[OSM-talk] mapnik rendering

[Steve Chilton]

I don't know whether I have missed something, or else am just lucky, but mapnik 
is rendering the things I am editing super-fast. Two new and different renders 
of an area in about 30 minutes.
 
STEVE

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] mapnik rendering

[Richard Weait]

On Thu, Sep 10, 2009 at 2:34 PM, Steve Chilton wrote:
> I don't know whether I have missed something, or else am just lucky, but 
> mapnik is rendering the things I am editing super-fast. Two new and different 
> renders of an area in about 30 minutes.

Now the renderer is sucking up to the cartographers?

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] mapnik rendering

[Shaun McDonald]

On 10 Sep 2009, at 19:40, Richard Weait wrote:

> On Thu, Sep 10, 2009 at 2:34 PM, Steve  
> Chilton wrote:
>> I don't know whether I have missed something, or else am just  
>> lucky, but mapnik is rendering the things I am editing super-fast.  
>> Two new and different renders of an area in about 30 minutes.
>
> Now the renderer is sucking up to the cartographers?
>

It's really old news, that tile.openstreetmap.org is using the  
minutely diffs.

Shaun


___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] mapnik rendering

[Lennard]

Shaun McDonald wrote:

> It's really old news, that tile.openstreetmap.org is using the  
> minutely diffs.

But Steve was out at a conference when we switched over to the new tile 
server. Which is fast enough that it doesn't drop any render requests 
any more.

-- 
Lennard

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] mapnik rendering

[Steve Chilton]

Well thanks for those responsible for the change - it is certainly appreciated.
And thanks to Lennard for putting the case for the defence!
So impressed, I am going to edit some more 
 
Cheers
STEVE

-Original Message- 
From: Lennard [mailto:l...@xs4all.nl] 
Sent: Thu 10/09/2009 22:45 
To: Shaun McDonald 
Cc: Richard Weait; talk@openstreetmap.org; Steve Chilton 
Subject: Re: [OSM-talk] mapnik rendering



Shaun McDonald wrote:

> It's really old news, that tile.openstreetmap.org is using the 
> minutely diffs.

But Steve was out at a conference when we switched over to the new tile
server. Which is fast enough that it doesn't drop any render requests
any more.

--
Lennard



___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] mapnik rendering

[Roman Neumüller]

>>> I don't know whether I have missed something, or else am just
>>> lucky, but mapnik is rendering the things I am editing super-fast.
>>> Two new and different renders of an area in about 30 minutes.
>>
>> Now the renderer is sucking up to the cartographers?
>>
> It's really old news, that tile.openstreetmap.org is using the
> minutely diffs.
> Shaun

But seemingly only on zoomlevel 12 and higher numbers, right?
Roman

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] mapnik rendering

[Ciarán Mooney]

Hi,

I noticed the blisteringly fast rendering the other day, it had
started rendering before I'd even finished my editing session. Means I
will need to hone my JOSM skills so that I'm not the slowest part of
the system!

Thanks Sys Admin Ninja's, we love you all. :)

Ciarán

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] mapnik rendering

[Peteris Krisjanis]

2009/9/11 Roman Neumüller :
 I don't know whether I have missed something, or else am just
 lucky, but mapnik is rendering the things I am editing super-fast.
 Two new and different renders of an area in about 30 minutes.
>>>
>>> Now the renderer is sucking up to the cartographers?
>>>
>> It's really old news, that tile.openstreetmap.org is using the
>> minutely diffs.
>> Shaun
>
> But seemingly only on zoomlevel 12 and higher numbers, right?

Use a browser's refresh button :) Seriously, it helps.

Again, thanks sysadmins for almost instant rendering :)

Cheers,
Peter.

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] mapnik rendering

[Hillsman, Edward]

I'd noticed the increased speed as well. Thank you to all involved in making it 
happen.

Ed

Edward L. Hillsman, Ph.D.
Senior Research Associate
Center for Urban Transportation Research
University of South Florida
4202 Fowler Ave., CUT100
Tampa, FL  33620-5375
813-974-2977 (tel)
813-974-5168 (fax)
hills...@cutr.usf.edu   
http://www.cutr.usf.edu




___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] mapnik rendering

[Mike N.]

> I'd noticed the increased speed as well. Thank you to all involved in 
> making it happen.

   I noticed this also.   That's a great help - Thanks!
 


___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] mapnik rendering

[Valent Turkovic]

On Fri, 11 Sep 2009 09:07:01 -0400, Mike N. wrote:

>I noticed this also.   That's a great help - Thanks!

Thank you all, it is blindingly fast!



-- 
pratite me na twitteru - www.twitter.com/valentt
http://kernelreloaded.blog385.com/
linux, blog, anime, spirituality, windsurf, wireless
registered as user #367004 with the Linux Counter, http://counter.li.org.
ICQ: 2125241, Skype: valent.turkovic


___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] mapnik rendering

[Jon Burgess]

On Fri, 2009-09-11 at 07:48 +0300, Roman Neumüller wrote:
> >>> I don't know whether I have missed something, or else am just
> >>> lucky, but mapnik is rendering the things I am editing super-fast.
> >>> Two new and different renders of an area in about 30 minutes.
> >>
> >> Now the renderer is sucking up to the cartographers?
> >>
> > It's really old news, that tile.openstreetmap.org is using the
> > minutely diffs.
> > Shaun


We moved the tile rendering to a new server[1] last weekend and this is
rendering tiles several times faster than the old server. Currently it
is managing to render all the tiles faster than the request rate so the
changes are showing up very quickly. 


> But seemingly only on zoomlevel 12 and higher numbers, right?
> Roman

The tile expiry script only marks tiles between zoom 13 and 18 for
automatic re-rendering. Limiting the minimum zoom is a performance
optimisation, otherwise the low zoom tiles get re-rendered far too
often. I'll try changing the minimum to 12 or 11 this weekend.

As a fallback I may setup a forced weekly render of the low zoom tiles
so they don't get left completely behind.

   Jon

1: "yevaud" http://wiki.openstreetmap.org/wiki/Servers/yevaud



___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

[OSM-talk] Mapnik rendering update

[Gervase Markham]

Anyone know why Mapnik still hasn't rendered the data I put here:
http://www.openstreetmap.org/?lat=54.6134&lon=-2.6175&zoom=12&layers=B0FT
over a week ago? Osmarender has it fine.

I've tried using the "rerender" command on informationfreeway. (Is that 
Osmarender only?) I've tried "waiting until next Wednesday" and then 
waiting a bit longer. I've tried going to the individual tile URL and 
appending "/dirty"; "/status" then saying it's awaiting rerender, and 
later it says it no longer is (so presumably it's been done) but yet, no 
change :-(

Sorry if I'm being thick. What am I missing?

Gerv


___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk

[OSM-talk] Mapnik rendering artifacts

[Karl Newman]

I'm seeing faint gray lines on the main site slippy map Mapnik base layer,
and I'm at a loss to explain their source. At first I thought they were
state borders, but I'm seeing the lines running through the middle of
states, too (here
http://www.openstreetmap.org/?lat=42.001&lon=-120.048&zoom=9&layers=B0FT it
runs vertically through the middle of Oregon). Then I thought they were tile
boundaries, but that's not consistent, either. Even weirder is on the
California-Nevada border, there are two vertical lines right next to each
other, not always parallel. (See here:
http://www.openstreetmap.org/?lat=41.62&lon=-120.053&zoom=9&layers=B0FT). I
downloaded that area in JOSM but all I saw was a single state border way. It
doesn't seem to have a source in the OSM data, so maybe it's something
caused by the conversion to pgsql or by the renderer itself (maybe a
projection issue)? I don't think it's related to the recent mod_tile
changes, because as far as I know that only affects how the tiles are stored
and served, not how they're rendered. (If I've said something stupid, please
forgive me. I *think* I understand the process...)

Any thoughts, anyone?

Karl
___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk

[OSM-talk] Mapnik rendering questions

[Maarten Deen]

I have two questions about some Mapnik renderings I encountered on the 
main OSM map.


1)
Why is this road not rendered as a tunnel?
http://www.openstreetmap.org/way/435805359

This is not a general error in the sylesheet, see this road which is 
rendered as a tunnel.

http://www.openstreetmap.org/way/24179161

2)
When rendering public transport platforms, the layer tag does not seem 
to be taken into account, making platforms that are areas render below 
roads that are running under the platform.
Example: http://www.openstreetmap.org/way/222561489 but there are a lot, 
like here https://www.openstreetmap.org/way/442106261

This probably is a bug in the stylesheet?

Regards,
Maarten

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

Re: [OSM-talk] Mapnik rendering update

[Jon Burgess]

On Fri, 2008-01-25 at 17:49 +, Thomas Wood wrote:
> I believe the mapnik tile server has been a little overloaded recently
> - queuing of new tiles was disabled for a time.
> 
> Rerender on ifw is for osmarender only (and isn't really required
> much, now that changed-tiles is being run to pick up data changes)
> 
> On Jan 25, 2008 5:36 PM, Gervase Markham <[EMAIL PROTECTED]> wrote:
> > Anyone know why Mapnik still hasn't rendered the data I put here:
> > http://www.openstreetmap.org/?lat=54.6134&lon=-2.6175&zoom=12&layers=B0FT
> > over a week ago? Osmarender has it fine.
> >
> > I've tried using the "rerender" command on informationfreeway. (Is that
> > Osmarender only?) I've tried "waiting until next Wednesday" and then
> > waiting a bit longer. I've tried going to the individual tile URL and
> > appending "/dirty"; "/status" then saying it's awaiting rerender, and
> > later it says it no longer is (so presumably it's been done) but yet, no
> > change :-(
> >
> > Sorry if I'm being thick. What am I missing?
> >

There is a major update to the way tiles are rendered occurring at the
moment, let me explain...

As Thomas mentions the rendering on the tile server has been overloaded
recently. I've been working on developing an improved version of the
software rendering framework called mod_tile for a few months and I took
the plunge this week to install and run it on tile. 

The old system used a Mysql database to store the tiles, the new one
uses the filesystem. I am currently running the list of 30M tiles from
the MySQL database through the new mod_tile rendering engine to build up
the new tiles from scratch. Right now all the tiles you are seeing are
coming from the MySQL DB. 

The tiles are still being rendered with the Mapnik renderer so the new
map will look the same as the old one. Many major changes have been made
to the other parts of the rendering setup to massively improve the
overall rendering speed.

Once this first pass of the rendering is complete the mod_tile setup
will have all the existing tiles fully rendered with the current planet
dump and current map style. At that point I will switch over the URLs on
the server to make the new tiles live.

The new system has already rendered 5M tiles in just under 24 hours.
This covers all of zooms 0 - 12 and about halfway through zoom 13. At
the current rate all the old 30M tiles will be rendered by about
Wednesday next week. 

Given how well things are progressing with the rendering I don't want to
attempt any hybrid schemes to mix the new and old tiles or transfer the
old tiles from the database to the filesystem.

The new rendering rate is about 5 times faster than the old setup and
allow us to be able to keep the tiles more up to date in future. The
weekly import of the planet dump data will still be a limitation for the
foreseeable future, but you should see updates appearing on the map much
sooner after the Wednesday each week.

Hopefully you can all live with a few more days delay in the updates
before you get a nice new set of tiles with faster updates in future.

Jon



___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk

Re: [OSM-talk] Mapnik rendering update

[Gervase Markham]

Jon Burgess wrote:
> There is a major update to the way tiles are rendered occurring at the
> moment, let me explain...


Makes perfect sense. Thanks for the update, and your hard work. I look 
forward to seeing the shiny new map next Wednesday or so :-)

Gerv


___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk

Re: [OSM-talk] Mapnik rendering update

[Jon Burgess]

On Fri, 2008-01-25 at 22:35 +, Robert (Jamie) Munro wrote:
> -BEGIN PGP SIGNED MESSAGE-
> Hash: SHA1
> 
> Jon Burgess wrote:
> | Once this first pass of the rendering is complete the mod_tile setup
> | will have all the existing tiles fully rendered with the current planet
> | dump and current map style. At that point I will switch over the URLs on
> | the server to make the new tiles live.
> |
> | The new system has already rendered 5M tiles in just under 24 hours.
> | This covers all of zooms 0 - 12 and about halfway through zoom 13. At
> | the current rate all the old 30M tiles will be rendered by about
> | Wednesday next week.
> 
> Now that you've done the slowest-to-render low zoom tiles, Why not just
> bin the old DB and backlog and let the rest of the tiles render on demand?

I might do this once all the z16 or z17 tiles are done but with only
z0-12 I fear I'll get lots of complaints about stuff suddenly
disappearing from the map. The zoom levels 15 - 17 are by far the most
numerous according to the histogram of tiles per zoom.  

ZoomNumber of tiles in database
0   1
1   4
2   16
3   64
4   256
5   1024
6   4096
7   16384
8   65539
9   262163
10  1039716
11  1507579
12  1010639
13  1495247
14  2333230
15  5110638
16  5017451
17  8192471
18  3793287
Total = 29849805

Jon



___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk

Re: [OSM-talk] Mapnik rendering update

[Robert (Jamie) Munro]

-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

Jon Burgess wrote:
| Once this first pass of the rendering is complete the mod_tile setup
| will have all the existing tiles fully rendered with the current planet
| dump and current map style. At that point I will switch over the URLs on
| the server to make the new tiles live.
|
| The new system has already rendered 5M tiles in just under 24 hours.
| This covers all of zooms 0 - 12 and about halfway through zoom 13. At
| the current rate all the old 30M tiles will be rendered by about
| Wednesday next week.

Now that you've done the slowest-to-render low zoom tiles, Why not just
bin the old DB and backlog and let the rest of the tiles render on demand?

Robert (Jamie) Munro
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.6 (Darwin)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFHmmRMz+aYVHdncI0RApeyAJ4izpTGTi2tn8u3X5vPZHfFRsSbVQCdGqC5
WcenOewOb/FpqChyFeULhBk=
=/LLL
-END PGP SIGNATURE-

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk

Re: [OSM-talk] Mapnik rendering update

[Thomas Wood]

I believe the mapnik tile server has been a little overloaded recently
- queuing of new tiles was disabled for a time.

Rerender on ifw is for osmarender only (and isn't really required
much, now that changed-tiles is being run to pick up data changes)

On Jan 25, 2008 5:36 PM, Gervase Markham <[EMAIL PROTECTED]> wrote:
> Anyone know why Mapnik still hasn't rendered the data I put here:
> http://www.openstreetmap.org/?lat=54.6134&lon=-2.6175&zoom=12&layers=B0FT
> over a week ago? Osmarender has it fine.
>
> I've tried using the "rerender" command on informationfreeway. (Is that
> Osmarender only?) I've tried "waiting until next Wednesday" and then
> waiting a bit longer. I've tried going to the individual tile URL and
> appending "/dirty"; "/status" then saying it's awaiting rerender, and
> later it says it no longer is (so presumably it's been done) but yet, no
> change :-(
>
> Sorry if I'm being thick. What am I missing?
>
> Gerv
>
>
> ___
> talk mailing list
> talk@openstreetmap.org
> http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
>



-- 
Regards,
Thomas Wood
(Edgemaster)

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk

Re: [OSM-talk] Mapnik rendering update

[Robert (Jamie) Munro]

-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

Jon Burgess wrote:
| On Fri, 2008-01-25 at 22:35 +, Robert (Jamie) Munro wrote:
|> -BEGIN PGP SIGNED MESSAGE-
|> Hash: SHA1
|>
|> Jon Burgess wrote:
|> | Once this first pass of the rendering is complete the mod_tile setup
|> | will have all the existing tiles fully rendered with the current planet
|> | dump and current map style. At that point I will switch over the
URLs on
|> | the server to make the new tiles live.
|> |
|> | The new system has already rendered 5M tiles in just under 24 hours.
|> | This covers all of zooms 0 - 12 and about halfway through zoom 13. At
|> | the current rate all the old 30M tiles will be rendered by about
|> | Wednesday next week.
|>
|> Now that you've done the slowest-to-render low zoom tiles, Why not just
|> bin the old DB and backlog and let the rest of the tiles render on
demand?
|
| I might do this once all the z16 or z17 tiles are done but with only
| z0-12 I fear I'll get lots of complaints about stuff suddenly
| disappearing from the map. The zoom levels 15 - 17 are by far the most
| numerous according to the histogram of tiles per zoom.

Are you able to say how many times each high zoom tile has been viewed?
I suspect that it most cases it will be 1 if they are rendered on
demand, and 0 if they are rendered from a backlog. There will be tiles
that have been viewed a lot - tiles in important cities or places with
lots of OSM activity, for example, but I suspect a lot of the 8 million
z17 tiles will never be looked again, at least not until their content
has been changed.

If we do get a backlog, then it needs to give priority to tiles that
have been viewed more than once, so that on a second viewing, a tile
will be rendered live, even if the system is a bit busy (but not a lot
busy, as it were).

If you do get complaints, you can always switch back.

Out of interest, do you have rough stats on how long tiles at each zoom
level take to render on average?

Robert (Jamie) Munro

-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.6 (Darwin)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFHmzviz+aYVHdncI0RAtMhAJ9uY+Nvw8NgRllh1VErN2wdAV+jiACgoulq
CxZ2MbuLSGPJFK1qR+SfeJU=
=knlt
-END PGP SIGNATURE-

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk

Re: [OSM-talk] Mapnik rendering update

[Joerg Ostertag (OSM Munich/Germany)]

On Freitag 25 Januar 2008, Gervase Markham wrote:
> Jon Burgess wrote:
> > There is a major update to the way tiles are rendered occurring at the
> > moment, let me explain...
>
> 
>
> Makes perfect sense. Thanks for the update, and your hard work. I look
> forward to seeing the shiny new map next Wednesday or so :-)

A great thanks from Germany too !!

-- 

Joerg

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk

Re: [OSM-talk] Mapnik rendering update

[Jon Burgess]

On Fri, 2008-01-25 at 23:28 +, Jon Burgess wrote:
> On Fri, 2008-01-25 at 22:35 +, Robert (Jamie) Munro wrote:
> > -BEGIN PGP SIGNED MESSAGE-
> > Hash: SHA1
> > 
> > Jon Burgess wrote:
> > | Once this first pass of the rendering is complete the mod_tile setup
> > | will have all the existing tiles fully rendered with the current planet
> > | dump and current map style. At that point I will switch over the URLs on
> > | the server to make the new tiles live.
> > |
> > | The new system has already rendered 5M tiles in just under 24 hours.
> > | This covers all of zooms 0 - 12 and about halfway through zoom 13. At
> > | the current rate all the old 30M tiles will be rendered by about
> > | Wednesday next week.
> > 

I've made the new tiles from mod_tile live on the main OSM tile server
(on the default Mapnik layer). The initial impression I get is that the
tiles are being served to more responsively and the on-access rendering
seems to be working well too. Hopefully this will all continue smoothly.


There are a couple of issues that I've noticed so far. I don't think any
of these are serious enough to force us to go back to the old rendering
setup. They should all be resolved in due course.

* The new colour reduction algorithm causes some odd colour changes in a
few tiles. 99% are fine but a few have a background colour or road fill
which is slightly wrong.

* Some of the z16/z17 tiles from the old system have not been rendered
yet. They should get rendered and appear shortly after you access them.
The new system uses a 8x8 meta tile rendering approach so overall we
actually have more tiles than the old system (40M vs 30M previously).

* The 'Expires' HTTP header has been broken causing tiles not to be
cached as well as they should be in your browser (or other proxy/cache).

* The munin stats for Apache volume have broken, you can see the real
throughput is still OK by looking at the eth0 stats.

Jon



___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk

Re: [OSM-talk] Mapnik rendering update

[Artem Pavlenko]

On 29 Jan 2008, at 23:40, Jon Burgess wrote:

>
> On Fri, 2008-01-25 at 23:28 +, Jon Burgess wrote:
>> On Fri, 2008-01-25 at 22:35 +, Robert (Jamie) Munro wrote:
>>> -BEGIN PGP SIGNED MESSAGE-
>>> Hash: SHA1
>>>
>>> Jon Burgess wrote:
>>> | Once this first pass of the rendering is complete the mod_tile  
>>> setup
>>> | will have all the existing tiles fully rendered with the  
>>> current planet
>>> | dump and current map style. At that point I will switch over  
>>> the URLs on
>>> | the server to make the new tiles live.
>>> |
>>> | The new system has already rendered 5M tiles in just under 24  
>>> hours.
>>> | This covers all of zooms 0 - 12 and about halfway through zoom  
>>> 13. At
>>> | the current rate all the old 30M tiles will be rendered by about
>>> | Wednesday next week.
>>>
>
> I've made the new tiles from mod_tile live on the main OSM tile server
> (on the default Mapnik layer). The initial impression I get is that  
> the
> tiles are being served to more responsively and the on-access  
> rendering
> seems to be working well too. Hopefully this will all continue  
> smoothly.
>
>

Fantastic.
>
> * The new colour reduction algorithm causes some odd colour changes  
> in a
> few tiles. 99% are fine but a few have a background colour or road  
> fill
> which is slightly wrong.


I had a look and there are quite a few of wrong tile. I committed  
changes that seems to fix this in r625.

Great work!
Artem


___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk

Re: [OSM-talk] Mapnik rendering artifacts

[Dave Stubbs]

On Jan 30, 2008 12:41 PM, Karl Newman <[EMAIL PROTECTED]> wrote:

> I'm seeing faint gray lines on the main site slippy map Mapnik base layer,
> and I'm at a loss to explain their source. At first I thought they were
> state borders, but I'm seeing the lines running through the middle of
> states, too (here
> http://www.openstreetmap.org/?lat=42.001&lon=-120.048&zoom=9&layers=B0FTit 
> runs vertically through the middle of Oregon). Then I thought they were
> tile boundaries, but that's not consistent, either. Even weirder is on the
> California-Nevada border, there are two vertical lines right next to each
> other, not always parallel. (See here:
> http://www.openstreetmap.org/?lat=41.62&lon=-120.053&zoom=9&layers=B0FT).
> I downloaded that area in JOSM but all I saw was a single state border way.
> It doesn't seem to have a source in the OSM data, so maybe it's something
> caused by the conversion to pgsql or by the renderer itself (maybe a
> projection issue)? I don't think it's related to the recent mod_tile
> changes, because as far as I know that only affects how the tiles are stored
> and served, not how they're rendered. (If I've said something stupid, please
> forgive me. I *think* I understand the process...)
>
> Any thoughts, anyone?
>
>
These have always been there.

It's an artefact of the way the coastlines are rendered.
Country/state/region polygon shapes are split into a large grid to improve
performance. Each shape is then rendered as land. Unfortunately the edges
meet exactly, but don't overlap, and the boundary is never pixel perfect, so
what I think is an antialiasing effect causes the blue background to seep
through. So you'll see these lines at all the borders known in the shape
file, which include US states, European countries and others, as well as the
grid.
Your double line example is because the shape for california and the tiling
grid almost coincide, but not quite.

You can see a similar effect on the coastline checker, but that has more
tiles in it's grid: http://tile.openstreetmap.nl/coastlines.html
___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk

Re: [OSM-talk] Mapnik rendering artifacts

[Artem Pavlenko]

On 30 Jan 2008, at 12:41, Karl Newman wrote:

I'm seeing faint gray lines on the main site slippy map Mapnik base  
layer, and I'm at a loss to explain their source. At first I  
thought they were state borders, but I'm seeing the lines running  
through the middle of states, too (here http:// 
www.openstreetmap.org/?lat=42.001&lon=-120.048&zoom=9&layers=B0FT  
it runs vertically through the middle of Oregon). Then I thought  
they were tile boundaries, but that's not consistent, either. Even  
weirder is on the California-Nevada border, there are two vertical  
lines right next to each other, not always parallel. (See here:  
http://www.openstreetmap.org/? 
lat=41.62&lon=-120.053&zoom=9&layers=B0FT). I downloaded that area  
in JOSM but all I saw was a single state border way. It doesn't  
seem to have a source in the OSM data, so maybe it's something  
caused by the conversion to pgsql or by the renderer itself (maybe  
a projection issue)? I don't think it's related to the recent  
mod_tile changes, because as far as I know that only affects how  
the tiles are stored and served, not how they're rendered. (If I've  
said something stupid, please forgive me. I *think* I understand  
the process...)


Any thoughts, anyone?


railtracks ?

Artem




Karl
___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk


___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk

Re: [OSM-talk] Mapnik rendering artifacts

[Adrian Frith]

On Wed, 2008-01-30 at 04:41 -0800, Karl Newman wrote:
> I'm seeing faint gray lines on the main site slippy map Mapnik base
> layer, and I'm at a loss to explain their source. At first I thought
> they were state borders, but I'm seeing the lines running through the
> middle of states, too (here
> http://www.openstreetmap.org/?lat=42.001&lon=-120.048&zoom=9&layers=B0FT
>  it runs vertically through the middle of Oregon). Then I thought they were 
> tile boundaries, but that's not consistent, either. Even weirder is on the 
> California-Nevada border, there are two vertical lines right next to each 
> other, not always parallel. (See here: 
> http://www.openstreetmap.org/?lat=41.62&lon=-120.053&zoom=9&layers=B0FT).
>  I downloaded that area in JOSM but all I saw was a single state border way. 
> It doesn't seem to have a source in the OSM data, so maybe it's something 
> caused by the conversion to pgsql or by the renderer itself (maybe a 
> projection issue)? I don't think it's related to the recent mod_tile changes, 
> because as far as I know that only affects how the tiles are stored and 
> served, not how they're rendered. (If I've said something stupid, please 
> forgive me. I *think* I understand the process...)
> 
> Any thoughts, anyone?

County boundaries?


___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk

Re: [OSM-talk] Mapnik rendering artifacts

[Artem Pavlenko]

On 30 Jan 2008, at 12:41, Karl Newman wrote:

I'm seeing faint gray lines on the main site slippy map Mapnik base  
layer, and I'm at a loss to explain their source. At first I  
thought they were state borders, but I'm seeing the lines running  
through the middle of states, too (here http:// 
www.openstreetmap.org/?lat=42.001&lon=-120.048&zoom=9&layers=B0FT  
it runs vertically through the middle of Oregon). Then I thought  
they were tile boundaries, but that's not consistent, either. Even  
weirder is on the California-Nevada border, there are two vertical  
lines right next to each other, not always parallel. (See here:  
http://www.openstreetmap.org/? 
lat=41.62&lon=-120.053&zoom=9&layers=B0FT). I downloaded that area  
in JOSM but all I saw was a single state border way. It doesn't  
seem to have a source in the OSM data, so maybe it's something  
caused by the conversion to pgsql or by the renderer itself (maybe  
a projection issue)? I don't think it's related to the recent  
mod_tile changes, because as far as I know that only affects how  
the tiles are stored and served, not how they're rendered. (If I've  
said something stupid, please forgive me. I *think* I understand  
the process...)


Any thoughts, anyone?

Karl


Ok, I got it. There are artifacts from rendering adjacent coast-line  
polygons (which are tiled -  100x100km squares inland)
This will be fixed when we upgrade to new coastline shape files - new  
tiles overlap by 150m and don't have these faint lines


Artem



___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk


___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk

Re: [OSM-talk] Mapnik rendering questions

[Martin Koppenhoefer]

2017-07-06 11:43 GMT+02:00 Maarten Deen :

> I have two questions about some Mapnik renderings I encountered on the
> main OSM map.
>
> 1)
> Why is this road not rendered as a tunnel?
> http://www.openstreetmap.org/way/435805359




it is rendered as a tunnel, but you can't see it, because there's another
road drawn above it:
http://www.openstreetmap.org/way/437039882

Cheers,
Martin
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Mapnik rendering questions

[Martin Koppenhoefer]

2017-07-06 11:43 GMT+02:00 Maarten Deen :

> 2)
> When rendering public transport platforms, the layer tag does not seem to
> be taken into account, making platforms that are areas render below roads
> that are running under the platform.
> Example: http://www.openstreetmap.org/way/222561489 but there are a lot,
> like here https://www.openstreetmap.org/way/442106261
> This probably is a bug in the stylesheet?
>



not sure if it is a bug, but as a workaround for the renderer, you can add
covered=yes tags to the part of the road that is under the platform.

Cheers,
Martin
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Mapnik rendering questions

[Maarten Deen]

On 2017-07-06 12:11, Martin Koppenhoefer wrote:

2017-07-06 11:43 GMT+02:00 Maarten Deen :


I have two questions about some Mapnik renderings I encountered on
the main OSM map.

1)
Why is this road not rendered as a tunnel?
http://www.openstreetmap.org/way/435805359 [1]


it is rendered as a tunnel, but you can't see it, because there's
another road drawn above it:
http://www.openstreetmap.org/way/437039882


Ah, errors in the data like that are not helpful. I'll have a look at 
the data.


Regards,
Maarten

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

Re: [OSM-talk] Mapnik rendering questions

[Daniel Koć]

W dniu 06.07.2017 o 11:43, Maarten Deen pisze:


2)
When rendering public transport platforms, the layer tag does not seem 
to be taken into account, making platforms that are areas render below 
roads that are running under the platform.
Example: http://www.openstreetmap.org/way/222561489 but there are a 
lot, like here https://www.openstreetmap.org/way/442106261

This probably is a bug in the stylesheet?


The common problem with layers is that people expect them to act 
globally, while in our implementation they are "local" - they respect 
rendering priority just for a set of objects, not all of them.


--
"Like a halo in reverse" [M. Gore]


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

Re: [OSM-talk] Mapnik rendering questions

[Maarten Deen]

On 2017-07-06 12:11, Martin Koppenhoefer wrote:

2017-07-06 11:43 GMT+02:00 Maarten Deen :


I have two questions about some Mapnik renderings I encountered on
the main OSM map.

1)
Why is this road not rendered as a tunnel?
http://www.openstreetmap.org/way/435805359 [1]


it is rendered as a tunnel, but you can't see it, because there's
another road drawn above it:
http://www.openstreetmap.org/way/437039882


I have deleted the road above it two days ago and the tile is still 
unchanged. Is there something with the renderer queue? You used to be 
able to invalidate a tile, but that does not work anymore.


Maarten

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

Re: [OSM-talk] Mapnik rendering questions

[Martin Koppenhoefer]

sent from a phone

> On 8. Jul 2017, at 08:39, Maarten Deen  wrote:
> 
> You used to be able to invalidate a tile, but that does not work anymore


AFAIK it still works, although it is rather complicated now to get the tile 
URL, and there's reluctance from the admin to expose them in a convenient way 
(right click menu)

I'm doing it in firefox by hitting ctrl+i and browsing through the media urls 
until I find the right tile...

cheers,
Martin 
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Mapnik rendering questions

[Maarten Deen]

On 2017-07-08 10:44, Martin Koppenhoefer wrote:

sent from a phone


On 8. Jul 2017, at 08:39, Maarten Deen  wrote:

You used to be able to invalidate a tile, but that does not work 
anymore



AFAIK it still works, although it is rather complicated now to get the
tile URL, and there's reluctance from the admin to expose them in a
convenient way (right click menu)

I'm doing it in firefox by hitting ctrl+i and browsing through the
media urls until I find the right tile...


CTRL-i gives me the bookmarks. I'm not sure that's what you mean. Where 
can I find the url for the tile there?


Regards,
Maarten

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

Re: [OSM-talk] Mapnik rendering questions

[Maarten Deen]

On 2017-07-08 10:44, Martin Koppenhoefer wrote:

sent from a phone


On 8. Jul 2017, at 08:39, Maarten Deen  wrote:

You used to be able to invalidate a tile, but that does not work 
anymore



AFAIK it still works, although it is rather complicated now to get the
tile URL, and there's reluctance from the admin to expose them in a
convenient way (right click menu)

I'm doing it in firefox by hitting ctrl+i and browsing through the
media urls until I find the right tile...


It took me a while to get to the correct tile on zoom 15 [1] but it was 
clean and rendered at 30-6. I marked it dirty and it got rendered. In 
the process I noticed that z13 appeard correct too.

z16 is still incorrect [2]
"Tile is clean. Last rendered at Wed Jul 05 20:47:35 2017. Last accessed 
at Sat Jul 08 18:55:11 2017. Stored in 
file:///srv/tile.openstreetmap.org/tiles/default/16/0/149/15/249/0.meta"
I did my edit on the 6th. Why does the renderer not see that I deleted 
an item and mark the tile dirty?


I have to say: it is a nuisance that you can not get this information 
easily. Please someone who is in the map development and reads this: add 
an item to the right-click menu where you can see tile status. There is 
a complaint about it on the dutch forum as well. [3]


[1] http://tile.openstreetmap.org/15/18552/12232.png
[2] http://tile.openstreetmap.org/16/37105/24465.png/status
[3] https://forum.openstreetmap.org/viewtopic.php?pid=653871#p653871

Maarten

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

Re: [OSM-talk] Mapnik rendering questions

[Martin Koppenhoefer]

sent from a phone

> On 9. Jul 2017, at 08:36, Maarten Deen  wrote:
> 
> I have to say: it is a nuisance that you can not get this information easily. 
> Please someone who is in the map development and reads this: add an item to 
> the right-click menu where you can see tile status.


https://github.com/openstreetmap/openstreetmap-website/issues/1446

cheers,
Martin ___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Mapnik rendering questions

[mmd]

Am 09.07.2017 um 12:29 schrieb Martin Koppenhoefer:

> sent from a phone

?

> 
> On 9. Jul 2017, at 08:36, Maarten Deen  > wrote:
> 
>> I have to say: it is a nuisance that you can not get this information
>> easily. Please someone who is in the map development and reads this:
>> add an item to the right-click menu where you can see tile status.
> 
> 
> https://github.com/openstreetmap/openstreetmap-website/issues/1446
>

There even used to be a demo by TomH where you could right click on a
tile and get its current status as a nice summary on the left pane.

It was previously deployed on the on dev server for testing
https://tomh.apis.dev.openstreetmap.org/ but doesn't seem to be around
anymore unfortunately.

Commits for this feature are available at:
https://github.com/tomhughes/openstreetmap-website/commits/tile-details

-- 






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

Re: [OSM-talk] Mapnik rendering questions

[Maarten Deen]

On 2017-07-09 08:36, Maarten Deen wrote:

On 2017-07-08 10:44, Martin Koppenhoefer wrote:

sent from a phone


On 8. Jul 2017, at 08:39, Maarten Deen  wrote:

You used to be able to invalidate a tile, but that does not work 
anymore



AFAIK it still works, although it is rather complicated now to get the
tile URL, and there's reluctance from the admin to expose them in a
convenient way (right click menu)

I'm doing it in firefox by hitting ctrl+i and browsing through the
media urls until I find the right tile...


It took me a while to get to the correct tile on zoom 15 [1] but it
was clean and rendered at 30-6. I marked it dirty and it got rendered.
In the process I noticed that z13 appeard correct too.
z16 is still incorrect [2]
"Tile is clean. Last rendered at Wed Jul 05 20:47:35 2017. Last
accessed at Sat Jul 08 18:55:11 2017. Stored in
file:///srv/tile.openstreetmap.org/tiles/default/16/0/149/15/249/0.meta"
I did my edit on the 6th. Why does the renderer not see that I deleted
an item and mark the tile dirty?


The area still is not rendered again. 
http://www.openstreetmap.org/#map=19/41.43684/23.82636 still shows the 
old situation. I find it really odd that this change is not picked up by 
the renderer.


Maarten

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

[OSM-talk] Mapnik: rendering forest or wood

[Karl Eichwalder]

I think the mapnik rendering of forests could be improved.  ATM,
landuse=forest is not distinguishable from recreation_ground.  Even
if forest are often used as places for recreation in Germany,
rendering both areas the same way is not optimal.

For outside activities you want where wood or forest is located.

Thus I'd like to propose to render both landuse=forest and
natural=wood the same way in a darkish green.  That's also how
osmarender deals with these areas.




___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk

[OSM-talk] mapnik rendering or tagging issue?

[Stefan Monnier]

If you look at
http://openstreetmap.org/?lat=45.52&lon=-73.59&zoom=8&layers=B000FTFT
you should see an urban island labelled "Laval" and "Longueil".  If you
zoom out, only "Laval" is left, while if you zoom in you'll discover
that this is really the island of Montreal.

Is this a problem in mapnik or is it a problem with the OSM data?
If the latter, how can I fix it?


Stefan


___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

[OSM-talk] Mapnik rendering of Marinas changed

[Dave F.]

Hi

Is there a reason why Mapnik's rendering of leisure=marina has change 
from a blue area to a dashed border line?

http://osm.org/go/eukhlr0la--

Cheers
Dave F.

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Mapnik: rendering forest or wood

[Igor Brejc]

Karl Eichwalder wrote:
> I think the mapnik rendering of forests could be improved.  ATM,
> landuse=forest is not distinguishable from recreation_ground.  Even
> if forest are often used as places for recreation in Germany,
> rendering both areas the same way is not optimal.
>
> For outside activities you want where wood or forest is located.
>
> Thus I'd like to propose to render both landuse=forest and
> natural=wood the same way in a darkish green.  That's also how
> osmarender deals with these areas.
>
>
>   
I agree. Anyway, I don't really see the distinction between the two, at 
least not in the country I'm living in (Slovenia). And this unclarity 
has resulted in some OSM Slovenian users marking certain areas as 
natural=wood, while some other mark the similar areas as landuse=forest 
(example: 
http://www.openstreetmap.org/?lat=46.411&lon=15.481&zoom=11&layers=B0FT).

Regards,
Igor

-- 
http://igorbrejc.net


___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk

Re: [OSM-talk] Mapnik: rendering forest or wood

[Daniel Schmidt]

> I think the mapnik rendering of forests could be improved.
>

> Thus I'd like to propose to render both landuse=forest and
> natural=wood the same way in a darkish green.  That's also how
> osmarender deals with these areas.

Why not use bitmap tiles for forest/wood rendering?
I've just painted a small tile which I'm hereby releasing into public  
domain. (http://osm.planetschmidt.de/pub/forest_tile.png)
Here's a mockup I made in Photoshop: http://osm.planetschmidt.de/pub/forest.png

Greets,

Daniel Schmidt

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk

Re: [OSM-talk] Mapnik: rendering forest or wood

[Guilhem Bonnefille]

Beautiful! Good work!
I vote for. :-)

2008/2/23, Daniel Schmidt <[EMAIL PROTECTED]>:
>
>
> > I think the mapnik rendering of forests could be improved.
> >
>
> > Thus I'd like to propose to render both landuse=forest and
> > natural=wood the same way in a darkish green.  That's also how
> > osmarender deals with these areas.
>
> Why not use bitmap tiles for forest/wood rendering?
> I've just painted a small tile which I'm hereby releasing into public
> domain. (http://osm.planetschmidt.de/pub/forest_tile.png)
> Here's a mockup I made in Photoshop: 
> http://osm.planetschmidt.de/pub/forest.png
>
> Greets,
>
> Daniel Schmidt

-- 
Guilhem BONNEFILLE
-=- #UIN: 15146515 JID: [EMAIL PROTECTED] MSN: [EMAIL PROTECTED]
-=- mailto:[EMAIL PROTECTED]
-=- http://nathguil.free.fr/

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk

Re: [OSM-talk] Mapnik: rendering forest or wood

[Bone Killian]

I disagree.  I think there is value in the ability to visually
distinguish managed forests from natural woods. 

Consider this area:
http://openstreetmap.org/?lat=40.9567&lon=-77.1919&zoom=13&layers=B0FT

Here, the dark green (natural=wood) is a forest preserve[1] in the
middle of an actively managed forest (landuse=forest).  Logging is
permitted in the light green area, but not in the dark green area.

It seems to me that the best course of action would be to use another,
unique shade of green for recreation_ground.

Bone

[1] http://www.dcnr.state.pa.us/FORESTRY/oldgrowth/thehook.aspx

On Sat, 2008-02-23 at 06:51 +0100, Karl Eichwalder wrote:
> I think the mapnik rendering of forests could be improved.  ATM,
> landuse=forest is not distinguishable from recreation_ground.  Even
> if forest are often used as places for recreation in Germany,
> rendering both areas the same way is not optimal.
> 
> For outside activities you want where wood or forest is located.
> 
> Thus I'd like to propose to render both landuse=forest and
> natural=wood the same way in a darkish green.  That's also how
> osmarender deals with these areas.
> 
> 
> 
> 
> ___
> talk mailing list
> talk@openstreetmap.org
> http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk


___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk

Re: [OSM-talk] mapnik rendering or tagging issue?

[Ed Loach]

Stefan asked:

> Is this a problem in mapnik or is it a problem with the OSM
> data?
> If the latter, how can I fix it?

It looks like the data is OK. All three are cities, and each is
tagged as such. I suspect that the lowzoom Mapnik tiles just happen
to decide that rendering the outer two cities names would overlap
the name of the middle one, which is Montreal, so drops it. 

If we condoned tagging for rendering, you could perhaps move the
node that is used to label Montreal south slightly, so the captions
*might* not overlap, but it's probably best not to worry about it.

I'm wondering whether the Greater Montreal Area [1] is defined as a
region, and if so whether that label would take priority at certain
zoom levels?

Ed
[1] http://en.wikipedia.org/wiki/Greater_Montreal_Area




___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] mapnik rendering or tagging issue?

[Andy Allan]

On Mon, Oct 6, 2008 at 5:28 PM, Ed Loach <[EMAIL PROTECTED]> wrote:
> Stefan asked:
>
>> Is this a problem in mapnik or is it a problem with the OSM
>> data?
>> If the latter, how can I fix it?
>
> It looks like the data is OK. All three are cities, and each is
> tagged as such. I suspect that the lowzoom Mapnik tiles just happen
> to decide that rendering the outer two cities names would overlap
> the name of the middle one, which is Montreal, so drops it.

Yep, it's a common problem. If someone can extend the mapnik
stylesheet and osm2pgsql to do something more clever with place names,
that would be cool.

Zoom levels 4+5 come from VMAP0, hence the ascii-ness. Beyond that,
increasing zoom levels increases the number of types of place names
rendered (cities, towns etc) but there's no priority in placename
rendering. See 
http://trac.openstreetmap.org/browser/applications/rendering/mapnik/osm-template-fontset.xml#L5329
- it would be nice if there was a z-order for points (and an ORDER BY
clause, of course) that reflected that we'd rather draw city names
first, then towns if there were space etc. Even better, we'd probably
want to order cities by population so that a big city next to a
smaller one still gets its name rendered. Such z-order calculations
are probably best done in osm2pgsql, which is where the z-ordering of
road occurs (see
http://trac.openstreetmap.org/browser/applications/utils/export/osm2pgsql/output-pgsql.c#L88
).

Students will achieve extra credit for moving z-ordering into a
run-time configuration file ala default.style :-)

Cheers,
Andy

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

[OSM-talk] Mapnik rendering export has only coastline

[Ben Ward]

Hi, I thought I'd do some editing in Inkscape on an SVG/PDF version of a
map, and discovered Openstreetmap.org's export tab.

Unfortunately, though I get a result from the Osmarender option, the Mapnik
output doesn't work.

http://badgertrack.com/images/swindon-oxford.png
(roughly
http://openstreetmap.org/?lat=51.619&lon=-1.496&zoom=10&layers=B000FTFT)
...unless you zoom out...
http://badgertrack.com/images/mapnik-export-coast.png

This looks like a bug/problem with the Openstreetmap Mapnik Export
rendering.  Can anyone confirm, or fix?

Ben

-- 
b...@crouchingbadger.com | http://crouchingbadger.com
51.717817,-1.225855
___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

[OSM-talk] Mapnik rendering issue with *_link roads

[Maarten Deen]

I finally found out what the issue is with this [1] situation. My issue 
with it is that IMHO it is not good to see the unclassified road 
rendered on top of the primary_link road. I made a test here [2] where 
the left primary is a primary_link and the right primary is a primary 
proper. Apparently mapnik's renderering rules state that primary_link 
roads should be rendered below unclassified roads. I haven't inspected 
the rendering rules in detail (I'm also not familiar with them) but I 
have seen that tertiary_link roads are also rendered below unclassified 
roads. I expect that any *_link road is rendered below any other (or at 
least motorway - unclassified) road.


Is this behaviour of mapnik wanted? As I said: IMHO it is not pleasing 
to the eye to see the unclassified road rendered on top of the 
primary_link road. In order of priority, a *_link road is just below its 
* counterpart but above the next lower road (so primary -> primary_link 
-> secondary).


[1] http://www.openstreetmap.org/?lat=51.352555&lon=6.014996&zoom=18
[2] http://www.openstreetmap.org/?lat=51.352419&lon=6.010627&zoom=18

Regards,
Maarten

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Mapnik rendering of Marinas changed

[Patrick Kilian]

Hi all,

> Is there a reason why Mapnik's rendering of leisure=marina has change 
> from a blue area to a dashed border line?
IIRC there was a discussion in #osm which basically went like this:

"Does leisure=marina cover the land portion or the water portion of the
marina"

"hm, well don't know. wiki says it's the land part"


"well anyway. rendering it in the color of water is bad. if it is water
you don't see that it is a marina and if it is land it should be blue"

"ah well makes sense. lets change it to a border line. that isn't
confusing and opens more options when we want to start rendering tidal
waters."

"ok lets make it a blue dashed line and see who complains".

The discussion was about three week old and quoted from memory, so I
might be slightly wrong.


HTH,
Patrick "Petschge" Kilian


___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Mapnik rendering of Marinas changed

[Dave F.]

Patrick Kilian wrote:
> Hi all,
>
>   
> IIRC there was a discussion in #osm which basically went like this:

Hi

Which forum do you mean by #osm?

Ta
Dave F.

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Mapnik rendering of Marinas changed

[Jean-Marc Liotier]

Dave F. wrote:
> Patrick Kilian wrote:
>
> Which forum do you mean by #osm?

Thats certainly the IRC channel on OFTC :
irc://irc.oftc.net #osm


___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Mapnik rendering of Marinas changed

[Patrick Kilian]

Hi,

>> IIRC there was a discussion in #osm which basically went like this:
> Which forum do you mean by #osm?
The IRC channel #osm in the oftc network. (The one mentioned in
http://wiki.openstreetmap.org/wiki/IRC#IRC)

HTH,
Patrick "Petschge" Kilian

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Mapnik rendering of Marinas changed

[Jean-Marc Liotier]

Patrick Kilian wrote:
> 
>> Is there a reason why Mapnik's rendering of leisure=marina has change 
>> from a blue area to a dashed border line?
>
> IIRC there was a discussion in #osm which basically went like this:
> [..]
> "ok lets make it a blue dashed line and see who complains".

Count me as a complainer... I was wondering about that blue dashed line
too - it does not fit well with the surrounding estuary and open sea.

I'm in the process of mapping a marina and I spontaneously only tagged
as a leisure=marina the water portion that is occupied by the harbour.


___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Mapnik rendering of Marinas changed

[David James]

> Patrick Kilian wrote:
>> Hi all,
>>
>>
>> IIRC there was a discussion in #osm which basically went like this:
>
> Hi
>
> Which forum do you mean by #osm?

IRC channel I'd guess.
-- 
David James



___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Mapnik rendering of Marinas changed

[Patrick Kilian]

Hi all,

>>> Is there a reason why Mapnik's rendering of leisure=marina has change 
>>> from a blue area to a dashed border line?
>> IIRC there was a discussion in #osm which basically went like this:
>> [..]
>> "ok lets make it a blue dashed line and see who complains".
> 
> Count me as a complainer... I was wondering about that blue dashed line
> too - it does not fit well with the surrounding estuary and open sea.
> 
> I'm in the process of mapping a marina and I spontaneously only tagged
> as a leisure=marina the water portion that is occupied by the harbour.
Open a trac ticket for mapnik then. Neither Steve nor the others in that
discussion had a strong opinion. Note that I didn't take part in that
discussion but merely witnessed it.



Patrick "Petschge" Kilian

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Mapnik rendering of Marinas changed

[Richard Weait]

On Sat, Apr 17, 2010 at 8:38 AM, Patrick Kilian  wrote:
> Hi all,
>
 Is there a reason why Mapnik's rendering of leisure=marina has change
 from a blue area to a dashed border line?
>>> IIRC there was a discussion in #osm which basically went like this:
>>> [..]
>>> "ok lets make it a blue dashed line and see who complains".
>>
>> Count me as a complainer... I was wondering about that blue dashed line
>> too - it does not fit well with the surrounding estuary and open sea.
>>
>> I'm in the process of mapping a marina and I spontaneously only tagged
>> as a leisure=marina the water portion that is occupied by the harbour.
> Open a trac ticket for mapnik then. Neither Steve nor the others in that
> discussion had a strong opinion. Note that I didn't take part in that
> discussion but merely witnessed it.

I remember that discussion, or one similar.  The question at hand was
how to distinguish the marina land area and water area from the
surrounding land and water.  The old-style marina-as-water made marina
land area appear to be water.  An outline, like administrative
boundaries, does not have that flaw.

A similar argument was made for national parks.  Rather than painting
them all green, show the boundary, let the distinct areas within show
through.

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Mapnik rendering of Marinas changed

[Dave F.]

David James wrote:
>> Patrick Kilian wrote:
>> 
>>> Hi all,
>>>
>>>
>>> IIRC there was a discussion in #osm which basically went like this:
>>>   
>> Hi
>>
>> Which forum do you mean by #osm?
>> 
>
> IRC channel I'd guess.
>   
Sorry, brain fade!

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Mapnik rendering of Marinas changed

[Dave F.]

Richard Weait wrote:
> On Sat, Apr 17, 2010 at 8:38 AM, Patrick Kilian  wrote:
>   
>> Hi all,
>>
>> 
> Is there a reason why Mapnik's rendering of leisure=marina has change
> from a blue area to a dashed border line?
>   
 IIRC there was a discussion in #osm which basically went like this:
 [..]
 "ok lets make it a blue dashed line and see who complains".
 
>>> Count me as a complainer... I was wondering about that blue dashed line
>>> too - it does not fit well with the surrounding estuary and open sea.
>>>
>>> I'm in the process of mapping a marina and I spontaneously only tagged
>>> as a leisure=marina the water portion that is occupied by the harbour.
>>>   
>> Open a trac ticket for mapnik then. Neither Steve nor the others in that
>> discussion had a strong opinion. Note that I didn't take part in that
>> discussion but merely witnessed it.
>> 
>
> I remember that discussion, or one similar.  The question at hand was
> how to distinguish the marina land area and water area from the
> surrounding land and water.  The old-style marina-as-water made marina
> land area appear to be water.  An outline, like administrative
> boundaries, does not have that flaw.
>   

Old style?
When I checked up on how to tag them, the first time the examples I 
looked at all signified the water area & assumed this was used in the 
majority?
Is this not the case?

If there wasn't a strong opinion I'm not sure why they changed it.

Did they offer an alternative for the water area? Sea/River etc. doesn't 
seem to cover it.

Wiki:
"This is a mooring place for boats"

To me this implies the water. You can't moor a boat on land.

> A similar argument was made for national parks.  Rather than painting
> them all green, show the boundary, let the distinct areas within show
> through.
This will apply to almost any item that's an enclosed area.
I thought the layer=* tag was implemented to allow renderers to 
differentiate.

-

Going slightly off-topic, this appears to highlight a problem with IRC.

A unilateral decision has been made only by people in the same time zone 
or with similar sleep patterns.
Am I correct in thinking there's no record of Instant chat? Was this 
discussion transferred to a forum?

As this appears to be a disagreement about the tagging & not just a 
specific render, shouldn't this type of discussion be in a forum so that 
everybody can have their say?

Regards
Dave F.



 





___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Mapnik rendering of Marinas changed

[Vladimir Vyskocil]

> 
> Count me as a complainer... I was wondering about that blue dashed line
> too - it does not fit well with the surrounding estuary and open sea.
> 
> I'm in the process of mapping a marina and I spontaneously only tagged
> as a leisure=marina the water portion that is occupied by the harbour.

I think it's better now, as marina should be carved in the land by the 
coastline and nothing forbid to use also the marina tag.


___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Mapnik rendering of Marinas changed

[Jonas Häggqvist]

On 17-04-2010 14:38, Patrick Kilian wrote:
> Hi all,
>> Count me as a complainer... I was wondering about that blue dashed line
>> too - it does not fit well with the surrounding estuary and open sea.
>>
>> I'm in the process of mapping a marina and I spontaneously only tagged
>> as a leisure=marina the water portion that is occupied by the harbour.
> Open a trac ticket for mapnik then. Neither Steve nor the others in that
> discussion had a strong opinion. Note that I didn't take part in that
> discussion but merely witnessed it.

http://trac.openstreetmap.org/ticket/1681 contains a few back and forths, 
but as Patrick says, there was also some discussion on IRC (I wouldn't 
call it discussion, since there wasn't any disagreement, and a lot of 
indifference).

The way I read the wiki at the time (from various pages), leisure=marina 
was for the land belonging to the marina. Seeing it rendered as sea 
sparked my initial interest in this.

No matter which is "correct", the previous method of rendering 
leisure=marina as sea does not make much sense, since it effectively means 
not rendering it at all. Rendering it as a border means you can use it for 
either the land, sea or both, and have it rendered in a useful fashion. 
Seems ideal to me.

-- 
Jonas Häggqvist
rasher(at)rasher(dot)dk

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

[OSM-talk] Mapnik rendering labels for unrecognised tags

[Steve Bennett]

I'm not sure if this is a recent change (or I've just noticed it), but
it seems that tags that don't contain anything recognisable to mapnik
other than a name are getting rendered: http://osm.org/go/uG42g@6EB--

That "Melbourne;Geelong..." label is on a way that defines the edge of
the Nearmap imagery, and shouldn't be rendered.

If the new policy is that anything in the database is liable to be
rendered by the default mapnik stylesheet, then perhaps we need a way
to indicate objects that shouldn't be rendered (perhaps osm_meta=yes
?)

[Of course this change(?) has some benefits: the nearby Winneke
Treatment Plant which I just mapped (http://osm.org/go/uG42SFtBP--)
gets kind of rendered, even though there's no specific support for
man_made=waterworks.]

Steve

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

[OSM-talk] Mapnik rendering problem (was: Cycle route improvements)

[Ben Laenen]

On Wednesday 06 February 2008, Chris Jones wrote:
> For mapnik see http://wiki.openstreetmap.org/index.php/Mapnik
>
> For sections of the planet see
> http://wiki.openstreetmap.org/index.php/Planet.osm#Extracts

Thanks, it was also possible by just downloading an area with JOSM.

Anyway, my Mapnik installation refuses to render things like parks, 
cemeteries, woods, landuse areas etc. i.e. no polygons show up, except 
for water which is rendered nicely like everything else on the map. 
Does anyone know what could be the problem? I have absolutely no idea 
what could be wrong, and I'm just using the osm-template.xml file from 
svn...

Greetings
Ben

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk

Re: [OSM-talk] Mapnik rendering export has only coastline

[Richard Fairhurst]

Ben Ward wrote:
> This looks like a bug/problem with the Openstreetmap Mapnik 
> Export rendering.  Can anyone confirm, or fix?

Mapnik export doesn't work on Wednesdays while the database is reloaded. I
believe there's an intention to fix this in the medium term (help welcome no
doubt). Meanwhile I'll add it to the FAQ on the wiki.

cheers
Richard
-- 
View this message in context: 
http://www.nabble.com/Mapnik-rendering-export-has-only-coastline-tp22955618p22955846.html
Sent from the OpenStreetMap - General mailing list archive at Nabble.com.


___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Mapnik rendering export has only coastline

[Chris Hill]

Richard Fairhurst wrote:
> Ben Ward wrote:
>   
>> This looks like a bug/problem with the Openstreetmap Mapnik 
>> Export rendering.  Can anyone confirm, or fix?
>> 
>
> Mapnik export doesn't work on Wednesdays while the database is reloaded. I
> believe there's an intention to fix this in the medium term (help welcome no
> doubt). Meanwhile I'll add it to the FAQ on the wiki. 
>   
Maybe we should remove the Export tab when it is out of commision?

Cheers, Chris

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Mapnik rendering export has only coastline

[Tom Hughes]

Chris Hill wrote:
> Richard Fairhurst wrote:
>> Ben Ward wrote:
>>   
>>> This looks like a bug/problem with the Openstreetmap Mapnik 
>>> Export rendering.  Can anyone confirm, or fix?
>>> 
>> Mapnik export doesn't work on Wednesdays while the database is reloaded. I
>> believe there's an intention to fix this in the medium term (help welcome no
>> doubt). Meanwhile I'll add it to the FAQ on the wiki. 
>>   
> Maybe we should remove the Export tab when it is out of commision?

Yes, because the users of all the other export modes that aren't 
dependent on the mapnik database would love that.

Tom

-- 
Tom Hughes (t...@compton.nu)
http://www.compton.nu/

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Mapnik rendering export has only coastline

[Simon Ward]

On Wed, Apr 08, 2009 at 08:54:30PM +0100, Tom Hughes wrote:
> > Maybe we should remove the Export tab when it is out of commision?
> 
> Yes, because the users of all the other export modes that aren't 
> dependent on the mapnik database would love that.

There’s nothing like a bit of dry sarcasm to cheer the place up.

The whole Export tab is never out of commission, so how about just
disabling the Mapnik exports when we know they are not going to work?

Simon
-- 
A complex system that works is invariably found to have evolved from a
simple system that works.—John Gall


signature.asc
Description: Digital signature
___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Mapnik rendering export has only coastline

[Jon Burgess]

On Wed, 2009-04-08 at 21:54 +0100, Simon Ward wrote:
> On Wed, Apr 08, 2009 at 08:54:30PM +0100, Tom Hughes wrote:
> > > Maybe we should remove the Export tab when it is out of commision?
> > 
> > Yes, because the users of all the other export modes that aren't 
> > dependent on the mapnik database would love that.
> 
> There’s nothing like a bit of dry sarcasm to cheer the place up.
> 
> The whole Export tab is never out of commission, so how about just
> disabling the Mapnik exports when we know they are not going to work?

The current plan is to enhance the Mapnik update process so that the
weekly import will go into a temporary DB. The export to can carry on
rendering the existing data during the import. Then old data will be
dropped and the new data put in its place which should only take a few
seconds.

The above plan is fine in principle but I need to work out an automated
script which does all the steps in the right sequence an make sure that
the partition containing the database does not run out of disk space.

Jon



___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Mapnik rendering issue with *_link roads

[Richard Mann]

Logically, you need to know the lower of the two classifications being
linked, and it may also be useful to know the higher of the two being
linked. So I record that information in links_lower and links_higher tags.
Then it can be rendered very neatly.

But I got flamed last time I proposed this ought to be fixed, so this
information is offered on a tag-what-you-like basis, rather than a
render-as-I-do basis.

Richard

On Fri, May 4, 2012 at 1:09 PM, Maarten Deen  wrote:

> I finally found out what the issue is with this [1] situation. My issue
> with it is that IMHO it is not good to see the unclassified road rendered
> on top of the primary_link road. I made a test here [2] where the left
> primary is a primary_link and the right primary is a primary proper.
> Apparently mapnik's renderering rules state that primary_link roads should
> be rendered below unclassified roads. I haven't inspected the rendering
> rules in detail (I'm also not familiar with them) but I have seen that
> tertiary_link roads are also rendered below unclassified roads. I expect
> that any *_link road is rendered below any other (or at least motorway -
> unclassified) road.
>
> Is this behaviour of mapnik wanted? As I said: IMHO it is not pleasing to
> the eye to see the unclassified road rendered on top of the primary_link
> road. In order of priority, a *_link road is just below its * counterpart
> but above the next lower road (so primary -> primary_link -> secondary).
>
> [1] 
> http://www.openstreetmap.org/?**lat=51.352555&lon=6.014996&**zoom=18
> [2] 
> http://www.openstreetmap.org/?**lat=51.352419&lon=6.010627&**zoom=18
>
> Regards,
> Maarten
>
> __**_
> talk mailing list
> talk@openstreetmap.org
> http://lists.openstreetmap.**org/listinfo/talk
>
___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Mapnik rendering issue with *_link roads

[AJ Ashton]

On Fri, May 4, 2012 at 8:09 AM, Maarten Deen  wrote:
> Is this behaviour of mapnik wanted? As I said: IMHO it is not pleasing to
> the eye to see the unclassified road rendered on top of the primary_link
> road. In order of priority, a *_link road is just below its * counterpart
> but above the next lower road (so primary -> primary_link -> secondary).

I would guess that, yes, this is the intention. The example you point
out is only minorly aesthetically displeasing. But if links were
rendered on top of unclassified roads, the situation of a link merging
into an unclassified (rather than passing through) would look much
worse. Example:

http://dl.dropbox.com/u/2398828/scrot/link_order.png

Granted, links don't feed into unclassifieds as often as they do
higher classifications of road, but it still happens a lot.

-- 
AJ Ashton

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Mapnik rendering issue with *_link roads

[Richard Mann]

That's why you need to know the lower of the two classifications being
linked (so you can put the link just under the lower one)

On Fri, May 4, 2012 at 4:11 PM, AJ Ashton  wrote:

> On Fri, May 4, 2012 at 8:09 AM, Maarten Deen  wrote:
> > Is this behaviour of mapnik wanted? As I said: IMHO it is not pleasing to
> > the eye to see the unclassified road rendered on top of the primary_link
> > road. In order of priority, a *_link road is just below its * counterpart
> > but above the next lower road (so primary -> primary_link -> secondary).
>
> I would guess that, yes, this is the intention. The example you point
> out is only minorly aesthetically displeasing. But if links were
> rendered on top of unclassified roads, the situation of a link merging
> into an unclassified (rather than passing through) would look much
> worse. Example:
>
> http://dl.dropbox.com/u/2398828/scrot/link_order.png
>
> Granted, links don't feed into unclassifieds as often as they do
> higher classifications of road, but it still happens a lot.
>
> --
> AJ Ashton
>
> ___
> talk mailing list
> talk@openstreetmap.org
> http://lists.openstreetmap.org/listinfo/talk
>
___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Mapnik rendering labels for unrecognised tags

[Tom Hughes]

On 29/08/11 09:44, Steve Bennett wrote:


I'm not sure if this is a recent change (or I've just noticed it), but
it seems that tags that don't contain anything recognisable to mapnik
other than a name are getting rendered: http://osm.org/go/uG42g@6EB--


It's always been the case that names sometimes get rendered for objects 
that haven't been rendered, because the names are produced by separate 
rendering rules and trying to attach to those rules a set of filters 
which match the set of objects which are rendered is more or less 
impossible.


It's a bug in the mapnik stylesheet basically, but not one which is very 
to fix.


Tom

--
Tom Hughes (t...@compton.nu)
http://compton.nu/

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Mapnik rendering labels for unrecognised tags

[SomeoneElse]

On 29/08/2011 09:44, Steve Bennett wrote:

I'm not sure if this is a recent change (or I've just noticed it), but
it seems that tags that don't contain anything recognisable to mapnik
other than a name are getting rendered: http://osm.org/go/uG42g@6EB--


It's not a new thing, I don't think - in early 2009 I opened this:
http://trac.openstreetmap.org/ticket/1514

Cheers,
Andy


___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Mapnik rendering labels for unrecognised tags

[yvecai]

Maybe it is something for osm2pgsql to deal with ?
Yves


"SomeoneElse"  a écrit :

>On 29/08/2011 09:44, Steve Bennett wrote:
>> I'm not sure if this is a recent change (or I've just noticed it), but
>> it seems that tags that don't contain anything recognisable to mapnik
>> other than a name are getting rendered: http://osm.org/go/uG42g@6EB--
>
>It's not a new thing, I don't think - in early 2009 I opened this:
>http://trac.openstreetmap.org/ticket/1514
>
>Cheers,
>Andy
>
>
>___
>talk mailing list
>talk@openstreetmap.org
>http://lists.openstreetmap.org/listinfo/talk

-- 
Sent from my Android phone with K-9 Mail. Please excuse my brevity.

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Mapnik rendering labels for unrecognised tags

[Steve Bennett]

On Mon, Aug 29, 2011 at 7:03 PM, Tom Hughes  wrote:
> It's always been the case that names sometimes get rendered for objects that
> haven't been rendered, because the names are produced by separate rendering
> rules and trying to attach to those rules a set of filters which match the
> set of objects which are rendered is more or less impossible.
>
> It's a bug in the mapnik stylesheet basically, but not one which is very to
> fix.

Thanks for the explanation. So I guess we should avoid using name=* on
anything which should not be rendered.

Steve

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Mapnik rendering labels for unrecognised tags

[Tom Hughes]

On 29/08/11 15:45, Steve Bennett wrote:


On Mon, Aug 29, 2011 at 7:03 PM, Tom Hughes  wrote:

It's always been the case that names sometimes get rendered for objects that
haven't been rendered, because the names are produced by separate rendering
rules and trying to attach to those rules a set of filters which match the
set of objects which are rendered is more or less impossible.

It's a bug in the mapnik stylesheet basically, but not one which is very to
fix.


Thanks for the explanation. So I guess we should avoid using name=* on
anything which should not be rendered.


Not really, no.

As always you should avoid tagging for the renderer and tag what exists 
on the ground and not worry too much about exactly how things render 
with the current rendering technology, which is bound to improve over time.


Tom

--
Tom Hughes (t...@compton.nu)
http://compton.nu/

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Mapnik rendering labels for unrecognised tags

[Martin Koppenhoefer]

2011/8/29 Steve Bennett :
> On Mon, Aug 29, 2011 at 7:03 PM, Tom Hughes  wrote:
> Thanks for the explanation. So I guess we should avoid using name=* on
> anything which should not be rendered.


+1, IMHO it is also generally disputable to keep coverage-polygons of
other data providers in the OSM main database.

cheers,
Martin

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Mapnik rendering labels for unrecognised tags

[Tom Hughes]

On 29/08/11 16:00, Martin Koppenhoefer wrote:

2011/8/29 Steve Bennett:

On Mon, Aug 29, 2011 at 7:03 PM, Tom Hughes  wrote:
Thanks for the explanation. So I guess we should avoid using name=* on
anything which should not be rendered.



+1, IMHO it is also generally disputable to keep coverage-polygons of
other data providers in the OSM main database.


Yes I should have added that, in this case, my preference would be to 
remove the object altogether as aerial imagery coverage areas are not 
real on the ground objects that should be in our database.


Tom

--
Tom Hughes (t...@compton.nu)
http://compton.nu/

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Mapnik rendering labels for unrecognised tags

[Stephen Hope]

On 30 August 2011 01:01, Tom Hughes  wrote:
> Yes I should have added that, in this case, my preference would be to remove
> the object altogether as aerial imagery coverage areas are not real on the
> ground objects that should be in our database.

Particularly in this specific case, as nearmap coverage can't be used
to derive OSM objects any more.

Stephen

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Mapnik rendering labels for unrecognised tags

[Steve Bennett]

On Tue, Aug 30, 2011 at 10:29 AM, Stephen Hope  wrote:
> Particularly in this specific case, as nearmap coverage can't be used
> to derive OSM objects any more.

The same mechanism is used for Yahoo, Bing etc coverage. Yes, it's
debatable whether meta-objects should be stored in the OSM database,
but that debate would also extend to meta-tags (note=*, fixme=*)

I don't think the (itself questionable) maxim of "don't tag for the
renderer" applies here, as these aren't real world objects.

Steve

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Mapnik rendering labels for unrecognised tags

[Ian Sergeant]

Steve Bennett  wrote on 30/08/2011 11:14:32 AM:

> The same mechanism is used for Yahoo, Bing etc coverage. Yes, it's
> debatable whether meta-objects should be stored in the OSM database,
> but that debate would also extend to meta-tags (note=*, fixme=*)

When I encounter these "meta" ways/nodes with a name tag, I always change 
it to a note tag.  It is more aligned with a annotation than with a named 
object.

More generally I do agree there should be a better place to store this 
meta-layer info, or if not certainly a way to tag it so it can be ignored 
by those only interested in the map features.

Ian.___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Mapnik rendering labels for unrecognised tags

[Anthony]

On Mon, Aug 29, 2011 at 9:22 PM, Ian Sergeant  wrote:
>
> Steve Bennett  wrote on 30/08/2011 11:14:32 AM:
>
>> The same mechanism is used for Yahoo, Bing etc coverage. Yes, it's
>> debatable whether meta-objects should be stored in the OSM database,
>> but that debate would also extend to meta-tags (note=*, fixme=*)

That's pretty different, or at least potentially so.  A note=* or
fixme=* tag, at least in the way I've used them and seen them used, is
just a human-readable rather than machine-readable description of
something in the real world.

> When I encounter these "meta" ways/nodes with a name tag, I always change it
> to a note tag.  It is more aligned with a annotation than with a named
> object.

+1.  While I'd rather see these objects go away altogether, I think a
tag of "name=Melbourne;Geelong;South-Central NSW Area;Central Victoria
Area" on a closed way implies that this closed way represents an area
called "Melbourne;Geelong;South-Central NSW Area;Central Victoria
Area".

I'm not sure boundary=* is appropriate either, as this is not a
political or governmental or pseudo-governmental division, though
maybe I'm just not aware of how broadly that tag is used.

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Mapnik rendering labels for unrecognised tags

[Steve Bennett]

On Tue, Aug 30, 2011 at 12:07 PM, Anthony  wrote:
> +1.  While I'd rather see these objects go away altogether, I think a
> tag of "name=Melbourne;Geelong;South-Central NSW Area;Central Victoria
> Area" on a closed way implies that this closed way represents an area
> called "Melbourne;Geelong;South-Central NSW Area;Central Victoria
> Area".
>
> I'm not sure boundary=* is appropriate either, as this is not a
> political or governmental or pseudo-governmental division, though
> maybe I'm just not aware of how broadly that tag is used.

Cool. Well, since these areas aren't AFAIK used for anything except
these kinds of maps: http://wiki.openstreetmap.org/wiki/Nearmap ,
there's no particular need for them to have tags like name=* or
boundary=*. I've changed this one.

Steve

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Mapnik rendering labels for unrecognised tags

[Anthony]

On Tue, Aug 30, 2011 at 12:15 AM, Steve Bennett  wrote:
> On Tue, Aug 30, 2011 at 12:07 PM, Anthony  wrote:
>> +1.  While I'd rather see these objects go away altogether, I think a
>> tag of "name=Melbourne;Geelong;South-Central NSW Area;Central Victoria
>> Area" on a closed way implies that this closed way represents an area
>> called "Melbourne;Geelong;South-Central NSW Area;Central Victoria
>> Area".
>>
>> I'm not sure boundary=* is appropriate either, as this is not a
>> political or governmental or pseudo-governmental division, though
>> maybe I'm just not aware of how broadly that tag is used.
>
> Cool. Well, since these areas aren't AFAIK used for anything except
> these kinds of maps: http://wiki.openstreetmap.org/wiki/Nearmap ,
> there's no particular need for them to have tags like name=* or
> boundary=*. I've changed this one.

Sounds good.  I don't think storing these in OSM, with the
non-overlapping tags, is harmful.  While I'd love to see them in a
separate database or at least a separate layer, the fact of the matter
is that separate database and/or separate layer hasn't yet really been
implemented.

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Mapnik rendering labels for unrecognised tags

[Lester Caine]

Anthony wrote:

Sounds good.  I don't think storing these in OSM, with the
non-overlapping tags, is harmful.  While I'd love to see them in a
separate database or at least a separate layer, the fact of the matter
is that separate database and/or separate layer hasn't yet really been
implemented.


This is the real problem!
With more and more historic mapping material coming on line, and what seems like 
little support for the start_date/end_date tagging of physical objects that we 
have fairly accurate data on when they did make an appearance or when they were 
redeveloped, linking to other data sources where this information can be stored 
would at least allow it's integration?


And the creation of 'temporary' layers where material is being worked on but has 
not yet been fully integrated would also seem to be a way forward even for 
general mapping?


--
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//
Firebird - http://www.firebirdsql.org/index.php

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Mapnik rendering labels for unrecognised tags

[Anthony]

On Tue, Aug 30, 2011 at 3:36 AM, Lester Caine  wrote:
> Anthony wrote:
>>
>> Sounds good.  I don't think storing these in OSM, with the
>> non-overlapping tags, is harmful.  While I'd love to see them in a
>> separate database or at least a separate layer, the fact of the matter
>> is that separate database and/or separate layer hasn't yet really been
>> implemented.
>
> This is the real problem!
> With more and more historic mapping material coming on line, and what seems
> like little support for the start_date/end_date tagging of physical objects
> that we have fairly accurate data on when they did make an appearance or
> when they were redeveloped, linking to other data sources where this
> information can be stored would at least allow it's integration?
>
> And the creation of 'temporary' layers where material is being worked on but
> has not yet been fully integrated would also seem to be a way forward even
> for general mapping?

Yeah.  I don't think it would be too hard to hack something up that
implements the basic API (enough to run JOSM against) and uses SQLite
as a backend.  This would be runnable by basically anyone, and good
enough for small dataset "layers" like this one.  Add in the ability
to export to .osm format (very simple) and you could basically run
Mapnik against it with no changes, with or without first merging with
other datasets and/or OSM proper.

But then, while I say it "wouldn't be too hard", it would probably
take a few weeks of full time coding, which unfortunately is something
I don't have for something that isn't going to help me pay the bills.
So as much as I hate to see these things in the main db, I'm accepting
of it.

Hopefully someone else will read this email, understand what I'm
talking about, and love the idea, though :).

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Mapnik rendering problem (was: Cycle route improvements)

[Artem Pavlenko]

On 6 Feb 2008, at 16:23, Ben Laenen wrote:

> On Wednesday 06 February 2008, Chris Jones wrote:
>> For mapnik see http://wiki.openstreetmap.org/index.php/Mapnik
>>
>> For sections of the planet see
>> http://wiki.openstreetmap.org/index.php/Planet.osm#Extracts
>
> Thanks, it was also possible by just downloading an area with JOSM.
>
> Anyway, my Mapnik installation refuses to render things like parks,
> cemeteries, woods, landuse areas etc. i.e. no polygons show up, except
> for water which is rendered nicely like everything else on the map.
> Does anyone know what could be the problem? I have absolutely no idea
> what could be wrong, and I'm just using the osm-template.xml file from
> svn...

Ben,

First thing to check if you have parks,woods etc in postgresql.

try:

select count(1) from planet_osm_polygons where "natural" = 'wood' ;

or similar
>
> Greetings
> Ben
Artem
>
> ___
> talk mailing list
> talk@openstreetmap.org
> http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk


___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk

Re: [OSM-talk] Mapnik rendering problem (was: Cycle route improvements)

[Ben Laenen]

On Wednesday 06 February 2008, Artem Pavlenko wrote:
> Ben,
>
> First thing to check if you have parks,woods etc in postgresql.
>
> try:
>
> select count(1) from planet_osm_polygons where "natural" = 'wood' ;
>
> or similar

It all looks well. There are hundreds of them in the database (and I 
also made sure the ones I want to see rendered are in there). So it 
should be something else...

Ben

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk

Re: [OSM-talk] Mapnik rendering problem (was: Cycle route improvements)

[Artem Pavlenko]

On 6 Feb 2008, at 17:08, Ben Laenen wrote:

> On Wednesday 06 February 2008, Artem Pavlenko wrote:
>> Ben,
>>
>> First thing to check if you have parks,woods etc in postgresql.
>>
>> try:
>>
>> select count(1) from planet_osm_polygons where "natural" = 'wood' ;
>>
>> or similar
>
> It all looks well. There are hundreds of them in the database (and I
> also made sure the ones I want to see rendered are in there). So it
> should be something else...

OK, good.
Probably something wrong with your styles. Post your *.xml file and  
I'll have a look.
>
> Ben
Artem


___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk

Re: [OSM-talk] Mapnik rendering problem (was: Cycle route improvements)

[Artem Pavlenko]

On 6 Feb 2008, at 17:17, Ben Laenen wrote:

> On Wednesday 06 February 2008, Artem Pavlenko wrote:
>> OK, good.
>> Probably something wrong with your styles. Post your *.xml file and
>> I'll have a look.
>
> It's the original one from svn, but if you want to have a look... I
> can't see anything wrong with it, but I'll try some older ones once.

Ben,

Are you using template ? You should have replaced all occurrences of   
%params% with real things
Also, make sure you have postgis.input installed

HTH
Artem
>
> Greetings
> Ben


___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk

Re: [OSM-talk] Mapnik rendering problem (was: Cycle route improvements)

[Ben Laenen]

On Wednesday 06 February 2008, Artem Pavlenko wrote:
> Are you using template ? You should have replaced all occurrences of
> %params% with real things

I don't have to do that manually, do I? I just enter the variables in 
set-mapnik-env, call customize-mapnik-map which replaces all %param% in 
osm-template.xml which is saved to osm.xml.

The evidence that it's working is that all symbols like for pubs and 
stations are shown on the map for example.

> Also, make sure you have postgis.input installed

I don't know what you mean with postgis".input" exactly, but I've 
installed postgresql8.2-postgis and ran it on the database, just like 
is mentioned on http://wiki.openstreetmap.org/index.php/Mapnik

Are there some alternative osm.xml files somewhere which I could test 
out?

Ben

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk

Re: [OSM-talk] Mapnik rendering problem (was: Cycle route improvements)

[Artem Pavlenko]

On 6 Feb 2008, at 19:29, Ben Laenen wrote:

> On Wednesday 06 February 2008, Artem Pavlenko wrote:
>> Are you using template ? You should have replaced all occurrences of
>> %params% with real things
>
> I don't have to do that manually, do I? I just enter the variables in
> set-mapnik-env, call customize-mapnik-map which replaces all %param 
> % in
> osm-template.xml which is saved to osm.xml.
>
> The evidence that it's working is that all symbols like for pubs and
> stations are shown on the map for example.

>
>> Also, make sure you have postgis.input installed
>
> I don't know what you mean with postgis".input" exactly, but I've
> installed postgresql8.2-postgis and ran it on the database, just like
> is mentioned on http://wiki.openstreetmap.org/index.php/Mapnik
>
> Are there some alternative osm.xml files somewhere which I could test
> out?

Ok, if you have pubs etc showing means you have mapnik (+postgis plug- 
in) installed. Could you compile Mapnik with DEBUG=y and see SQL  
statements being sent to postmaster ?

Oh, one more thing, just an idea :

try running
vacuum analyze planet_osm_;

Artem

>
> Ben


___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk

Re: [OSM-talk] Mapnik rendering problem (was: Cycle route improvements)

[Jon Burgess]

On Wed, 2008-02-06 at 20:29 +0100, Ben Laenen wrote:
> On Wednesday 06 February 2008, Artem Pavlenko wrote:
> > Are you using template ? You should have replaced all occurrences of
> > %params% with real things
> 
> I don't have to do that manually, do I? I just enter the variables in 
> set-mapnik-env, call customize-mapnik-map which replaces all %param% in 
> osm-template.xml which is saved to osm.xml.

That sounds fine.

> The evidence that it's working is that all symbols like for pubs and 
> stations are shown on the map for example.

You are getting other data for Postgres OK so the input plugin is
working. The only time I saw something similar recently was when the
osm.xml had the following:

true

This caused it to lose parks other polygon features in some places of
the map. Setting it to false fixed this.

Jon



___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk

[OSM-talk] Mapnik rendering of paths + place=locality in general

[Igor Brejc]

Hello,

This question has come up several times (from what I could Google out), 
but let me ask it again: when can we expect Mapnik OSM layer to render 
highway=path? Nevermind the "path or footway" debate - it would be nice 
to have paths rendered (they can be rendered the same as footways, as 
far as I'm concerned). Anyway, I checked Tagwatch for Germany and it has 
some 17,000 paths, so it's not totally unused.

Second issue: I've just added an area tagged with place=locality, but 
Osmarender doesn't seem to render it (I don't know about Mapnik). This 
would be a nice feature to have (especially for cycling map).

Igor

-- 
http://igorbrejc.net


___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

[OSM-talk] Mapnik rendering of paths + place=locality in general

[Lauri Kytömaa]

>So: Someone thought that dropping the well-established and well-working 
>highway=footway was a good idea. He or she seems to have managed to 
>convince people to actually *change* existing data to fit his new idea, 
>without, obviously, spending a second thinking about the data consumers 
>(i.e. renderers and others).

I was the one who was quite active in the discussion about the proposal, 
but i *really* don't know how the Germans got the impression or decided 
that all or any footways would better be paths, or even any signed 
footways...

If I now look closely at the proposals wording, there's the clause "or 
intended for several uses", on which they might have constructed their 
reasoning. Yet a foot=designated on a cycleway is, IMO, equally sufficient 
to, well, designate it for both uses equally. Then a highway=path is only 
for, as the current wording in Tag:highway=path reads, "Paths for which 
*all and any* of highway=footway, highway=cycleway and highway=bridleway 
would be inappropriate or inadequate, yet usable for travel or 
navigation". This leaves for path any hiking trails and urban unofficial 
and totally unmaintained and unofficial narrow paths, i.e. the minor ones 
that snip a corner of a park or a forest, but also the snowmobile routes 
and others which just aren't footways nor tracks.

-- 
alv

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Mapnik rendering of paths + place=locality in general

[Tom Hughes]

Igor Brejc wrote:

> This question has come up several times (from what I could Google out), 
> but let me ask it again: when can we expect Mapnik OSM layer to render 
> highway=path? Nevermind the "path or footway" debate - it would be nice 
> to have paths rendered (they can be rendered the same as footways, as 
> far as I'm concerned). Anyway, I checked Tagwatch for Germany and it has 
> some 17,000 paths, so it's not totally unused.

How about if you want them rendered you tag them with the standard 
tagging of highway=footway that we've been using for the last several 
years and which mapnik already supports?

It's no use saying "never mind the debate". By that logic I could 
announce that in future I'm going to tag all motorways as "highway=frog" 
and demand that people "never mind the debate" and just render 
highway=frog as a motorway.

Tom

-- 
Tom Hughes ([EMAIL PROTECTED])
http://www.compton.nu/

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Mapnik rendering of paths + place=locality in general

[Steve Chilton]

Wednesday - patch by Bernhard H integrated and submitted yesterday.
Locality has been in mapnik style for a while and should render fine.

Cheers
STEVE

Steve Chilton, Learning Support Fellow
Manager of e-Learning Academic Development
Centre for Learning and Quality Enhancement
Middlesex University
phone/fax: 020 8411 5355
email: [EMAIL PROTECTED]
http://www.mdx.ac.uk/schools/hssc/staff/profiles/technical/chiltons.asp

Chair of the Society of Cartographers: http://www.soc.org.uk/

SoC conference 2008:
http://www.abdn.ac.uk/cartographers08/

-Original Message-
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of Igor Brejc
Sent: 03 November 2008 16:14
To: talk@openstreetmap.org
Subject: [OSM-talk] Mapnik rendering of paths + place=locality in
general

Hello,

This question has come up several times (from what I could Google out), 
but let me ask it again: when can we expect Mapnik OSM layer to render 
highway=path? Nevermind the "path or footway" debate - it would be nice 
to have paths rendered (they can be rendered the same as footways, as 
far as I'm concerned). Anyway, I checked Tagwatch for Germany and it has

some 17,000 paths, so it's not totally unused.

Second issue: I've just added an area tagged with place=locality, but 
Osmarender doesn't seem to render it (I don't know about Mapnik). This 
would be a nice feature to have (especially for cycling map).

Igor

-- 
http://igorbrejc.net


___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk



___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Mapnik rendering of paths + place=locality in general

[Igor Brejc]

Tom Hughes wrote:
> Igor Brejc wrote:
>
>> This question has come up several times (from what I could Google 
>> out), but let me ask it again: when can we expect Mapnik OSM layer to 
>> render highway=path? Nevermind the "path or footway" debate - it 
>> would be nice to have paths rendered (they can be rendered the same 
>> as footways, as far as I'm concerned). Anyway, I checked Tagwatch for 
>> Germany and it has some 17,000 paths, so it's not totally unused.
>
> How about if you want them rendered you tag them with the standard 
> tagging of highway=footway that we've been using for the last several 
> years and which mapnik already supports?
>
> It's no use saying "never mind the debate". By that logic I could 
> announce that in future I'm going to tag all motorways as 
> "highway=frog" and demand that people "never mind the debate" and just 
> render highway=frog as a motorway.
>
> Tom
>
Ok, but is there anywhere in OSM world a law that says "use footways and 
not paths"? Aren't we constantly reiterating the fact that tagging is 
"democratic"? Isn't there a description of "highway=path" on Map 
Features page?

I don't see a reason (and I don't intend to continue) the debate about 
this issue. I was merely asking when/if this feature will be rendered in 
Mapnik. And I didn't announce any plans to use or disuse this tag.

Best regards,
Igor

-- 
http://igorbrejc.net


___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Mapnik rendering of paths + place=locality in general

[Andy Allan]

On Mon, Nov 3, 2008 at 4:49 PM, Igor Brejc <[EMAIL PROTECTED]> wrote:
> Aren't we constantly reiterating the fact that tagging is
> "democratic"?

I think democratic is the wrong word. You can do what you like.
There's no tyranny of the masses, for a start - minorities and
majorities are both on equal footing with tagging. Consuming data is
entirely up to whoever is doing it to use or ignore OSM data as they
see fit.

> Isn't there a description of "highway=path" on Map
> Features page?

Ahahahahahahaha. So what?

Cheers,
Andy

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk