Re: [talk-au] Mapbox mapping activity

[Alex (Maxious) Sadleir]

It's not just us Aussies left scratching our heads, several country
mailing lists got this message.

DWG is aware https://github.com/mapbox/mapping/issues/409
> Please answer (on this issue and every other country issue) the same 
> questions that were asked at #387 , and make sure that a local mailing list 
> also sees those answers.
> Simply posting the same link without explanation for every country is not 
> good enough.
> -- Andy (SomeoneElse, from OSM's Data Working Group)
> Communication isn't occurring - hence the angry reply at 
> http://lists.openstreetmap.ch/pipermail/talk-ch/2022-April/011470.html ,
> Another is that the quality of some of Mapbox's mapping doesn't reflect 
> https://wiki.openstreetmap.org/wiki/How_Mapbox_Data_RAVE_team_work_on_receiving_feedback_from_OSM_users
>  . The DWG have had a number of issues recently where Mapbox editors have 
> changed something from "wrong, and detected by a QA metric" to "differently 
> wrong, but not detected by a QA metric". In the long term editing like this 
> harms the quality of OSM because it prevents mappers fixing problems properly.
> Obviously we (the DWG) raise issues with individual mappers and if necessary 
> with Mapbox directly but the worrying thing is that Mapbox as a company 
> appears not to be learning from its mistakes - Mapbox has been involved with 
> (and dependent on) OSM data for many years now but we're still seeing 
> examples of really poor quality mapping, of the "just remove a tag so that a 
> detected error goes away, never mind if it actually matches the actual 
> situation" variety.


On Sat, Apr 23, 2022 at 9:35 AM stevea  wrote:
>
> OK, I've found the links, but a couple of things:
>
> 1)  On the wiki page (about "linting," I've used "lint" in C code, is that 
> what you mean?), I don't want to see "examples," (although, they are better 
> than nothing), I want to see SPECIFICS.
>
> 2)  The GitHub page is also pretty vague.  You say "we use internal Mapbox 
> data validation jobs that search OSM data for errors for detecting issues 
> related to road network data in Australia. In total, we are going to review 
> issues in 9 categories."
>
> OK, WHAT 9 categories?
>
> I appreciate that you are "offering notice," but please be less vague.
>
> > On Apr 22, 2022, at 12:08 AM, Sergey Beliamei via Talk-au 
> >  wrote:
> >>
> >> Hello from the Mapbox Team!
> >> In April 2022 our team is going to start a mapping project in Australia. 
> >> As part of on-going work to improve the quality of OpenStreetMap data, our 
> >> team is planning to review a subset of the detections to better understand 
> >> the type of issues, and also fix any valid data issues directly in the 
> >> OSM. We're concentrating on road mistakes to improve map condition. In 
> >> Australia
> >> we plan to upload data once a week to see and fix the latest mistakes in 
> >> mapping after reviewing the 1st iteration.
> >> We would really appreciate your feedback, any questions you have about this
> >> project, as well as local insights that you think will help us better 
> >> understand the data.
> >> There's a link to our Github ticket to the related issue and a link to our 
> >> page in OSM Wiki.
> >>
> >>
> >> Best regards,
> >> Member of Mapbox team,
> >> Sergey
> >
>
>
> ___
> Talk-au mailing list
> Talk-au@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/talk-au

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

Re: [talk-au] Country homesteads?

[Graeme Fitzpatrick]

Thanks all!

Yeah, the various wiki comments re size as always are very European & don't
suit Oz conditions.

I think that place=farm + name=* seems to be the way to go, so I'll test
that & see how it works?

One of the factors in OP making them hamlets could have been that that will
render, so will have to see if =farm does as well?

Thanks

Graeme


On Fri, 22 Apr 2022 at 18:39, cleary  wrote:

>
> Generally, I would suggest a node at the hub of the farm (usually in the
> vicinity of the main residence)
> place=farm
> name=*
> operator=*
>
> In regard to farm boundaries, I think it has been OSM practice not to map
> lot/property boundaries.
>
> Most farms today don't employ many people and employees frequently live in
> nearby towns etc. so place=farm is appropriate.  In more isolated areas,
> some very large farms may still accommodate workers and sometimes their
> families - in such cases place=isolated_dwelling might be appropriate as I
> think it implies a larger population than a farm. I'd suggest that it has
> many years since Australian farms had enough residents to be considered as
> hamlets. In fact, increasingly farms have no residents. In some cases
> multiple farms are operated from a single homestead but, if each farm is
> separately named, I think each should still be separately mapped. Farms
> with animals such as dairy cattle or poultry etc may need people on site
> overnight but grazing cattle and crops such as cotton/wheat/rice etc may
> not need people in the immediate vicinity - and security cameras and alarms
> are increasingly used so that people can go to off-farm homes at night. The
> "homestead" may become an administrative office plus staff facilities when
> there is no-one resident on the property.
>
> In some cases, where public roads go through farms (usually cattle grids
> and signs at the respective boundaries), I have added  is_in:farm=*  on the
> section of road that is within the particular farm, but the "is_in" tag
> seems now less used than in the past.  Perhaps there is a case for mapping
> lot/property boundaries where the properties are very large farms but I
> will leave that for others to advocate. In South Australia's pastoral
> district, each farm has its own "suburb" boundaries in the official
> government suburb/locality database - but I am not aware of any farm
> boundaries designated in this way in any other state.
>
>
>
>
> On Fri, 22 Apr 2022, at 3:27 PM, Bob Cameron wrote:
> > Remote areas and larger farms generally have been troubling me too
> > Graeme. I make no distinction about numbers of people, just a
> > landuse=farm node. (so I copied a very prolific mapper!) Recently I
> > noted that landuse:farm has been deprecated and to use
> > landuse:farmland, but that complains about being a node. There is no
> > easy way to define a farm boundary. I think in terms of the mailbox,
> > driveway and largest concentration of activity being the node centre.
> >
> > And the name is the farm name, not the house name.. maybe!
> >
> > Remote cattle stations can support an extended family (in more than one
> > homestead) and other many onsite (staff) people. Are the working farm
> > staff include in any people sizing calculations? ouch!
> >
> > Personally I don't think it a good idea to tag a farm that creates
> > commercial income with any notion of the number of people. It gets a
> > bit blurry when it is an unusual group like a religious order or non
> > profit retreat, but they already have other tags.
> >
> > Cheers Bob
> >
> > On 22/4/22 14:55, Graeme Fitzpatrick wrote:
> >> Also bringing discussion out here from Discord.
> >>
> >> An anonymous user is hitting Notes with quite a few entries yesterday
> to say that remote homesteads are incorrectly tagged as hamlets eg
> https://www.openstreetmap.org/note/3145380, but looking at this
> particular place
> https://www.openstreetmap.org/edit?node=1829712552#map=17/-21.96106/148.80882,
> I'd say that "hamlet" was probably correct in that there could well be a
> couple of families living there?
> >>
> >> Other suggestions that have been made are place=isolated_dwelling or
> place=farm.
> >>
> >> Bit of a grey area, I guess? Isolated-dwelling says 1-2 families only,
> hamlet says 100-200 people, while place=farm says "a family of farmers".
> Guess it really depends on the particular property involved, which would
> require detailed local knowledge?
> >>
> >> Thoughts?
> >>
> >> Thanks
> >>
> >> Graeme
> >>
> >> ___
> >> Talk-au mailing list
> >> Talk-au@openstreetmap.org
> >> https://lists.openstreetmap.org/listinfo/talk-au
> > ___
> > Talk-au mailing list
> > Talk-au@openstreetmap.org
> > https://lists.openstreetmap.org/listinfo/talk-au
>
> ___
> Talk-au mailing list
> Talk-au@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/talk-au
>

Re: [talk-au] Mapbox mapping activity

[stevea]

OK, I've found the links, but a couple of things:

1)  On the wiki page (about "linting," I've used "lint" in C code, is that what 
you mean?), I don't want to see "examples," (although, they are better than 
nothing), I want to see SPECIFICS.

2)  The GitHub page is also pretty vague.  You say "we use internal Mapbox data 
validation jobs that search OSM data for errors for detecting issues related to 
road network data in Australia. In total, we are going to review issues in 9 
categories."

OK, WHAT 9 categories?

I appreciate that you are "offering notice," but please be less vague.

> On Apr 22, 2022, at 12:08 AM, Sergey Beliamei via Talk-au 
>  wrote:
>> 
>> Hello from the Mapbox Team!
>> In April 2022 our team is going to start a mapping project in Australia. As 
>> part of on-going work to improve the quality of OpenStreetMap data, our team 
>> is planning to review a subset of the detections to better understand the 
>> type of issues, and also fix any valid data issues directly in the OSM. 
>> We're concentrating on road mistakes to improve map condition. In Australia
>> we plan to upload data once a week to see and fix the latest mistakes in 
>> mapping after reviewing the 1st iteration.
>> We would really appreciate your feedback, any questions you have about this
>> project, as well as local insights that you think will help us better 
>> understand the data.
>> There's a link to our Github ticket to the related issue and a link to our 
>> page in OSM Wiki.
>> 
>> 
>> Best regards,
>> Member of Mapbox team,
>> Sergey
> 


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

Re: [talk-au] Mapbox mapping activity

[stevea]

Mmmm, yeah:  I'll second Graeme's "that doesn't make much sense to me."

What does "review a subset of the detections to better understand the types of 
issues" actually mean?

And, "1st iteration"...of WHAT?

What data?

How can we give feedback on a "related issue" that is neither stated nor 
"linked to (your) Github ticket"?  Which is NOT linked here?!

Try again, please, Sergey.  That initial email of yours was a non-starter.

> On Apr 22, 2022, at 12:08 AM, Sergey Beliamei via Talk-au 
>  wrote:
> 
> Hello from the Mapbox Team!
>  In April 2022 our team is going to start a mapping project in Australia. As 
> part of on-going work to improve the quality of OpenStreetMap data, our team 
> is planning to review a subset of the detections to better understand the 
> type of issues, and also fix any valid data issues directly in the OSM. We're 
> concentrating on road mistakes to improve map condition. In Australia
> we plan to upload data once a week to see and fix the latest mistakes in 
> mapping after reviewing the 1st iteration.
>  We would really appreciate your feedback, any questions you have about this
> project, as well as local insights that you think will help us better 
> understand the data.
> There's a link to our Github ticket to the related issue and a link to our 
> page in OSM Wiki.
> 
> 
> Best regards,
> Member of Mapbox team,
> Sergey


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

Re: [talk-au] Mapbox mapping activity

[Graeme Fitzpatrick]

On Fri, 22 Apr 2022 at 17:20, Sergey Beliamei via Talk-au <
talk-au@openstreetmap.org> wrote:

> our team is planning to review a subset of the detections to better
> understand the type of issues, and also fix any valid data issues directly
> in the OSM. We're concentrating on road mistakes to improve map condition.
>

Sorry, but to me at least, that doesn't make much sense?

Could you please explain what you're trying to do?

Thanks

Graeme
___
Talk-au mailing list
Talk-au@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-au

Re: [talk-au] Creating a Mapping Team?

[Andrew Harvey]

The organised editing guidelines at
https://wiki.osmfoundation.org/wiki/Organised_Editing_Guidelines should be
followed if the whole team is working together on mapping activities in a
coordinated way. Any questions feel free to ask.

On Fri, 22 Apr 2022 at 14:30, Andrew Hughes  wrote:

> Hi All,
>
> Would anyone be able to please point me in the right direction? I am
> looking to learn more about mapping teams and how to create one for the
> National Heavy Vehicle Regulator  https://www.nhvr.gov.au/
>
> I have looked high and low, but I can't anything.
>
> Thanks in advance,
> Andrew
>
>
> ___
> Talk-au mailing list
> Talk-au@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/talk-au
>
___
Talk-au mailing list
Talk-au@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-au

Re: [talk-au] Country homesteads?

[Warin]

Traditionally the 'homestead' is the main residential building and is 
usually named the same or similar to the farm. I have been mapping in 
NSW the building as building=farm and name=x from the DCS Base Map for 
location and name with DCS Imagery for the footprint. When you zoom in 
on the Base Map the name will be associated with one building by 
proximity and the building will be a filled in black square, other non 
residential buildings tend to be hollow black squares.


The homestead might get 'homestead' appended to the name, and the farm 
might have appended 'Station' or 'Down'. I think this depends on how old 
you are and your local knowledge. My first visit many years ago to Mount 
Wood in the north west of NSW is a case in point .. I was told 'Mount 
Wood Station' for the farm and 'Mount Wood Homestead' for the 
residence... IIRC 'Station' is for cattle and 'Down' is for sheep...


Outback farms tend to have the mail box on the main road and that can be 
quite some distance from the homestead.



On 22/4/22 15:27, Bob Cameron wrote:


Remote areas and larger farms generally have been troubling me too 
Graeme. I make no distinction about numbers of people, just a 
landuse=farm node. (so I copied a very prolific mapper!) Recently I 
noted that landuse:farm has been deprecated and to use 
landuse:farmland, but that complains about being a node. There is no 
easy way to define a farm boundary. I think in terms of the mailbox, 
driveway and largest concentration of activity being the node centre.


And the name is the farm name, not the house name.. maybe!

Remote cattle stations can support an extended family (in more than 
one homestead) and other many onsite (staff) people. Are the working 
farm staff include in any people sizing calculations? ouch!


Personally I don't think it a good idea to tag a farm that creates 
commercial income with any notion of the number of people. It gets a 
bit blurry when it is an unusual group like a religious order or non 
profit retreat, but they already have other tags.


Cheers Bob

On 22/4/22 14:55, Graeme Fitzpatrick wrote:

Also bringing discussion out here from Discord.

An anonymous user is hitting Notes with quite a few entries yesterday 
to say that remote homesteads are incorrectly tagged as hamlets eg 
https://www.openstreetmap.org/note/3145380 
, but looking at this 
particular place 
https://www.openstreetmap.org/edit?node=1829712552#map=17/-21.96106/148.80882 
, 
I'd say that "hamlet" was probably correct in that there could well 
be a couple of families living there?


Other suggestions that have been made are place=isolated_dwelling or 
place=farm.


Bit of a grey area, I guess? Isolated-dwelling says 1-2 families 
only, hamlet says 100-200 people, while place=farm says "a family of 
farmers". Guess it really depends on the particular property 
involved, which would require detailed local knowledge?


Thoughts?

Thanks

Graeme

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


___
Talk-au mailing list
Talk-au@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-au___
Talk-au mailing list
Talk-au@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-au

[talk-au] Mapbox mapping activity

[Sergey Beliamei via Talk-au]

Hello from the Mapbox Team!
 In April 2022 our team is going to start a mapping project in Australia.
As part of on-going work to improve the quality of OpenStreetMap data, our
team is planning to review a subset of the detections to better understand
the type of issues, and also fix any valid data issues directly in the OSM.
We're concentrating on road mistakes to improve map condition. In Australia
we plan to upload data once a week to see and fix the latest mistakes in
mapping after reviewing the 1st iteration.
 We would really appreciate your feedback, any questions you have about this
project, as well as local insights that you think will help us better
understand the data.
There's a link  to our Github
ticket to the related issue and a link

to our page in OSM Wiki.


Best regards,
Member of Mapbox team,
Sergey
___
Talk-au mailing list
Talk-au@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-au