of the data in> OSM is produced by people who have never been near the place they map> based on the belief that what the image layer they use depicts what is> actually there (which is sometimes not the case - like with the iconic> demolished buildings) and that what people think they see in the images> actually resembles the impression they would have on the ground (which> is also frequently not the case - prominent example would be here:> http://www.openstreetmap.org/relation/7142214). But the key is every> information in OSM has to be able to stand up to scrutiny by local> mappers - a local mapper standing in front of the object needs to be> able to tell if it is correct or not - even if different mappers could> in borderline cases come to different assessments on the matter.>> --> Christoph Hormann> http://www.imagico.de/>> ___> talk mailing list> talk@openstreetmap.org> https://lists.openstreetmap.org/listinfo/talk___talk mailing listtalk@openstreetmap.orghttps://lists.openstreetmap.org/listinfo/talk___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Adding wikidata tags to the remaining objects with only wikipedia tag

[dkiselev]

Hi everybody.

We already accepted wikipedia links keep in mind that wiki article isn't the 
same abstraction as OSM object.
And the way we make a reference on wiki articles differs over time.
It was a link, it was an article name, it was a name with language prefix.

Wikidata id is a way to make a refererence onto wikipedia article 
which wiki community states to be the right and consistent one.
Why don't we just accept the recomended way of referencing wiki articles?

That doesn't make wikipedia articles more or less verifable on the ground.
But that makes the link persistent agains the changes in wikipedia.

01.10.2017, 16:38, "Christoph Hormann" :
> On Sunday 01 October 2017, Marc Gemis wrote:
>>  If I noticed an inception date on a information sign next to a
>>  building, is this original research or a secondary source ?
>
> The date on the sign is verifiable as a signed date, not necessarily as
> a date connected to the building. Historic information like dates from
> before the timeframe witnessed by today's mappers is something that is
> problematic in OSM in general.
>
> As any historian will be able to tell you history as a science is not
> about analyzing events of the past in the same way you analyze present
> day observations in empiric science disciplines but about analyzing and
> interpreting sources - in this way it resembles the approach of
> Wikipedia more than that of OSM.
>
>>  If a Wikidata contributor/Wikipedian researches websites like Orbis,
>>  Onroerend erfgoed, etc and finds different inception dates and
>>  records them all, is it bad that s/he uses secondary sources instead
>>  of the information sign on the ground ?
>
> I don't want to judge the different approaches of Wikipedia and OSM -
> both have their pros and cons. As a contributor i am more comfortable
> with the OSM approach but i would never say that collecting information
> from secondary sources is inherently less valuable than collecting
> empiric data.
>
>>  I wonder whether OSM is really always based on original research, or
>>  whether we sometimes just want to believe this.
>
> Well - human perception is of course always connected to our past
> experience and beliefs, we cannot observe our environment in a truly
> neutral way. And of course with armchair mapping much of the data in
> OSM is produced by people who have never been near the place they map
> based on the belief that what the image layer they use depicts what is
> actually there (which is sometimes not the case - like with the iconic
> demolished buildings) and that what people think they see in the images
> actually resembles the impression they would have on the ground (which
> is also frequently not the case - prominent example would be here:
> http://www.openstreetmap.org/relation/7142214). But the key is every
> information in OSM has to be able to stand up to scrutiny by local
> mappers - a local mapper standing in front of the object needs to be
> able to tell if it is correct or not - even if different mappers could
> in borderline cases come to different assessments on the matter.
>
> --
> Christoph Hormann
> http://www.imagico.de/
>
> ___
> talk mailing list
> talk@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/talk

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