Re: [Talk-GB] UK Quarterly Projects

[SK53]

A while back I played with trying to create watersheds
 using available OSM data. Each group of
waterways which interconnect (note I used st_intersects() on osm2pgsql
data, not the true topological relations in way_nodes) is given a different
colour. The map gives a fairly good synoptic overview of density and
completeness of waterway mapping. Of course it doesn't show accuracy, and
much of the dense coverage in Wales was done by Steve Chilton many years
ago from NPE maps.

Also I didnt try and handle rivers which pass through a lake or
rivers/streams only mapped as areas (this is really obvious on the
equivalent map for Ireland as the Shannon appears as several disconnected
watersheds). For this reason, and concave hull throwing errors on PostGis I
have not followed through with the aim of creating watershed polygons.

Jerry

On 14 March 2016 at 21:09, Andy Townsend  wrote:

> On 14/03/2016 20:50, Rob Nickerson wrote:
>
>> My concern with rivers is that we don't have tools to measure progress -
>> taginfo gives a count of nodes/ways/relations whereas we'd want total
>> length of features per region (in miles/km).
>>
>
> I'd agree that tracking progress with that one would be tricky. I've
> mapped lots of rivers and streams from a combination of survey, imagery and
> OS OpenData, and while we've got something approaching the "length" of
> major rivers and streams, what we've got in some areas is largely ex-NPE,
> and in many cases some distance from the actual waterway on the ground.
>
> GPS traces (unless there are lots) and Bing imagery can of course be
> misplaced, but OS OpenData is normally pretty good (actually better than
> the OS vector data that people have imported in a couple of areas) although
> it can be wrong when watercourses have changed, and the top end of small
> Welsh streams is often a bit "wishful thinking" in OSSV - where in reality
> there's just a boggy mess the OS sometimes has well-defined streams.
>
> The other problem with the waterways we've got in OSM is that many are
> just either "stream" or "river" - with things that I'd normally map as
> drains and ditches just in as "stream".  Obviously changing a tag
> post-survey is pretty straightforward, but it's something else to bear in
> mind.
>
> There is a real benefit of having OSSV streams, ditches and drains in
> though - it's often clear that a watercourse hasn't moved for years, and it
> can then be used to help align imagery and GPS traces.
>
> Cheers,
>
> Andy (SomeoneElse)
>
>
>
> ___
> Talk-GB mailing list
> Talk-GB@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/talk-gb
>
___
Talk-GB mailing list
Talk-GB@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-gb

Re: [Talk-GB] UK Quarterly Projects

[Lester Caine]

On 14/03/16 21:09, Andy Townsend wrote:
> There is a real benefit of having OSSV streams, ditches and drains in
> though - it's often clear that a watercourse hasn't moved for years, and
> it can then be used to help align imagery and GPS traces.

That would be useful where the source has been properly tagged. The ons
I had problems with often had no source, although NPE ones were fairly
obvious when one zoomed out.

-- 
Lester Caine - G8HFL
-
Contact - http://lsces.co.uk/wiki/?page=contact
L.S.Caine Electronic Services - http://lsces.co.uk
EnquirySolve - http://enquirysolve.com/
Model Engineers Digital Workshop - http://medw.co.uk
Rainbow Digital Media - http://rainbowdigitalmedia.co.uk

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

Re: [Talk-GB] UK Quarterly Projects

[Andy Townsend]

On 14/03/2016 20:50, Rob Nickerson wrote:
My concern with rivers is that we don't have tools to measure progress 
- taginfo gives a count of nodes/ways/relations whereas we'd want 
total length of features per region (in miles/km).


I'd agree that tracking progress with that one would be tricky. I've 
mapped lots of rivers and streams from a combination of survey, imagery 
and OS OpenData, and while we've got something approaching the "length" 
of major rivers and streams, what we've got in some areas is largely 
ex-NPE, and in many cases some distance from the actual waterway on the 
ground.


GPS traces (unless there are lots) and Bing imagery can of course be 
misplaced, but OS OpenData is normally pretty good (actually better than 
the OS vector data that people have imported in a couple of areas) 
although it can be wrong when watercourses have changed, and the top end 
of small Welsh streams is often a bit "wishful thinking" in OSSV - where 
in reality there's just a boggy mess the OS sometimes has well-defined 
streams.


The other problem with the waterways we've got in OSM is that many are 
just either "stream" or "river" - with things that I'd normally map as 
drains and ditches just in as "stream".  Obviously changing a tag 
post-survey is pretty straightforward, but it's something else to bear 
in mind.


There is a real benefit of having OSSV streams, ditches and drains in 
though - it's often clear that a watercourse hasn't moved for years, and 
it can then be used to help align imagery and GPS traces.


Cheers,

Andy (SomeoneElse)



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

Re: [Talk-GB] UK Quarterly Projects

[Rob Nickerson]

My concern with rivers is that we don't have tools to measure progress -
taginfo gives a count of nodes/ways/relations whereas we'd want total
length of features per region (in miles/km).

Would anyone on this list be able to produce an online tracking tool?

*Rob*

On 14 March 2016 at 20:47, Rob Nickerson  wrote:

> Thanks Brian,
>
> I've also put this on Loomio [1] 
>
[1] https://www.loomio.org/d/xagnZoK4
>
> *Rob*
>
___
Talk-GB mailing list
Talk-GB@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-gb

Re: [Talk-GB] UK Quarterly Projects

[Rob Nickerson]

Thanks Brian,

I've also put this on Loomio [1] because at some stage we will need to
decide and I find it difficult to interpret messages on talk-gb which often
provide some pro, some cons but then don't say yes/no clearly enough to
judge opinion.

For those interested, the healthcare data is all available at
http://systems.hscic.gov.uk/data/ods/datadownloads

I would propose limiting ourselves to just a small, manageable section.
[1] https://www.loomio.org/d/xagnZoK4

*Rob*
___
Talk-GB mailing list
Talk-GB@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-gb

Re: [Talk-GB] UK Quarterly Projects

[SK53]

Anything covered by FHRS has decent Open Data coverage across the country:
hospitals, care homes, pharmacies.

Charity Care Commission Open Data has a significant overlap with doctors,
dentists, care homes, clinics etc.

Both contain postcodes so one can locate things approximately, but in many
cases there's plenty of other visual cues on aerial imagery or OSSV to
locate these. For smaller GP practices and many dentists a survey may be
necessary because they will be run from an ordinary house.

Some councils provide Open Data listings of Places of Worship (Nottingham
for instance). Be aware that many smaller christian groups and many mosques
may be in atypical premises. I have come across a number of Hindu temples &
Gurdwaras making use of former industrial buildings. The New Basford
Industrial area  in Nottingham
has a surprising number of small places of worship which we would not have
found without the council data.

Although OSSV shows most Places of Worship, unless they are obviously
anglican churches by position, size, and presence in a churchyard, it is
difficult to deduce much more about them. I don't know if OpenMap Local has
any more information. For instance I gather that the distribution of
unitarian churches in Wales is rather interesting, but the information we
have is scant in the extreme.

It should be possible to locate the older established churches in England &
Scotland (and also Church in Wales) through use of old maps &
out-of-copyright guides. GSGS 3906 is very useful for Northern Ireland as
it often includes both the name & denomination of churches. Newer churches
in Northern Ireland may be recognised both by structure & large car parks,
especially when in the rural countryside. As an example there are currently
just over 2400 places of worship in Wales. There are/were about 1400
parishes many of which would have had several non-conformist chapels as
well as the parish church. Only around 350 had denomination=anglican, and
1900 religion=christian. (An additional problem in Wales is that churches,
and non-conformist chapels in particular, have been closing quite rapidly
over the past couple of decades). Finding ways to improve tagging of
existing places of worship, whether through survey or using other
resources, would be useful in its own right.

Jerry

On 14 March 2016 at 14:15, Stuart Reynolds  wrote:

> It might also be worth doing some kind of survey to see what motivated
> participants in teh Schools project, what they liked, what they didn't
> like, how we could improve, how they heard of the project (if indeed some
> of the single editors were even aware they were being counted as part of a
> project)
>
>
> My interest lies primarily with points of interest data. As many of you
> may recall, I work for one of the traveline regions for public transport
> journey planning. We are most interested, therefore, in the places that
> people actually want to travel to. Hospitals and other medical facilities
> are obviously one of those, as are schools. Churches also seem to be a key
> destination, and I have had contact from organisations that are trying to
> create directories of Mosques and other places of worship.
>
> At present we buy in our point of interest data. However, we would like to
> use OSM because it is free, and because we don’t have to worry about POI
> locations moving or being missing - we can just create them in the map
> ourselves, and have it in our next data release. On that basis, I
> participated in the schools project because I was slightly horrified at the
> number of schools that were missing in a relatively small area like
> Southend. I didn’t participate in the previous project (which, if I recall,
> was postboxes) because I frankly don’t care about such “micro" objects when
> large objects like schools are missing, and for a future project I would be
> interested in mapping larger objects that we can reasonably accurately fix
> in space and track. So hospitals / medical facilities are one, places of
> worship another, and probably also parks / woods / nature reserves. These
> latter fall into the category of large areas which are not suited to
> points, but where there are defined access points and possibly central
> “information” points that could be considered “main entrances” or the focal
> point of the area.
>
> Regards,
> Stuart
>
> 
> Stuart Reynolds
> for traveline south east & anglia
>
>
>
> On 13 Mar 2016, at 17:31, Brian Prangle  wrote:
>
> Hi everyone
>
> Following the extraordinary success of the Schools project we need  to
> decide how to proceed with the next quarter's project.
>
> There is a view we shoud rollover the project for another quarter so as to
> approach completion.
>
> I favour rolling it into an ongoing UK national project (similar to the
> Irish Townlands project), keeping the existing tools 

Re: [Talk-GB] UK Quarterly Projects

[Lester Caine]

On 13/03/16 17:31, Brian Prangle wrote:
> Following the extraordinary success of the Schools project we need  to
> decide how to proceed with the next quarter's project.
> 
> There is a view we shoud rollover the project for another quarter so as
> to approach completion.

Given that there is a clean set of data to compare against, keeping the
current tools available would be appreciated. I'm hoping to have another
session myself, but moving away from my local area, that will take more
time. The data is available so lets use it.

One thing I did find while working on the schools was a lot of poor
quality data that had to be tidied at the same time. Watercourses were
the main problem with crude blue lines running straight through the land
areas I was mapping. Some were fairly easy to relocate as traces of the
real path were visible, but I was tempted at times to simply delete
them, opting instead to push them to one side where the route was not
actually detectable ... unless they run under the school buildings :) So
is a more accurate set of data available for these?

What other databases such as edubase are available?

-- 
Lester Caine - G8HFL
-
Contact - http://lsces.co.uk/wiki/?page=contact
L.S.Caine Electronic Services - http://lsces.co.uk
EnquirySolve - http://enquirysolve.com/
Model Engineers Digital Workshop - http://medw.co.uk
Rainbow Digital Media - http://rainbowdigitalmedia.co.uk

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

Re: [Talk-GB] UK Quarterly Projects

[Stuart Reynolds]

It might also be worth doing some kind of survey to see what motivated 
participants in teh Schools project, what they liked, what they didn't like, 
how we could improve, how they heard of the project (if indeed some of the 
single editors were even aware they were being counted as part of a project)

My interest lies primarily with points of interest data. As many of you may 
recall, I work for one of the traveline regions for public transport journey 
planning. We are most interested, therefore, in the places that people actually 
want to travel to. Hospitals and other medical facilities are obviously one of 
those, as are schools. Churches also seem to be a key destination, and I have 
had contact from organisations that are trying to create directories of Mosques 
and other places of worship.

At present we buy in our point of interest data. However, we would like to use 
OSM because it is free, and because we don’t have to worry about POI locations 
moving or being missing - we can just create them in the map ourselves, and 
have it in our next data release. On that basis, I participated in the schools 
project because I was slightly horrified at the number of schools that were 
missing in a relatively small area like Southend. I didn’t participate in the 
previous project (which, if I recall, was postboxes) because I frankly don’t 
care about such “micro" objects when large objects like schools are missing, 
and for a future project I would be interested in mapping larger objects that 
we can reasonably accurately fix in space and track. So hospitals / medical 
facilities are one, places of worship another, and probably also parks / woods 
/ nature reserves. These latter fall into the category of large areas which are 
not suited to points, but where there are defined access points and possibly 
central “information” points that could be considered “main entrances” or the 
focal point of the area.

Regards,
Stuart


Stuart Reynolds
for traveline south east & anglia



On 13 Mar 2016, at 17:31, Brian Prangle 
> wrote:

Hi everyone

Following the extraordinary success of the Schools project we need  to decide 
how to proceed with the next quarter's project.

There is a view we shoud rollover the project for another quarter so as to 
approach completion.

I favour rolling it into an ongoing UK national project (similar to the Irish 
Townlands project), keeping the existing tools in place to monitor progress. As 
a byproduct do a major revamp of the UK projects wiki page to bring it up to 
date.

As for the subject for next quarter's project if we move the Schools project to 
a national project and don't roll it over, a number of ideas have been put 
forward:

1. Water: add sewage works, rivers,streams and ponds from OSSV, improve lake 
outlines, improve alignments, separate natural and reservoir tags,improve 
coastlines, add bridges or tunnels where waterways cross highways and railways 
etc.
2. Healthcare: add hospitals, doctors, dentists, pharmacies. Or possibly just 
add doctors which is a smaller target. I think national single Open Data 
sources exist for all of these. It will need more surveying than Schools (not 
visible on OSSV or aerial imagery)
3. Highway Maxheights, maxwidths, maxweights: various data sources exist but I 
don't  think there's one national resource. Improving this data will make our 
data more usable for routing. There is the possibility to involve couriers, 
haulage companies etc either organisationally, or individual drivers
4.There's also a suggestion to group-mentor a GSoC (Google Summer of Code) 
project, which I believe doesn't fit well as a quarterly project, as it won't 
involve a wide swathe of the community. We could explore it as an additional 
project
5. And a light-hearted suggestion - map some monkey puzzle trees as part of 
this project

It might also be worth doing some kind of survey to see what motivated 
participants in teh Schools project, what they liked, what they didn't like, 
how we could improve, how they heard of the project (if indeed some of the 
single editors were even aware they were being counted as part of a project)

As ever: opinions and other suggestions welcome

Regards

Brian

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

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

Re: [Talk-GB] UK Quarterly Projects

[Jez Nicholson]

For the quarterly project, would you say that we're looking for something
that does all of the following?:
* include/attract lots of people
* improve the UK map in either accuracy or coverage
* attract outside attention
* be sort-of interesting to us

On Sun, 13 Mar 2016 at 17:33 Brian Prangle  wrote:

> Hi everyone
>
> Following the extraordinary success of the Schools project we need  to
> decide how to proceed with the next quarter's project.
>
> There is a view we shoud rollover the project for another quarter so as to
> approach completion.
>
> I favour rolling it into an ongoing UK national project (similar to the
> Irish Townlands project), keeping the existing tools in place to monitor
> progress. As a byproduct do a major revamp of the UK projects wiki page to
> bring it up to date.
>
> As for the subject for next quarter's project if we move the Schools
> project to a national project and don't roll it over, a number of ideas
> have been put forward:
>
> 1. Water: add sewage works, rivers,streams and ponds from OSSV, improve
> lake outlines, improve alignments, separate natural and reservoir
> tags,improve coastlines, add bridges or tunnels where waterways cross
> highways and railways etc.
> 2. Healthcare: add hospitals, doctors, dentists, pharmacies. Or possibly
> just add doctors which is a smaller target. I think national single Open
> Data sources exist for all of these. It will need more surveying than
> Schools (not visible on OSSV or aerial imagery)
> 3. Highway Maxheights, maxwidths, maxweights: various data sources exist
> but I don't  think there's one national resource. Improving this data will
> make our data more usable for routing. There is the possibility to involve
> couriers, haulage companies etc either organisationally, or individual
> drivers
> 4.There's also a suggestion to group-mentor a GSoC (Google Summer of Code)
> project, which I believe doesn't fit well as a quarterly project, as it
> won't involve a wide swathe of the community. We could explore it as an
> additional project
> 5. And a light-hearted suggestion - map some monkey puzzle trees as part
> of this project 
>
> It might also be worth doing some kind of survey to see what motivated
> participants in teh Schools project, what they liked, what they didn't
> like, how we could improve, how they heard of the project (if indeed some
> of the single editors were even aware they were being counted as part of a
> project)
>
> As ever: opinions and other suggestions welcome
>
> Regards
>
> Brian
>
> ___
> Talk-GB mailing list
> Talk-GB@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/talk-gb
>
___
Talk-GB mailing list
Talk-GB@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-gb

[Talk-GB] UK Quarterly Projects

[Brian Prangle]

Hi everyone

Following the extraordinary success of the Schools project we need  to
decide how to proceed with the next quarter's project.

There is a view we shoud rollover the project for another quarter so as to
approach completion.

I favour rolling it into an ongoing UK national project (similar to the
Irish Townlands project), keeping the existing tools in place to monitor
progress. As a byproduct do a major revamp of the UK projects wiki page to
bring it up to date.

As for the subject for next quarter's project if we move the Schools
project to a national project and don't roll it over, a number of ideas
have been put forward:

1. Water: add sewage works, rivers,streams and ponds from OSSV, improve
lake outlines, improve alignments, separate natural and reservoir
tags,improve coastlines, add bridges or tunnels where waterways cross
highways and railways etc.
2. Healthcare: add hospitals, doctors, dentists, pharmacies. Or possibly
just add doctors which is a smaller target. I think national single Open
Data sources exist for all of these. It will need more surveying than
Schools (not visible on OSSV or aerial imagery)
3. Highway Maxheights, maxwidths, maxweights: various data sources exist
but I don't  think there's one national resource. Improving this data will
make our data more usable for routing. There is the possibility to involve
couriers, haulage companies etc either organisationally, or individual
drivers
4.There's also a suggestion to group-mentor a GSoC (Google Summer of Code)
project, which I believe doesn't fit well as a quarterly project, as it
won't involve a wide swathe of the community. We could explore it as an
additional project
5. And a light-hearted suggestion - map some monkey puzzle trees as
part of this
project 

It might also be worth doing some kind of survey to see what motivated
participants in teh Schools project, what they liked, what they didn't
like, how we could improve, how they heard of the project (if indeed some
of the single editors were even aware they were being counted as part of a
project)

As ever: opinions and other suggestions welcome

Regards

Brian
___
Talk-GB mailing list
Talk-GB@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-gb