Yes - agreed. But at the moment, this would lead to useless information
being displayed, because OSM contributers haven't followed the rules
either. There isn't a strong incentive for app developers to adopt it it
just leads to garbage being displayed.
How do we get OSM contributers themselves to follow the scheme? Once we
have good data we can then work with app developers to follow the scheme.
Bjoern
On Sat, 8 Apr 2023, 10:04 David Davis, <feline1...@googlemail.com> wrote:
> Re those misbehaving map renderers -
> Probably worth contacting them and pointing out that they're not following
> the tagging scheme set out on OSM wiki. They're probably then raise bugs
> and correct it in due course.
>
> On Sat, 8 Apr 2023, 00:41 Bjoern Hassler, <bjohas...@gmail.com> wrote:
>
>> Hi Rob,
>>
>> thanks!
>>
>> I appreciate that the tagging scheme excludes 'Exit 1', and suggests to
>> not include the station. I've extracted exits via overpass and most exits
>> don't follow the scheme. :) Most exits are simply the station name, and/or
>> have the exit number written into the name. Some have what it says on the
>> sign ("Exit 1. ..."), and a few do follow the scheme and just have what's
>> on the sign minus the "Exit 1".
>>
>> While I appreciate the nature of OSM relations that give more info, ...
>> and am aware that we're not tagging for the renderer.... but Organic Maps
>> (and Maps.Me) renders the station name, and I can search for "Bank Exit 1"
>> and find it (because that's in the name). I can search for Argyll Street,
>> and will find "Exit 8" (but cannot see which station it is.) In principle,
>> that seems very usable for people who would need to rely on using certain
>> exits (e.g., for access).
>>
>> So... I do appreciate the scheme and that the scheme hasn't been changed
>> in a while... but, given the observations above: *Do you/others think
>> there's scope for bringing the labeling scheme in line with what apps can
>> find while at the same time bringing the osm name in line with signage?*
>>
>> I appreciate the view that it's the fault of the app not rendering
>> adequately - but tagging is far from perfect, e.g. 'ref' is a bit random -
>> so why would apps go out of their way to use imperfect tagging? Quite
>> possibly apps are just doing to use free text searches. I'd be quite happy
>> to put some time into regularising the exit names. However, I'd want the
>> end result to work reasonably well in apps, so it's actually of practical
>> use to somebody. At the same time, I could also do this outside OSM, i.e.,
>> make a set of bookmarks for OrganiseMaps that people can overlay. However,
>> it would be nicer to store the data in OSM, so others can find it. *What
>> do you think?*
>>
>> I realise that it doesn't help much that the four entrances which are
>>> members of that relation were incorrectly tagged 5 years ago with
>>> name="Aldgate East" (understandable) and ref=* values of A,B,C and F
>>> (which seem very odd).
>>
>>
>> I can maybe offer some observations/heuristics here. I don't know what
>> the 'ref' values are, but the values for 'ref' are labels that can be found
>> on the axiometric projections, see
>> https://www.ianvisits.co.uk/articles/3d-maps-of-every-underground-station-ab-14630/.
>> They are also referenced in some of the open data TfL makes available,
>> e.g., in the unique exit id.
>>
>> (1) An agency did some mapping about 4-5 years ago on behalf of TfL.
>> https://mentz.net/ E.g.,
>> https://www.openstreetmap.org/node/475381373/history, #9. I raised some
>> issues with the tagging at the time, but it wasn't really possible to get
>> hold of the people doing the mapping. The ref=A,B,C were introduced at the
>> time.
>>
>> (2) As far as I can see, those labels are purely of internal use for TfL
>> (if of any use at all). I agree that they could be replaced... but see (1).
>> So maybe there needs to be a different tag that can hold the ref=A,B,C and
>> ref can then be used for the entrance number. (TfL does have the unique ids
>> for station exits, which can be looked up in their open data on the basis
>> of the ref=A,B,C. They could be added and make the ref=A,B,C redundant.)
>>
>> Maybe it's all a bit too random too fix, but I'd be happy to put some
>> energy into fixing this on OSM, because I have experienced accurate exit
>> information being useful (and I value open data). But, at the same time, it
>> could be done outside OSM as well. Opinions welcome.
>>
>> This is clearly useless for routing, but luckily
>>> for me CityMapper gets its station information from somewhere else.
>>>
>>
>> I had a look at CityMapper, but I couldn't see station exits used.
>>
>> Björn
>> _______________________________________________
>> Talk-gb-london mailing list
>> Talk-gb-london@openstreetmap.org
>> https://lists.openstreetmap.org/listinfo/talk-gb-london
>>
>
_______________________________________________
Talk-gb-london mailing list
Talk-gb-london@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-gb-london
