where we at
86 views
Skip to first unread message
Hamish
Jan 27, 2013, 3:22:43 PM1/27/13
to nzopengis
Hi everyone,
Seems like a good time for a status update. We're finally in production
mode, uploading data to the OSM servers in earnest. There's still lots
of inter-tagging work to do, but we've enough ready to go now that we
won't be slowed down by that for a long while yet.
Glen continues to work on Northland, and has started with some roads! That's
some sort of long awaited major milestone for this project.
Steve continues to quietly improve things and bring more online, I don't
presume to speak for them so I'll let him and Glen comment on their own
about what they're up to if they like, just to say that I'm excited to see all
the progress.
The Antipodes are now completely uploaded:
http://www.openstreetmap.org/?lat=-49.6795&lon=178.77&zoom=13&layers=M
The river_cl -> stream went pretty well. The geographic_names layer containing
the stream names show up as red circles with lines through them in JOSM, so
are easy to spot & apply to the way tagging. There is plenty of merging
duplicate nodes at the confluences (can we automate that in the web app?)
and way reversals to get them all pointing downstream, so one more tagging
action really isn't too much of a burden.
Much of the land cover for Stewart Island is now up:
http://www.openstreetmap.org/?lat=-46.993&lon=167.809&zoom=10&layers=M
The day-walk tracks to the west of Oban are not uploaded; there was significant
disagreement between what was already there and what was in the to-be-uploaded
data that I pass that on to someone with a GPS when they're down there next. The
bush is thick enough to make the satellite images not much help.
Waiheke is doing pretty well:
http://www.openstreetmap.org/?lat=-36.8011&lon=175.0958&zoom=13&layers=M
Great Barrier seems well mapped out -- turns out it was done with the LINZ
data but without any attribution or data versioning, so we'll have to redo
the whole thing. (not as much work as it sounds, but unfortunate to have to
double handle it)
Good progress on native bush in the Tararuas and Cormandel:
http://www.openstreetmap.org/?lat=-40.718&lon=175.347&zoom=10&layers=M
http://www.openstreetmap.org/?lat=-36.871&lon=175.593&zoom=10&layers=M
There's continuing effort to upload land cover for the lower North
Island and (less so)
around the Franz Joseph glacier. As always, any and all help is most welcome;
it's pretty quick to get your favourite small town on the map with all
sorts of detail.
Bet the first in your town to... etc etc (hey can we get local
chambers of commerce/
rotary groups involved?)
And if anyone was wondering, our Antarctic data doesn't go as far south as the
Queen Alexandra Range (AFAICT).
regards,
Hamish
Seems like a good time for a status update. We're finally in production
mode, uploading data to the OSM servers in earnest. There's still lots
of inter-tagging work to do, but we've enough ready to go now that we
won't be slowed down by that for a long while yet.
Glen continues to work on Northland, and has started with some roads! That's
some sort of long awaited major milestone for this project.
Steve continues to quietly improve things and bring more online, I don't
presume to speak for them so I'll let him and Glen comment on their own
about what they're up to if they like, just to say that I'm excited to see all
the progress.
The Antipodes are now completely uploaded:
http://www.openstreetmap.org/?lat=-49.6795&lon=178.77&zoom=13&layers=M
The river_cl -> stream went pretty well. The geographic_names layer containing
the stream names show up as red circles with lines through them in JOSM, so
are easy to spot & apply to the way tagging. There is plenty of merging
duplicate nodes at the confluences (can we automate that in the web app?)
and way reversals to get them all pointing downstream, so one more tagging
action really isn't too much of a burden.
Much of the land cover for Stewart Island is now up:
http://www.openstreetmap.org/?lat=-46.993&lon=167.809&zoom=10&layers=M
The day-walk tracks to the west of Oban are not uploaded; there was significant
disagreement between what was already there and what was in the to-be-uploaded
data that I pass that on to someone with a GPS when they're down there next. The
bush is thick enough to make the satellite images not much help.
Waiheke is doing pretty well:
http://www.openstreetmap.org/?lat=-36.8011&lon=175.0958&zoom=13&layers=M
Great Barrier seems well mapped out -- turns out it was done with the LINZ
data but without any attribution or data versioning, so we'll have to redo
the whole thing. (not as much work as it sounds, but unfortunate to have to
double handle it)
Good progress on native bush in the Tararuas and Cormandel:
http://www.openstreetmap.org/?lat=-40.718&lon=175.347&zoom=10&layers=M
http://www.openstreetmap.org/?lat=-36.871&lon=175.593&zoom=10&layers=M
There's continuing effort to upload land cover for the lower North
Island and (less so)
around the Franz Joseph glacier. As always, any and all help is most welcome;
it's pretty quick to get your favourite small town on the map with all
sorts of detail.
Bet the first in your town to... etc etc (hey can we get local
chambers of commerce/
rotary groups involved?)
And if anyone was wondering, our Antarctic data doesn't go as far south as the
Queen Alexandra Range (AFAICT).
regards,
Hamish
Glen Barnes
Jan 28, 2013, 7:35:13 PM1/28/13
to nzop...@googlegroups.com
On Monday, 28 January 2013 09:22:43 UTC+13, Hamish wrote:
Glen continues to work on Northland, and has started with some roads! That's
some sort of long awaited major milestone for this project.
Steve continues to quietly improve things and bring more online, I don't
presume to speak for them so I'll let him and Glen comment on their own
about what they're up to if they like, just to say that I'm excited to see all
the progress.
I started with some road uploading during the weekend and at least in the Far North the new data is generally better or the same as the existing data. I've found it easier to keep the new data in all cases and be careful to make sure the road tags are kept the same. I have had to change some 'track' types to 'unclassified' as this better fits with the description (Most NZ public dirt roads would be unclassified). The only pain is joining the new segments of SH 1 to the massive State Highway 1 relation.
There is also a job to get rid of the big forest up north and make a relation of all the LINZ data. I'm still trying to find the LINZ layer which 'fills in' the other parts of the forest but have yet to find it. I wonder what the best way is to find "All the layers that cover this area"?
Hamish
Jan 28, 2013, 11:34:02 PM1/28/13
to nzop...@googlegroups.com
Glen:
ways, typically an inner hole in some relation and an area of bush or
something in that hole?
> I'm thinking there is probably some for of script we could run that would deal with these
> issues and I will probably show Stephen and Rob to see what they think.
I've just been leaving them for now, along with the overlapping
coastline/bush "duplicate nodes on unclosed way" which won't auto-fix
in bulk. I figure a tool within the Validator will come around in
time.
> There is also a job to get rid of the big forest up north and make a relation of all the LINZ
> data. I'm still trying to find the LINZ layer which 'fills in' the other parts of the forest but
> have yet to find it. I wonder what the best way is to find "All the layers that cover this area"?
nztopomaps.com is pretty good to let you glance at the topo50 sheet,
and you can do the same by adding a custom WMS layer in JOSM linking
to the LINZ data service.
for big forests often there's a single green "node" describing the
whole thing, you need to checkout the rectangle containing that green
dot, many of the rectangles around it turn up as 0 features even if
there is coverage from the neighboring rectangle.
Just having a peek at the topo50 sheets for northland, 90 mile beach
seems to be all exotic_poly, but you've already completed those
workslices. South of there the bush seems pretty well broken up and
about half pine trees, so the areas shouldn't get too huge.
So actually I'm not really sure where you are looking. URL?
If the LINZ data isn't already in a relation automatically, I wouldn't
bother putting it in one for the sake of being in one.
2c,
Hamish
> I've run into a few issues with having duplicate nodes where different polygons
> share that same outline and are part of a relation.
Do you mean the Validator complaining that there are two overlapping
> share that same outline and are part of a relation.
ways, typically an inner hole in some relation and an area of bush or
something in that hole?
> I'm thinking there is probably some for of script we could run that would deal with these
> issues and I will probably show Stephen and Rob to see what they think.
coastline/bush "duplicate nodes on unclosed way" which won't auto-fix
in bulk. I figure a tool within the Validator will come around in
time.
> There is also a job to get rid of the big forest up north and make a relation of all the LINZ
> data. I'm still trying to find the LINZ layer which 'fills in' the other parts of the forest but
> have yet to find it. I wonder what the best way is to find "All the layers that cover this area"?
and you can do the same by adding a custom WMS layer in JOSM linking
to the LINZ data service.
for big forests often there's a single green "node" describing the
whole thing, you need to checkout the rectangle containing that green
dot, many of the rectangles around it turn up as 0 features even if
there is coverage from the neighboring rectangle.
Just having a peek at the topo50 sheets for northland, 90 mile beach
seems to be all exotic_poly, but you've already completed those
workslices. South of there the bush seems pretty well broken up and
about half pine trees, so the areas shouldn't get too huge.
So actually I'm not really sure where you are looking. URL?
If the LINZ data isn't already in a relation automatically, I wouldn't
bother putting it in one for the sake of being in one.
2c,
Hamish
Glen Barnes
Jan 29, 2013, 4:17:44 AM1/29/13
to nzop...@googlegroups.com
On Tuesday, 29 January 2013 17:34:02 UTC+13, Hamish wrote:
Glen:
> I've run into a few issues with having duplicate nodes where different polygons
> share that same outline and are part of a relation.
Do you mean the Validator complaining that there are two overlapping
ways, typically an inner hole in some relation and an area of bush or
something in that hole?
Yes.
> I'm thinking there is probably some for of script we could run that would deal with these
> issues and I will probably show Stephen and Rob to see what they think.
I've just been leaving them for now, along with the overlapping
coastline/bush "duplicate nodes on unclosed way" which won't auto-fix
in bulk. I figure a tool within the Validator will come around in
time.
Right - The actual 'proper' solution is to:
- Merge ways that are < 2000 nodes combined to a single way.
- Move the tags to a relation if they are > 2000 nodes combined. This way you don't get an unclosed way issue.
> There is also a job to get rid of the big forest up north and make a relation of all the LINZ
> data. I'm still trying to find the LINZ layer which 'fills in' the other parts of the forest but
> have yet to find it. I wonder what the best way is to find "All the layers that cover this area"?
nztopomaps.com is pretty good to let you glance at the topo50 sheet,
and you can do the same by adding a custom WMS layer in JOSM linking
to the LINZ data service.
for big forests often there's a single green "node" describing the
whole thing, you need to checkout the rectangle containing that green
dot, many of the rectangles around it turn up as 0 features even if
there is coverage from the neighboring rectangle.
Just having a peek at the topo50 sheets for northland, 90 mile beach
seems to be all exotic_poly, but you've already completed those
workslices. South of there the bush seems pretty well broken up and
about half pine trees, so the areas shouldn't get too huge.
So actually I'm not really sure where you are looking. URL?
Basically someone previously has drawn Aupori Forest as a really big forrest and I don't want to blow it all away. I would prefer to make a relation with that name and add all the LINZ objects to it before deleting it.
The aerial imagery shows a bunch of different stuff but it is a big forestry area up there so it could be out of date.
Hamish
Jan 29, 2013, 4:29:59 AM1/29/13
to nzop...@googlegroups.com
Glen wrote:
> # Merge ways that are < 2000 nodes combined to a single way.
After dealing with cleaning up the horrible mess of ways that went
over the 2000 limit a number of times, I'd strongly argue to keep the
polylines at a starting point of ~495 nodes each. It gives you (i.e.
not really you, but some helpful 3rd party who doesn't know about the
limit) a bit of leeway to mix and match and combine without being too
splintered, but not too close to going over the cliff either.
> I would prefer to make a relation with that name and add all the LINZ objects to it before
> deleting it.
ok
Hamish
On 1/29/13, Glen Barnes <barnacl...@gmail.com> wrote:
>
>
> On Tuesday, 29 January 2013 17:34:02 UTC+13, Hamish wrote:
>>
>> Glen:
>> > I've run into a few issues with having duplicate nodes where different
>> polygons
>> > share that same outline and are part of a relation.
>>
>> Do you mean the Validator complaining that there are two overlapping
>> ways, typically an inner hole in some relation and an area of bush or
>> something in that hole?
>>
>
> Yes.
>
>
>>
>> > I'm thinking there is probably some for of script we could run that
>> would deal with these
>> > issues and I will probably show Stephen and Rob to see what they think.
>> >
>>
>> I've just been leaving them for now, along with the overlapping
>> coastline/bush "duplicate nodes on unclosed way" which won't auto-fix
>> in bulk. I figure a tool within the Validator will come around in
>> time.
>>
>
> Right - The actual 'proper' solution is to:
>
>
> - Merge ways that are < 2000 nodes combined to a single way.
> - Move the tags to a relation if they are > 2000 nodes combined. This
> You received this message because you are subscribed to the Google Groups
> "nzopengis" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to nzopengis+...@googlegroups.com.
> To post to this group, send email to nzop...@googlegroups.com.
> Visit this group at http://groups.google.com/group/nzopengis?hl=en.
> For more options, visit https://groups.google.com/groups/opt_out.
>
>
>
> # Merge ways that are < 2000 nodes combined to a single way.
After dealing with cleaning up the horrible mess of ways that went
over the 2000 limit a number of times, I'd strongly argue to keep the
polylines at a starting point of ~495 nodes each. It gives you (i.e.
not really you, but some helpful 3rd party who doesn't know about the
limit) a bit of leeway to mix and match and combine without being too
splintered, but not too close to going over the cliff either.
> I would prefer to make a relation with that name and add all the LINZ objects to it before
> deleting it.
Hamish
On 1/29/13, Glen Barnes <barnacl...@gmail.com> wrote:
>
>
> On Tuesday, 29 January 2013 17:34:02 UTC+13, Hamish wrote:
>>
>> Glen:
>> > I've run into a few issues with having duplicate nodes where different
>> polygons
>> > share that same outline and are part of a relation.
>>
>> Do you mean the Validator complaining that there are two overlapping
>> ways, typically an inner hole in some relation and an area of bush or
>> something in that hole?
>>
>
> Yes.
>
>
>>
>> > I'm thinking there is probably some for of script we could run that
>> would deal with these
>> > issues and I will probably show Stephen and Rob to see what they think.
>> >
>>
>> I've just been leaving them for now, along with the overlapping
>> coastline/bush "duplicate nodes on unclosed way" which won't auto-fix
>> in bulk. I figure a tool within the Validator will come around in
>> time.
>>
>
> Right - The actual 'proper' solution is to:
>
>
> - Move the tags to a relation if they are > 2000 nodes combined. This
> way you don't get an unclosed way issue.
>
>
>> > There is also a job to get rid of the big forest up north and make a
>> relation of all the LINZ
>> > data. I'm still trying to find the LINZ layer which 'fills in' the other
>> >
>> parts of the forest but
>> > have yet to find it. I wonder what the best way is to find "All the
>> layers that cover this area"?
>>
>> nztopomaps.com is pretty good to let you glance at the topo50 sheet,
>> and you can do the same by adding a custom WMS layer in JOSM linking
>> to the LINZ data service.
>>
>> for big forests often there's a single green "node" describing the
>> whole thing, you need to checkout the rectangle containing that green
>> dot, many of the rectangles around it turn up as 0 features even if
>> there is coverage from the neighboring rectangle.
>>
>> Just having a peek at the topo50 sheets for northland, 90 mile beach
>> seems to be all exotic_poly, but you've already completed those
>> workslices. South of there the bush seems pretty well broken up and
>> about half pine trees, so the areas shouldn't get too huge.
>> So actually I'm not really sure where you are looking. URL?
>>
>>
> <https://www.evernote.com/shard/s2/sh/cedc6f54-8811-4912-8145-23b2d3fe42c9/2aa734310d8eb415357369cba40474c5>
>
>
>> > There is also a job to get rid of the big forest up north and make a
>> relation of all the LINZ
>> > data. I'm still trying to find the LINZ layer which 'fills in' the other
>> >
>> parts of the forest but
>> > have yet to find it. I wonder what the best way is to find "All the
>> layers that cover this area"?
>>
>> nztopomaps.com is pretty good to let you glance at the topo50 sheet,
>> and you can do the same by adding a custom WMS layer in JOSM linking
>> to the LINZ data service.
>>
>> for big forests often there's a single green "node" describing the
>> whole thing, you need to checkout the rectangle containing that green
>> dot, many of the rectangles around it turn up as 0 features even if
>> there is coverage from the neighboring rectangle.
>>
>> Just having a peek at the topo50 sheets for northland, 90 mile beach
>> seems to be all exotic_poly, but you've already completed those
>> workslices. South of there the bush seems pretty well broken up and
>> about half pine trees, so the areas shouldn't get too huge.
>> So actually I'm not really sure where you are looking. URL?
>>
>>
>
> Yes
> -
> https://www.evernote.com/shard/s2/sh/cedc6f54-8811-4912-8145-23b2d3fe42c9/2aa734310d8eb415357369cba40474c5
>
> Basically someone previously has drawn Aupori Forest as a really big
> forrest and I don't want to blow it all away. I would prefer to make a
> relation with that name and add all the LINZ objects to it before deleting
> it.
>
> The aerial imagery shows a bunch of different stuff but it is a big
> forestry area up there so it could be out of date.
>
>
>>
>>
>
> --
> Yes
> -
> https://www.evernote.com/shard/s2/sh/cedc6f54-8811-4912-8145-23b2d3fe42c9/2aa734310d8eb415357369cba40474c5
>
> Basically someone previously has drawn Aupori Forest as a really big
> forrest and I don't want to blow it all away. I would prefer to make a
> relation with that name and add all the LINZ objects to it before deleting
> it.
>
> The aerial imagery shows a bunch of different stuff but it is a big
> forestry area up there so it could be out of date.
>
>
>>
>>
>
> You received this message because you are subscribed to the Google Groups
> "nzopengis" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to nzopengis+...@googlegroups.com.
> To post to this group, send email to nzop...@googlegroups.com.
> Visit this group at http://groups.google.com/group/nzopengis?hl=en.
> For more options, visit https://groups.google.com/groups/opt_out.
>
>
>
JAYSONHOOGY
Jan 29, 2013, 4:12:22 PM1/29/13
to nzop...@googlegroups.com
G'day
I am currently working on redoing my work around Dunedin with the correct tagging for the approved layers.
Nice to see the progress around NZ.
Cheers
Jayson
Hamish
Feb 23, 2013, 3:27:33 AM2/23/13
to nzop...@googlegroups.com
[northland]
Glen wrote:
> I'm still trying to find the LINZ layer which 'fills in' the other parts of
> the forest but have yet to find it.
It's scrub_poly.
have fun,
Hamish
Glen wrote:
> I'm still trying to find the LINZ layer which 'fills in' the other parts of
> the forest but have yet to find it.
have fun,
Hamish
Andin
Jun 18, 2013, 4:36:03 AM6/18/13
to nzop...@googlegroups.com
Hey mate, noticing all the updates you're doing awesome stuff.
Just a heads up that for each workslice you upload, make sure you enter in a comment.
As Hamish told me when I first started,
Cheers,
mrsid
Just a heads up that for each workslice you upload, make sure you enter in a comment.
As Hamish told me when I first started,
when you do an upload, could you cut & paste in the osm changeset
number into the linz2osm workslice page? due to a quirk in the web app
you need to put at least one character in the comments box as well to
get it to accept the post.
the web app automagically sees the changeset number and turns it into
a link back to the osm database.
number into the linz2osm workslice page? due to a quirk in the web app
you need to put at least one character in the comments box as well to
get it to accept the post.
the web app automagically sees the changeset number and turns it into
a link back to the osm database.
Cheers,
mrsid
Reply all
Reply to author
Forward
0 new messages