Re: [OSM-talk] non-temporary usage of highway=road

[Simon Poole]

Aun

The use of highway=road goes back to before we had access to aerial
imagery. Often you would simply have a raw GPX track without any further
meta-data and you might simply wanted to indicate that there might be
something there for later survey and there was neither the amount of
correct road data nor the tools we have available at this point in time.

Use of it, IMHO, doesn't really make sense at all any more, particularly
if you have imagery available, a conservative best guess plus a fixme
tag would seem to be much more useful. We shouldn't forget that way back
we were limited by the speed at which we could add geometry, these days
it's, even with humans doing it :-), the fast bit of adding data to OSM
and it rarely makes sense to do so in advance if there is no way to
infer at least basic meta data or you are doing it immediately before a
survey. 

QA tools and editors already tend to report/highlight such ways (iD
perhaps not), finding them shouldn't be an issue.

Simon


Am 27.09.2016 um 03:42 schrieb Aun Johnsen:
> I see a lot of usage of the generic tag highway=road, which is meant as a 
> temporary tag and a “low level entry tag” for beginners. The problem with 
> this is that because of its ambiguous meaning, it is impossible for data 
> consumers to process this correctly. 
>
> i.e., highway=road means anything between a small footway to motorway, so a 
> routing engine might wrongfully send cars down a set of steps or a hiker up a 
> motorway
>
> As described on the proposal page, this tag is meant to be temporary until 
> more data can be obtained from survey.
>
> I have now had a few rounds of cleanup of the usage of this tag in Brazil. My 
> cleanup run last year found highway=road that had been unedited for 3 years, 
> so hardly temporary. This year I found several highway=road added by armchair 
> mappers from Europe, and I doubt they ever will travel to the remote areas of 
> Brazil to correct this.
>
> I am still in progress of checking each of these elements manually, I started 
> this years cleaning with more than 3600 highway road in Brazil alone, roughly 
> 14 months after I had completely cleaned Brazil the last time.
>
> In my opinion, a warning about its temporary state and non-capability with 
> data consumers should be added to the wiki, and it should be removed from the 
> standard presets of all the common editors. Also editors with validation 
> functions should give a warning about the existence of this temporary tag so 
> that it can be dealt with properly by people editing in the areas, and 
> further QA tools should highlight them as items needing attention.
>   
> Aun Johnsen
>
>
> ___
> talk mailing list
> talk@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/talk




signature.asc
Description: OpenPGP digital signature
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

[OSM-talk] non-temporary usage of highway=road

[Aun Johnsen]

I see a lot of usage of the generic tag highway=road, which is meant as a 
temporary tag and a “low level entry tag” for beginners. The problem with this 
is that because of its ambiguous meaning, it is impossible for data consumers 
to process this correctly. 

i.e., highway=road means anything between a small footway to motorway, so a 
routing engine might wrongfully send cars down a set of steps or a hiker up a 
motorway

As described on the proposal page, this tag is meant to be temporary until more 
data can be obtained from survey.

I have now had a few rounds of cleanup of the usage of this tag in Brazil. My 
cleanup run last year found highway=road that had been unedited for 3 years, so 
hardly temporary. This year I found several highway=road added by armchair 
mappers from Europe, and I doubt they ever will travel to the remote areas of 
Brazil to correct this.

I am still in progress of checking each of these elements manually, I started 
this years cleaning with more than 3600 highway road in Brazil alone, roughly 
14 months after I had completely cleaned Brazil the last time.

In my opinion, a warning about its temporary state and non-capability with data 
consumers should be added to the wiki, and it should be removed from the 
standard presets of all the common editors. Also editors with validation 
functions should give a warning about the existence of this temporary tag so 
that it can be dealt with properly by people editing in the areas, and further 
QA tools should highlight them as items needing attention.
  
Aun Johnsen


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

Re: [OSM-talk] Filling in the gaps in the map with crowdsourcing techniques

[Simon Poole]

I assume everybody saw my talk yesterday :-).

Seriously, even if you don't expose geometry to the end user your editor
needs to be geometry-aware (that means have at least internal access to
all variants of how the object of interest can be modeled in OSM) and
make the user aware of the presence of nearby objects.

As I pointed out in the talk, building a simple editor is hard. So hard
that we've not managed to actually develop one on any platform to date.
And there are number of things in OSM that actually work against that
(freeform tagging, complicated tagging schemes, different interpretation
of the schemes in the preset systems and so on).

Simon

On 26.09.2016 15:10, john whelan wrote:
> >From evidence of the edits I'm still to be convinced as to the usefulness
> of these smart phone devices. Often inaccurate & unclear (Navmii) or
> unable to recognise entities mapped as polygons & limit presets.
> (MAPS.ME  & others)
>
> A _very_ experienced mapper recently added POIs to places already
> mapped as closed ways:
> "I was testing out MAPS.ME  yesterday while out and
> about. But it's a bit limiting. Clearly I needed a shop=tea preset for
> example!"
>
> That is very much in my mind, buildings mapped as a way do not show up
> in OSMand for example for adding tags only POIs are available.
>
> Cheerio John
>
> On 26 September 2016 at 08:59, Dave F  > wrote:
>
>
> On 25/09/2016 19:11, john whelan wrote:
>
> OSMand and POI editing springs to mind...
>
>
> From evidence of the edits I'm still to be convinced as to the
> usefulness of these smart phone devices. Often inaccurate &
> unclear (Navmii) or unable to recognise entities mapped as
> polygons & limit presets. (MAPS.ME  & others)
>
> A _very_ experienced mapper recently added POIs to places already
> mapped as closed ways:
> "I was testing out MAPS.ME  yesterday while out
> and about. But it's a bit limiting. Clearly I needed a shop=tea
> preset for example!"
>
> Dave F. (Walking boots/bike, GPS, Camera, Pub)
>
>
>
> ___
> 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] Filling in the gaps in the map with crowdsourcing techniques

[john whelan]

>From evidence of the edits I'm still to be convinced as to the usefulness
of these smart phone devices. Often inaccurate & unclear (Navmii) or unable
to recognise entities mapped as polygons & limit presets. (MAPS.ME & others)

A _very_ experienced mapper recently added POIs to places already mapped as
closed ways:
"I was testing out MAPS.ME yesterday while out and about. But it's a bit
limiting. Clearly I needed a shop=tea preset for example!"

That is very much in my mind, buildings mapped as a way do not show up in
OSMand for example for adding tags only POIs are available.

Cheerio John

On 26 September 2016 at 08:59, Dave F  wrote:

>
> On 25/09/2016 19:11, john whelan wrote:
>
>> OSMand and POI editing springs to mind...
>>
>
> From evidence of the edits I'm still to be convinced as to the usefulness
> of these smart phone devices. Often inaccurate & unclear (Navmii) or unable
> to recognise entities mapped as polygons & limit presets. (MAPS.ME &
> others)
>
> A _very_ experienced mapper recently added POIs to places already mapped
> as closed ways:
> "I was testing out MAPS.ME yesterday while out and about. But it's a bit
> limiting. Clearly I needed a shop=tea preset for example!"
>
> Dave F. (Walking boots/bike, GPS, Camera, Pub)
>
>
>
> ___
> 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] Filling in the gaps in the map with crowdsourcing techniques

[Dave F]

On 25/09/2016 19:11, john whelan wrote:

OSMand and POI editing springs to mind...


From evidence of the edits I'm still to be convinced as to the 
usefulness of these smart phone devices. Often inaccurate & unclear 
(Navmii) or unable to recognise entities mapped as polygons & limit 
presets. (MAPS.ME & others)


A _very_ experienced mapper recently added POIs to places already mapped 
as closed ways:
"I was testing out MAPS.ME yesterday while out and about. But it's a bit 
limiting. Clearly I needed a shop=tea preset for example!"


Dave F. (Walking boots/bike, GPS, Camera, Pub)


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

Re: [OSM-talk] Map features page on wiki

[François Lacombe]

Understood Dalibor,

Obviously, no offense intended while replacing the whole template like I
did few days ago.

The problem isn't the taglist itself but the lack of translation of keys'
pages on wiki, I agree with that too.

Nevertheless I think it's a bad idea to wait for all translations of all
templates in all languages : we'll never implement this useful feature from
taginfo.
Can't we take this as an advantage to encourage wiki contributors to move
to this new format by creating local translations for the key they use ?
I mean no one will ever move if we say "ok there is a new way of defining
features lists but we put it aside for a while". And the wiki will be
messier.

Is there any other problem I didn't see regarding this ?


All the best

François

2016-09-26 10:51 GMT+02:00 Dalibor Jelínek :

> Hi Francois,
>
> it seems that we are talking about two different things.
>
> Yes, you are right that when all the tags in Taglist table
>
> are translated then it shows them correctly in given language.
>
> But when the tags themselves are not translated then
>
> there is nothing to show in that language.
>
[...]
>
>
>
> Regards,
>
> Dalibor (chrabros)
>
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Map features page on wiki

[François Lacombe]

Hi Matthijs,

2016-09-23 14:27 GMT+02:00 Matthijs Melissen :

> Yes, for the moment it is necessary to create a new page under a
> different name. If you move over the current template to TagList,
> non-English language wiki pages will break.
>

For me It worked like a charm.
I don't see that problem on power features list. The list is ok for
English, French and Italian languages.

http://wiki.openstreetmap.org/wiki/Map_Features#Power

Can you show me an example of bad case please ?

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