[talk-au] Contributions to Road Geometry in Perth,

Sun Sep 20 05:47:03 UTC 2020

On 3/9/20 9:38 pm, Ian Steer wrote:
>> Date: Thu, 03 Sep 2020 19:23:21 +1000
>> From: "Sebastian S." <mapping at consebt.de>
>> Subject: Re: [talk-au] Contributions to Road Geometry in Perth,
>> 	Australia
>> Hi,
>> I have made excessive use of the node tag for islands.
>> Particularly for pedestrian crossing.
>>
>> Splitting the road into two separate ways for only a few metres seems excessive to me. Even when there is a several Meter long raised kerb separating the lanes I would not >split the road.
> Could you please elaborate on this method you have been using ?
>
> Ian

Hi Ian,

have a look at Marianna Road

https://www.openstreetmap.org/way/176612937#map=19/-33.69876/151.28378

where I have used islands for stretches in the road that are separated
by a concrete barrier. (Height is like a curb so it could be argued if
this is segregated enough)

but even if the road was split by a proper island for the length of the
intersection I would not split the way.

Seb

>
>
> On 1 September 2020 10:05:42 pm AEST, Andrew Harvey <andrew.harvey4 at gmail.com> wrote:
>> Heads up, looks like their team has started to map in Perth, see on 
>> OSMCha
>> -> https://osmcha.org/?aoi=80b50a6d-6bb5-48cb-8ac4-4b2ddd9d5d76
>>
>> Mostly looks okay to me, and mostly minor tweaks, though I raised a few 
>> questions and issues on changeset comments but also listed most of them
>> here:
>>
>> https://www.openstreetmap.org/way/840589945/history was added but the 
>> existing road name and other applicable attributes were not applied.
>> This
>> same issue happens in quite a few other places too so appears to be 
>> systemic. I've raised some changeset comments but worth including this 
>> as part of the standard practice by your editing team.
>>
>> https://www.openstreetmap.org/way/842851495/history is that a 
>> roundabout? I can't tell from the Maxar imagery, yet that is the 
>> claimed source, how could you tell from the imagery what this is?
>>
>> I personally find splitting ways for a traffic island at roundabouts 
>> like in https://www.openstreetmap.org/way/840189281/history a tad to 
>> excessive (would prefer to just tag the node as traffic island and use 
>> one way, gives a much cleaner dataset as the transition between dual 
>> and single carriageways is always messy) but I guess it's not wrong and 
>> both styles are popular in OSM currently. Does the community have a 
>> view on this?
>>
>> Unclear source of the turn restriction in
>> https://www.openstreetmap.org/changeset/90223764#map=18/-32.04553/115.8
>> 0953
>>
>> On Sun, 16 Aug 2020 at 21:28, OSM NextBillion. AI <osm at nextbillion.ai>
>> wrote:
>>
>>> Thank you cleary for valuable insights, we would be more cautious
>> while
>>> mapping in such areas. While Satellite Imagery is our prime resource,
>> we’d
>>> consider mapillary photos as well wherever available. We do have some 
>>> expert assistance in our team for interpreting satellite imagery and
>> map
>>> something only if we’re double sure of it’s existence. We will refer
>> to
>>> mappers history before editing existing data to understand if it was 
>>> created using local expertise and would change only if there is
>> conclusive
>>> evidence from satellite and mapillary imageries.
>>>
>>> We will reach out to local mapping experts through forum and/or
>> changeset
>>> comments if we require further help.
>>>
>>> Thank you all once again for the suggestions, we look forward to
>> working
>>> with you all. :)
>>>
>>>
>>> On Sun, 16 Aug 2020 at 05:35, cleary <osm at 97k.com> wrote:
>>>
>>>> Thanks for the interest in mapping in Australia and thanks for
>> posting
>>>> your plans on this list.
>>>>
>>>> I would add to the caution expressed by others.  I live in an urban 
>>>> location in Australia but I have travelled in other areas within 
>>>> Australia.  It has taken me quite some time to learn to interpret
>> satellite
>>>> imagery and I still have a lot to learn about this country.  After 
>>>> personally visiting areas and noting what I see, and sometimes
>> taking
>>>> photographs, I then return home and compare my notes with what I see
>> in the
>>>> imagery and I am still surprised.  I think it can be quite
>> precarious to
>>>> map features using just satellite imagery unless you have expert
>> assistance
>>>> in interpreting the imagery.  For example, a common error by others
>> has
>>>> been to map lines of cleared vegetation as roads when they are
>> actually
>>>> fences. Even where an unmapped road exists, it is probably still
>> unmapped
>>>> because it is a private road and not accessible by the public - many
>> of the
>>>> roads on rural properties in Australia are private and, if added to
>> the
>>>> map, need to marked as such. Farmers get annoyed about intruders on
>> their
>>>> farms especially as biosecurity is a significant concern in parts of 
>>>> Australia.
>>>>
>>>> So while I appreciate contributions to the map, I suggest that
>> "armchair"
>>>> mapping needs to be undertaken with a lot of caution.
>>>>
>>>>
>>>>
>>>>
>>>> On Sat, 15 Aug 2020, at 2:17 AM, OSM NextBillion. AI wrote:
>>>>> Hi all,
>>>>>
>>>>>
>>>>>
>>>>> We’re a small team based out of Hyderabad, India. We would be
>> doing
>>>>> minimal edits in Perth and contribute to OSM in the next couple of 
>>>>> weeks, in-line with OSM and Australia specific tagging guidelines
>> [Link
>> <https://wiki.openstreetmap.org/wiki/Australian_Tagging_Guidelines>].
>>>>>
>>>>> Please refer our Wiki
>>>>> <https://wiki.openstreetmap.org/wiki/NextBillion.AI-OSM> and
>> Github
>>>>> <https://github.com/NextBillionAI-OSM/OSM/issues/3> project pages
>> for
>>>>> more information.
>>>>>
>>>>> Looking forward to suggestions, if any ☺
>>>>>
>>>>> Thanking you in advance,
>>>>> Team NextBillion
>>>>> _______________________________________________
>
> _______________________________________________
> Talk-au mailing list
> Talk-au at openstreetmap.org
> https://lists.openstreetmap.org/listinfo/talk-au