Re: [OSM-talk-be] Fwd: [osmose-backend] Missing Parent Tag for all highways with bicycle = yes (#1)

2013-03-24 Thread Marc Gemis
I think anyone can remove them. Of course, it would be better to also
contact whoever added them, so they do not make the same mistake again.

Osmose seems to be rather stupid, and decides that if x number of objects
has certain tag combinations, all objects that have one of the tags, should
also have the others. It does not verify if the tag is appropriate for the
object (node, way, relation). But it is still a valuable tool.

So feel free to remove the route=bicycle if you have the time

regards

m




On Sun, Mar 24, 2013 at 12:45 AM, André Pirard Papou 
a.pirard.pa...@gmail.com wrote:

  On 2013-03-22 21:32, Marc Gemis wrote :

 This is the reply I got on my bug report to osmose.

  Do we have to locate those ways with route=bicycle ?

  m


 It makes no sense.
 route=bicycle is a tag for a relation, not for a way.
 It's very surprising that Osmose commands to repeat such a tag on ever
 cycling way !!!
 How could we identify a cycling route if all the cycling ways contained
 route=bicycle 

 Either Osmose removes that test.
 Or the authors remove that tag (I bcc: one of them).
 Or I can remove 28 of them that I have ready after my selection.

 Your choice?

 Cheers,

   André.

  -- Forwarded message --
 From: frodrigo notificati...@github.com
 Date: Fri, Mar 22, 2013 at 9:12 PM
 Subject: Re: [osmose-backend] Missing Parent Tag for all highways with
 bicycle = yes (#1)
 To: osm-fr/osmose-backend osmose-back...@noreply.github.com
 Cc: marcgemis marc.ge...@gmail.com


 Missing parent tag is a statistical analysis. Osmose suggest to add
 route=bicycle because this tag value is already a key as bicycle=yes, and
 more over because there is more than 50 ways with route=bicycle +
 bicycle=yes in Belgium.

 Maybe we need to apply this analysis on relative value inside of just 50.

 —
 Reply to this email directly or view it on 
 GitHubhttps://github.com/osm-fr/osmose-backend/issues/1#issuecomment-15319079
 .



 ___
 Talk-be mailing 
 listTalk-be@openstreetmap.orghttp://lists.openstreetmap.org/listinfo/talk-be



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


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


Re: [OSM-talk-be] Meeting with SPW (some elements of the discussion may also interest Flemish people)

2013-03-24 Thread Jo
2013/3/23 Julien Fastré jul...@fastre.info

 Hi,

 This is some news following the contact that eMerzh, Benoit Coumont and
 I had with the SPW (Service Public de Wallonie) on Friday.

 The meeting was rather positive. They think about applying an
 ODbL-compatible licence on some data.

 - at first, they will think about letting us use the new orthophoto
 (precision 25cm). The year-2012 pictures will be published on the
 website before summer ;
 - we agreed on the fact that beginning slowly is better. They will
 discuss about releasing three kind of data under open-licence:
 - hydrological data (there isn't a lot of rivers and streams in the db)
 - path and footways, because we were thinking that osm is quite
 useful for hikers, we thought it was an added value for everybody ;
 - and Arbres et haies remarquables because it is easy and quite
 fun :-)

 They were speaking having a decision for their master plan which
 should be adopted in September 2013. But they also added that it will be
 easy to add opendata items in their master plan if a first experience
 works before. So, one of those items could be released before.

 The SPW has also a request for our community (this is were flemish
 people might be more interested !)

 They would like to use OSM to track changes in Wallonia. They would like
 to know where to send their teams, where there is some new roads, or
 where the roads or elements are modified.

 The problem is: how to track those changes ? If I add a new road in OSM,
 is it a road build in the past month, or is it a road build years ago,
 but which didn't exist in the osm'db ? And if I update a line highway,
 am I correcting some mistake, or adapting our map to some new reality ?

 We thought about one solution: a tag since to object which are updated
 or build in the reality and modified in the db. Example: if a road has a
 new bike lane, I update the line with JOSM and add the tag
 since=2013-01-01 on the object. Within SPW, they track the tags
 since and see where are changes.

 What do you think ?

 I think start_date is indeed what you want, then they can use Overpass API
regularly to download all the elements with a start date.

How was the presentation you gave on Tuesday received?

Jo
___
Talk-be mailing list
Talk-be@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk-be


[OSM-talk-be] Fwd: [osmose-backend] Missing Parent Tag for all highways with bicycle = yes (#1)

2013-03-24 Thread Frédéric Rodrigo

 It makes no sense.
 route=bicycle is a tag for a relation, not for a way.
 It's very surprising that Osmose commands to repeat such a tag on
 ever cycling way !!!
 How could we identify a cycling route if all the cycling ways
 contained route=bicycle 

 Osmose seems to be rather stupid, and decides that if x number of
 objects has certain tag combinations, all objects that have one of
 the tags, should also have the others. It does not verify if the tag
 is appropriate for the object (node, way, relation). But it is still
 a valuable tool.

It's an automatic analysis based on statistical and an arbitrary trigger 
at 50 count. I can change the tigger, eg with relative value.


Error was triggered with a parent tag like in highway=construction and 
construction=primary.


The analysis have already a tag blacklist. I can add route=bicycle.

 These tags were used on some ways for some routes: if a route goes
 over a big square, an additional way is created going across the
 square.

Really need tag on this ways ? Or add the area to route relation, or 
just no way at all.


 and I guess osmose wouldn't like untagged ways either

Osmose report error on untagged way not part of relation (with role).

 But I haven't found any ways with both route=bicycle and bicycle=yes
 though.

I found lot of them here :
http://overpass-turbo.eu/?Q=way(%7B%7Bbbox%7D%7D)%5B%22route%22%3D%22bicycle%22%5D%5B%22bicycle%22%3D%22yes%22%5D%3B(._%3B%3E%3B)%3Bout%20meta%3B%0AC=50.67634;4.60636;12R

There is some choses:
- change the data
- black list route=bicycle
- use relative value on analyser to trigger at higher count.

Frédéric
Osmose analyser


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


[OSM-talk-be] Re : Re: Meeting with SPW (some elements of the discussion may also interest Flemish people)

2013-03-24 Thread Julien Fastré
Ben,

The process should tend to be the most automatised possible. That's why we 
should think about this tag 'since' or 'start_date' (which seems to be more 
appropriate because already used).

Reading your answer you seems to have perfectly understood :-)

Julien

Envoyé depuis mon téléphone

Ben Abelshausen ben.abelshau...@gmail.com a écrit :

Julien,

Starting slow is good I think and the part about updating is very very
interesting!

Can't they judge an update based on a comparison between OSM and their own
data? Or is this supposed to be an automated process?

Can't they also ask users directly why they made a change if there is some
doubt about something?

(I probably don't understand completely! :-) )

Met vriendelijke groeten,
Best regards,

Ben Abelshausen
ben.abelshau...@gmail.com ben.abelshau...@gmail.be
http://twitter.com/xivk http://twitter.com/xivk



On Sat, Mar 23, 2013 at 9:55 PM, Julien Fastré jul...@fastre.info wrote:

 Hi,

 This is some news following the contact that eMerzh, Benoit Coumont and
 I had with the SPW (Service Public de Wallonie) on Friday.

 The meeting was rather positive. They think about applying an
 ODbL-compatible licence on some data.

 - at first, they will think about letting us use the new orthophoto
 (precision 25cm). The year-2012 pictures will be published on the
 website before summer ;
 - we agreed on the fact that beginning slowly is better. They will
 discuss about releasing three kind of data under open-licence:
 - hydrological data (there isn't a lot of rivers and streams in the db)
 - path and footways, because we were thinking that osm is quite
 useful for hikers, we thought it was an added value for everybody ;
 - and Arbres et haies remarquables because it is easy and quite
 fun :-)

 They were speaking having a decision for their master plan which
 should be adopted in September 2013. But they also added that it will be
 easy to add opendata items in their master plan if a first experience
 works before. So, one of those items could be released before.

 The SPW has also a request for our community (this is were flemish
 people might be more interested !)

 They would like to use OSM to track changes in Wallonia. They would like
 to know where to send their teams, where there is some new roads, or
 where the roads or elements are modified.

 The problem is: how to track those changes ? If I add a new road in OSM,
 is it a road build in the past month, or is it a road build years ago,
 but which didn't exist in the osm'db ? And if I update a line highway,
 am I correcting some mistake, or adapting our map to some new reality ?

 We thought about one solution: a tag since to object which are updated
 or build in the reality and modified in the db. Example: if a road has a
 new bike lane, I update the line with JOSM and add the tag
 since=2013-01-01 on the object. Within SPW, they track the tags
 since and see where are changes.

 What do you think ?

 Julien Fastré

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


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


Re: [OSM-talk-be] AGIV watuning

2013-03-24 Thread Jo
Unfortunately they don't work for me. This is the one I learned to live
with:

wms:
http://wms.agiv.be/ogc/wms/omkl?FORMAT=image/jpegVERSION=1.1.1SERVICE=WMSREQUEST=GetMapLAYERS=OrthoSTYLES=SRS={proj}WIDTH={width}HEIGHT={height}BBOX={bbox}

Jo

2013/3/25 André Pirard a.pirard.pa...@gmail.com

  Hi,

 Aren't these two JOSM configurations what you wanted?
 Do they not work?
 Aren't you satisfied?


 wms:
 http://grb.agiv.be/geodiensten/raadpleegdiensten/geocache/?FORMAT=image/jpegVERSION=1.1.1SERVICE=WMSREQUEST=GetMapSTYLES=SRS=
 {proj}WIDTH={width}HEIGHT={height}BBOX={bbox}LAYERS=orthoklmx

 wms:
 http://grb.agiv.be/geodiensten/raadpleegdiensten/geocache/?FORMAT=image/pngVERSION=1.1.1SERVICE=WMSREQUEST=GetMapSTYLES=SRS=
 {proj}WIDTH={width}HEIGHT={height}BBOX={bbox}TRANSPARENT=YESLAYERS=grb_bsk,grb_gbg,grb_sel,grb_adp

 Cheers,

   André.
 On 2013-03-19 03:17, A.Pirard wrote :

 On 2013-03-18 19:53, Jo wrote :

 zoom in on the area of interest on this site:

 http://ogc.beta.agiv.be/gdiviewer/?simple=true

 That  viewer issues requests like this:

 http://grb.agiv.be/geodiensten/raadpleegdiensten/geocache/tms/1.0.0/orthoklm@BPL72VL/13/4828/3716.png
 (for some beta time anyway)

 The zoom goes up to 15, but the maximum resolution it at 13 (as shown)
 after which the server stretches the image (scales it).
 These are nonstandard zoom levels to which JOSM's zoom level should be
 adjusted.
 JOSM configuration should be
 tms:
 http://grb.agiv.be/geodiensten/raadpleegdiensten/geocache/tms/1.0.0/orthoklm@WGS84VL/
 {zoom-*z*}/{x}/{y}.png
 But I was unable to find a working *z*.  That's always the problem with
 TMS when it's not standard.
 x and y could be swapped or x be bottom up, etc... A puzzle if it's not
 documented.
 The advantage of TMS is that the images (tiles) are not modified by the
 server.

 Anyway, the corresponding WMS configurations are:

 wms:
 http://grb.agiv.be/geodiensten/raadpleegdiensten/geocache/?FORMAT=image/jpegVERSION=1.1.1SERVICE=WMSREQUEST=GetMapSTYLES=SRS=
 {proj}WIDTH={width}HEIGHT={height}BBOX={bbox}LAYERS=orthoklmx

 It seems to me to have fairly the same resolution as the viewer.
 Don't forget to right-click-change resolution on the layer, especially
 after you zoomed in.

 wms:
 http://grb.agiv.be/geodiensten/raadpleegdiensten/geocache/?FORMAT=image/pngVERSION=1.1.1SERVICE=WMSREQUEST=GetMapSTYLES=SRS=
 {proj}WIDTH={width}HEIGHT={height}BBOX={bbox}TRANSPARENT=YESLAYERS=grb_bsk,grb_gbg,grb_sel,grb_adp

 png is (much) better for the grb_* and jpeg for the orthos (equivalent but
 smaller images).
 Unfortunately, the grb images are not transparent (as in Wallonie).
 I still used TRANSPARENT=YES in case they would.
 Play with the layers' transparency if you need so.

 compare the maximum resolution after you
 Install a wms layer on JOSM:
 I used this uri, but maybe it can be 'optimised':

 wms[12]:
 http://wms.agiv.be/ogc/wms/omkl?FORMAT=image/jpegVERSION=1.1.1SERVICE=WMSREQUEST=GetMapLAYERS=OrthoSTYLES=SRS={proj}WIDTH={width}HEIGHT={height}BBOX={bbox}

 The 12 between [] stands for the maximum zoom level. I tried with values
 of 19, 21, 25, but it doesn't seem to have any effect. So I guess one could
 also simply omit it.

 The zoom limit prevents querying servers that return blank or error images
 when the resolution limit is exceeded.
 If you can determine the JOSM zoom level corresponding to the resolution
 limit (13 for TMS) you can use it.
 This will prevent uselessly getting stretched images from the server when
 JOSM itself can stretch them to the screen.

 I also tried to use it as tms layer, but no luck with that either.

 Did you have the same problem as I did (no image at any speed) or another?
 Tell me how you could get TMS working, anyone.


 Thanks for having a look at it.

 It's great to have 2 imagery sources. For some places it works like a time
 machine. For others sometimes Bing is clearer, ofthen AGIV is more
 revealing, as they had the good sense to shoot pictures in wintertime, when
 there is less foliage on the trees.

 It's even greater to have more ;-)  All sorts of maps can help figuring
 what one sees on a photo :-)
 Beware of Bing's error offset. Always zoom out to check or... check with
 another map.

 At the maximum possible resolution it would be just about perfect.

 I hope it will.  Tell me.  Have fun.

 Cheers,

   Rambo.

 Jo

 Op 18 maart 2013 19:18 schreef A.Pirard.Papou a.pirard.pa...@gmail.comhet 
 volgende:

 On 2013-03-12 08:13, Jo wrote :

  Anderzijds is het wel wat vreemd dat wat JOSM in de WMS-laag laat zien
 duidelijk een lagere resolutie heeft, dan wat er op de website van AGIV
 zelf zichtbaar is. Het klinkt misschien wat ondankbaar, maar weten dat er
 betere kwaliteit beschikbaar is en het dan met minder duidelijke beelden
 moeten doen, is toch wat frustrerend.

 Het zou wel kunnen dat de URI die we gebruiken niet geoptimaliseerd is.
 Ik zal dit ook eens op josm-dev posten en dan vraag ik tegelijk hoe we
 ervoor kunnen zorgen dat JOSM 

[OSM-talk] Crossroad names

2013-03-24 Thread Hans Schmidt
Hello,

some weeks/months ago I started a discussion here that OSM needs to show
crossroad names. I then opened a bug report on the bug tracker, but
there was absolutely no reaction there – not even a comment that it is
not needed (as it is at least usually done on OSS projects).

As I announced, I would not give up so easily, so that is why I write
another mail.

How can we get this feature supported? For comparison, try to imagine
what a European OSM map would look like without street names. We could
try to delete them for a while, and I bet that it would be supported
again in no time.

Thanks.


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


Re: [OSM-talk] Crossroad names

2013-03-24 Thread malenki
Am Sun, 24 Mar 2013 12:48:21 +0100
schrieb Hans Schmidt z0idb...@gmx.de:

 some weeks/months ago I started a discussion here that OSM needs to
 show crossroad names. I then opened a bug report on the bug tracker,
 but there was absolutely no reaction there – not even a comment that
 it is not needed (as it is at least usually done on OSS projects).

It seems there are not a lot of people in need of this feature.

 As I announced, I would not give up so easily, so that is why I write
 another mail.
 
 How can we get this feature supported?

Render a map supporting name= on crossroads. Or make an overlay using
Overpass API.

 For comparison, try to imagine what a European OSM map would look
 like without street names. 

Apples and pears?

regards
Thomas



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


Re: [OSM-talk] Crossroad names

2013-03-24 Thread Tom Hughes

On 24/03/13 11:48, Hans Schmidt wrote:


some weeks/months ago I started a discussion here that OSM needs to show
crossroad names. I then opened a bug report on the bug tracker, but
there was absolutely no reaction there – not even a comment that it is
not needed (as it is at least usually done on OSS projects).

As I announced, I would not give up so easily, so that is why I write
another mail.

How can we get this feature supported? For comparison, try to imagine
what a European OSM map would look like without street names. We could
try to delete them for a while, and I bet that it would be supported
again in no time.


Well the stylesheet is basically stalled at the moment, but we are 
working to revive it.


As always though, there is no infinite pool people waiting to jump on 
tickets and implement them so the best way to get something in is always 
to provide a patch to implement it.


In this case the way to do that would be to work against the new carto 
based stylesheet,which you can find here:


  https://github.com/gravitystorm/openstreetmap-carto

That is not yet in production but we are working to get it into 
production as a way of rebooting the stylesheet maintenance.


Tom

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

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


Re: [OSM-talk] Crossroad names

2013-03-24 Thread Hans Schmidt
Am 24.03.2013 13:07, schrieb malenki:
 It seems there are not a lot of people in need of this feature.
About 160 mio people is not a lot of people?
 Render a map supporting name= on crossroads. Or make an overlay using
 Overpass API.
You miss my point: This is an essential map feature. And why should I
make a map for me personally, if this is a general issue for everyone?

 Apples and pears?
Is it? It seems that you have just made a comment without _any_
knowledge and without even the slightest bit of effort to understand
what this is about.
Sorry, but opinions like yours are ...


For you in an abbreviated form, so that you can understand it:
Crossroad names are in Japan and Korea similar to what street names in
Europe are. They give a help for orientation. Without them, a map in
Japan or Korea is extremely hard to read.

Now you say that there are no people who need this: I would rather say
that the general number of people from Japan or Korea who are using OSM
is low, not the one number of people who would need it. So then we can
analyze why few people from there use OSM. I would definitly say because

1. The entire project is extremely european centric
2. Major features who needed there are not supported (for example
crossroad names)
3. If the map is so unsuitable for non-european regions, nobody will use
it - nobody will participate - it seems that nobody needs these features

So: There _are_ basic features who need to be supported in order that
people from these regions will participate. And this is exactly why I
made the comparison with street names. This is not apples and pears,
this is chicken and egg.

Of course, you can say that OSM is a European project, but then you
should just delete every non-European data and restrict the map to
Europe, so that nobody dares to create maps in their own country.


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


Re: [OSM-talk] Crossroad names

2013-03-24 Thread Frederik Ramm

Hans,

   OSM is a help yourself project. Street names are important in 
Europe, that's why Europeans helped themselves by making a database that 
records, and a map that shows, street names. People in countries where 
crossroads names are very important are invited to help themselves and 
improve OSM to the point where it becomes easy to record and display 
crossroads names.


(OSM is quite popular in Japan so I'm surprised to hear from you that 
OSM should be extremely hard to read there. The Japanese OSM community 
does possess considerable technical skills so I would have assumed that 
if crossroads names are as important as you say, they would have 
developed something to implement them in the mean time.)



1. The entire project is extremely european centric
2. Major features who needed there are not supported (for example
crossroad names)
3. If the map is so unsuitable for non-european regions, nobody will use
it - nobody will participate - it seems that nobody needs these features


This logic doesn't work. OSM has bootstrapped itself from nothing to 
what it is today in Europe. People in Europe participated in OSM before 
the map was usable, and made it usable. People in Japan and Korea can do 
the same.


OSM is not an European project, OSM is a project that gives everyone on 
the planet the chance to participate and make a good map for their area.


To my knowledge there hasn't been a proposal from Japan or Korea about 
how to map or render crossroads names; I'm sure it would be favourably 
considered by the wider community. It would make no sense for someone in 
Europe to develop something that he believes is useful in Japan or Korea 
when he doesn't even know the local circumstances. As I sad, OSM is a 
help-yourself project; people in Japan or Korea are welcome to help 
themselves.


The way things like this often happen is that someone invents some kind 
of hack to achieve what they want - for example, it would be slightly 
incorrect but possible to place a node at a named intersection and tag 
it place=locality, name=blah blah. This would be rendered on the map. 
If it turns out that there's demand for this kind of information and 
people start to add it more frequently, someone would perhaps say uh 
guys, place=locality is not really good for an intersection name, let's 
make up something better, and things would run their course.


Are you in touch with mappers in Japan or Korea, and if so, what is 
their opinion regarding intersection names? Are they waiting for someone 
to tell them what to do, or have they invented some kind of hack to add 
this (according to you) very important information? If they haven't, 
then why not?


Bye
Frederik

--
Frederik Ramm  ##  eMail frede...@remote.org  ##  N49°00'09 E008°23'33

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


Re: [OSM-talk] Crossroad names

2013-03-24 Thread Martin Koppenhoefer
2013/3/24 Frederik Ramm frede...@remote.org:
  OSM is a help yourself project. Street names are important in Europe,
 that's why Europeans helped themselves by making a database that records,
 and a map that shows, street names. People in countries where crossroads
 names are very important are invited to help themselves and improve OSM to
 the point where it becomes easy to record and display crossroads names.

 (OSM is quite popular in Japan so I'm surprised to hear from you that OSM
 should be extremely hard to read there. The Japanese OSM community does
 possess considerable technical skills so I would have assumed that if
 crossroads names are as important as you say, they would have developed
 something to implement them in the mean time.)


But there were times when it was easier for the Europeans to help
themselves, i.e propose a patch for one of the 2 main open map styles
(osmarender and mapnik) and have hope that it would be integrated
(actually in Osmarender you could simply change whatever you felt was
reasonable). AFAIK almost all recently closed mapnik tickets were
marked either invalid, won't fix, worksforme or duplicate, but I
couldn't find hardly anything that actually got integrated. The only 2
tickets I found that were opened since Jan 2012 and got fixed for
mapnik are these two:
https://trac.openstreetmap.org/ticket/4226
https://trac.openstreetmap.org/ticket/4523  (doesn't appear to have
been a style problem)

While I agree that mapnik is (at least for European areas) a
reasonably mature style there are still lots of smaller issues that
might be addressed (and maybe are not because of issues with mapnik on
windows or key not available in db or maybe not so much available time
that the style sheet developers can dedicate). Anyway, for whatever
reason, in the past few years there were very few modifications to
what the style displays and how (while at the same time tagging
evolved a lot).

cheers,
Martin

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


Re: [OSM-talk] Crossroad names

2013-03-24 Thread Hans Schmidt
Am 24.03.2013 14:45, schrieb Frederik Ramm:
 The way things like this often happen is that someone invents some
 kind of hack to achieve what they want - for example, it would be
 slightly incorrect but possible to place a node at a named
 intersection and tag it place=locality, name=blah blah. This would
 be rendered on the map. If it turns out that there's demand for this
 kind of information and people start to add it more frequently,
 someone would perhaps say uh guys, place=locality is not really good
 for an intersection name, let's make up something better, and things
 would run their course.

I can understand that there is a need to first create the data and then
look which is the best. But in some cases, I think it is better to
create a solution and a recommendation beforehand. The thing is that
these features are needed (everyone can see this who takes a look at a
Japanese map), and with some planning in before, this would be easier. A
well developed solution how this is done can enable everyone to tag the
data correctly. If you do a dirty hack, you have to correct everything
afterwards again.


 Are you in touch with mappers in Japan or Korea, and if so, what is
 their opinion regarding intersection names? Are they waiting for
 someone to tell them what to do, or have they invented some kind of
 hack to add this (according to you) very important information? If
 they haven't, then why not?

I have asked the question on the Japanese mailing list, and there the
reply was something like “Yeah, we had the discussion before, and the
general opinion was ‘it would be better to have them displayed’”.

As to why they are not more eager to have them displayed – I don’t know.
Maybe it is a language issue?

Still, there is one problem though: I rather get the feeling that if
this is not supported on the main map, there will be some forks, where
the country specific issues are resolved on a separate page. And I think
this is a problem which is really not desirable. Of course, “OSM is only
a database, not a map”, but osm.org is de facto the place to go if you
want to see the OSM map. If you then have to use tens of different maps
for every country, this is not good.
 
And as Martin said: If a project has gained a certain size, it is harder
to get suggestions implemented. If a project is small, a small issue
will likely be implemented. If it is larger, it is usually considered as
not as important.

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


Re: [OSM-talk] Area tags in Overpass API

2013-03-24 Thread Tobias Knerr
On 23.03.2013 02:13, Paul Norman wrote:
 The canonical list of areas is all polygon entries in
 http://svn.openstreetmap.org/applications/utils/export/osm2pgsql/default.sty
 le
 
 Anything with those tags will be considered an area by the osm2pgsql style
 used for rendering tile.osm.org, which is the largest consumer.

But that list is quite short - surely there are more area keys and tags
documented in the wiki and used in the database than that.

The wiki actually offers machine readable templates on the Key:* and
Tag:* pages with boolean flags such as onWay and onArea, which are
currently parsed by Taginfo and offered though the Taginfo API. Without
doubt there are errors in these wiki templates, but I believe that they
are the most comprehensive source of documented area tags.

Whether it's suitable for Overpass is a different question, of course.
If stable behaviour is a vital goal, that might not be fully achieved
when using a source as potentially volatile as our wiki.

Tobias

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


Re: [OSM-talk] Crossroad names

2013-03-24 Thread Martin Koppenhoefer
2013/3/24 Hans Schmidt z0idb...@gmx.de:
 And as Martin said: If a project has gained a certain size, it is harder
 to get suggestions implemented.


While I'd agree with this, it is not what I wrote. I said: it seems
that mapnik style development has stuck in the recent past (1-2
years), while other fields like tagging have continued to evolve in a
dynamic way as usual.

cheers,
Martin

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


Re: [OSM-talk] Crossroad names

2013-03-24 Thread Cartinus
Have you started working on a patch for the stylesheet yet?

Until you do please stop spamming.

-- 
---
m.v.g.,
Cartinus

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


Re: [OSM-talk] Crossroad names

2013-03-24 Thread Peter Körner

Am 24.03.2013 15:10, schrieb Martin Koppenhoefer:

But there were times when it was easier for the Europeans to help
themselves, i.e propose a patch for one of the 2 main open map styles
(osmarender and mapnik) and have hope that it would be integrated
(actually in Osmarender you could simply change whatever you felt was
reasonable). AFAIK almost all recently closed mapnik tickets were
marked either invalid, won't fix, worksforme or duplicate, but I
couldn't find hardly anything that actually got integrated.


Currently ist's really hard for anyone (not just Europeans) to do any 
change to the mapnik style due tu its complexity. That's why nobody 
really works on it; not because they don't care but because its freaking 
hard to get the style to do exactly what you want without destroying 
sth. else.


Well, someone *is* working on it, he's porting it to Carto to make it 
accessable for the broader public, too:

https://github.com/gravitystorm/openstreetmap-carto

So your best option would be to fork the above mentioned repo, 
incorporate your changes, do some test-renderings of areas in 
Japan/Korea *and* Europe and put all together into a pull request.


The carto style is not yet deployed on osm.org but I'm really sure it 
will land there, soon. The more new features it has over the current 
style, the sooner.


Peter


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


Re: [OSM-talk] Crossroad names

2013-03-24 Thread malenki
Am Sun, 24 Mar 2013 14:12:47 +0100
schrieb Hans Schmidt z0idb...@gmx.de:

 Am 24.03.2013 13:07, schrieb malenki:
  It seems there are not a lot of people in need of this feature.
 About 160 mio people is not a lot of people?

Since you didn't go into details in you OP, where from should I know
this?
What I wanted to say with what I said above is along with Frederik
Ramm's opinion: If there would be/is a real need to render such a map,
somebody would have done/ will do it. :)

  Render a map supporting name= on crossroads. Or make an overlay
  using Overpass API.
 You miss my point: This is an essential map feature. And why should I
 make a map for me personally, if this is a general issue for everyone?

I did not say you should render it for yourself. You can make it
public.

  Apples and pears?
 Is it? It seems that you have just made a comment without _any_
 knowledge and without even the slightest bit of effort to understand
 what this is about.

Like IT people like to say: crap in, crap out.
You say a lot of people need xy, what if we took away ab from
Europe. Luckily in your reply you gave some details (below)
 Sorry, but opinions like yours are ...

You are welcome :)

 [nice explanation why rendered crossroad names are not unimportant]
 [...]
 3. If the map is so unsuitable for non-european regions, nobody will
 use it - nobody will participate - it seems that nobody needs these
 features

In 2006 the map was totally unusable. Despite this fact a lot of
people contributed and made it usable and are going on doing so.

 So: There _are_ basic features who need to be supported in order that
 people from these regions will participate. 

Imho a reason for only a small number of participants in the regions
you mentioned can also be the number of imports dumped into the
database without any cleanup. Just mentioning KSJ2 (tons of
intersecting ways of all kind, thirteen useless tags¹ on lot of nodes),
YahooJapanALPS_Data (lot of unconnected/intersecting highways) or what
user cyana did in South Korea.

 And this is exactly why I made the comparison with street names.
 This is not apples and pears

With your explanations not anymore.

Thomas

¹ http://taginfo.openstreetmap.org/keys/KSJ2%3AINT#combinations



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


Re: [OSM-talk] Crossroad names

2013-03-24 Thread malenki
Am Sun, 24 Mar 2013 15:35:06 +0100
schrieb Hans Schmidt z0idb...@gmx.de:

 Still, there is one problem though: I rather get the feeling that if
 this is not supported on the main map, there will be some forks, where
 the country specific issues are resolved on a separate page. And I
 think this is a problem which is really not desirable.

Why not? E.g. openstreetmap.de has it's own rendering style.

 Of course, “OSM is only a database, not a map”, but osm.org is de
 facto the place to go if you want to see the OSM map. If you then
 have to use tens of different maps for every country, this is not
 good. 

Then why these sites exist (random picks):
openstreetmap.de (own style)
openstreetmap.nl
openstreetmap.jp
openstreetmap.sk
(redirect to awesome freemap.sk)
openstreetmap.cz
openstreetmap.ru
openstreetmap.by (own style)



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


Re: [OSM-talk] Area tags in Overpass API

2013-03-24 Thread Martin Raifer

Am 23.03.2013, 02:13 Uhr, schrieb Paul Norman penor...@mac.com:


The canonical list of areas is all polygon entries in
http://svn.openstreetmap.org/applications/utils/export/osm2pgsql/default.sty
le


I came up with a more elaborate area-detection algorithm for  
overpass-turbo [1], which respects our official documentation (the wiki)  
much better than osm2pgsql does.


It is mostly based on the information found in the wiki and my own  
understanding and interpretation of OSM. It is surely (still) quite  
incomplete. Thus, I would be glad to hear about possible improvements for  
this algorithm!


Martin / tyr_asd

[1] http://wiki.openstreetmap.org/wiki/Overpass_turbo/Polygon_Features

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


Re: [OSM-talk] Area tags in Overpass API

2013-03-24 Thread Martin Raifer

Am 24.03.2013, 15:40 Uhr, schrieb Tobias Knerr o...@tobias-knerr.de:


The wiki actually offers machine readable templates on the Key:* and
Tag:* pages with boolean flags such as onWay and onArea, which are
currently parsed by Taginfo and offered though the Taginfo API. Without
doubt there are errors in these wiki templates, but I believe that they
are the most comprehensive source of documented area tags.


The wiki (and thus also taginfo) has information about where a tag can be  
used on, but not whether the tag defines an area or not. See the following  
examples:


* natural=cliff has onWay=true and onArea=true and defines a way as an  
area, if the way forms a closed loop.
* name=* also has onWay=true and onArea=true, but does not define any way  
as an area.
* highway=* has onArea=false, but if combined with area=yes (or any other  
area-defining tag) the way is still an area.


We would need a flag like definesArea (in addition to onArea) in order  
to write an area-detection algorithm based on the wiki alone.


Martin / tyr_asd

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


Re: [OSM-talk] Area tags in Overpass API

2013-03-24 Thread Tobias Knerr
On 24.03.2013 17:01, Martin Raifer wrote:
 The wiki (and thus also taginfo) has information about where a tag can
 be used on, but not whether the tag defines an area or not. See the
 following examples:
 
 * natural=cliff has onWay=true and onArea=true and defines a way as an
 area, if the way forms a closed loop.
 * name=* also has onWay=true and onArea=true, but does not define any
 way as an area.
 * highway=* has onArea=false, but if combined with area=yes (or any
 other area-defining tag) the way is still an area.

Well, the first example is the one that contradicts my expectations. I
assumed the general rule was that only area-only tags (including
area=yes) can turn a closed way into an area.

I did not know that there were actually features that can be linear, but
where a closed way nevertheless represents an area by default and you
need an area=no if you want it to be linear. If that is accepted
practice, though, then the wiki does indeed not have enough data to make
the distinction.

Tobias

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


Re: [OSM-talk] Crossroad names

2013-03-24 Thread Kevin Peat
On 24 March 2013 15:00, Peter Körner osm-li...@mazdermind.de wrote:

 Currently ist's really hard for anyone (not just Europeans) to do any change
 to the mapnik style due tu its complexity. That's why nobody really works on
 it; not because they don't care but because its freaking hard to get the
 style to do exactly what you want without destroying sth. else.


The current stylesheet has been stuck for ages, even in Europe there
are a lot of useful things unrendered. How about OSMF paying someone
to complete this port? Possibly a corporate sponsor could be found or
have a kickstarter or something.

Kevin

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


Re: [OSM-talk] Crossroad names

2013-03-24 Thread Tom Hughes

On 24/03/13 16:32, Kevin Peat wrote:

On 24 March 2013 15:00, Peter Körner osm-li...@mazdermind.de wrote:


Currently ist's really hard for anyone (not just Europeans) to do any change
to the mapnik style due tu its complexity. That's why nobody really works on
it; not because they don't care but because its freaking hard to get the
style to do exactly what you want without destroying sth. else.


The current stylesheet has been stuck for ages, even in Europe there
are a lot of useful things unrendered. How about OSMF paying someone
to complete this port? Possibly a corporate sponsor could be found or
have a kickstarter or something.


What makes you think money is the problem?

The problem is that the original XML stylesheet is all but impossible to 
edit which is why it is being redeveloped in carto and preparations are 
being made for deployment of that redeveloped version.


Tom

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

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


Re: [OSM-talk] Crossroad names

2013-03-24 Thread Kevin Peat
On 24 March 2013 16:38, Tom Hughes t...@compton.nu wrote:

 What makes you think money is the problem?


Money could help to speed up the process by buying time which people
may not be able to give as a volunteer. Crowd sourcing map data works
great because it is fun (for OSMers at least!) but building
stylesheets is fun for way less people so some inducement may be
required.

Kevin

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


Re: [OSM-talk] Crossroad names

2013-03-24 Thread Hans Schmidt
Am 24.03.2013 16:15, schrieb malenki:
 Since you didn't go into details in you OP, where from should I know
 this?

Well, I mentioned it some weeks before. I didn’t want to write anything
again, because this should have been just a running-up thread for the
bug tracker.

 Why not? E.g. openstreetmap.de has it's own rendering style. 

Because what should be the advantage over one page for everything?

I admit, in order to use different languages, this used to be necessary
(more or less). But with the multilingual map, this problem is hopefully
solved in the future. With different websites, we rather have the
problem that the user has to know every single different page. Imagine
if a German person would like to see the map of Russia? Should he then
first search for the Russian OSM page? Why not just use one map for
everything?


Am 24.03.2013 16:00, schrieb Peter Körner:
 So your best option would be to fork the above mentioned repo,
 incorporate your changes, do some test-renderings of areas in
 Japan/Korea *and* Europe and put all together into a pull request.

About this: Is is not possible to make some styles specific to one
region, so that it would not wreak havoc in another? For example, if we
have crossroad names only in Japan and Korea, would it not be possible
to limit these changes only to these countries? So that the other
countries’ stylesheets are not even affected? Then it would be possible
to “play a little bit around” in one country, without affecting the
entire OSM page.
This would not only be relevant for my issue here. Imagine that in one
country, one type of shop would be so ubiquitous that it needed to be
supported. In another country, that kind of shop is non-existant or not
really relevant.
Or again concerning the crossroads: in Japan, they are usually displayed
with a rectangle around the name. In Korea, this may be different (not
sure about that, though). If we only have one stylesheet for the whole
world, this would inevitably cause problems or create some kind of
substandard “average”, where nobody is really content with.

Of course, this is a software issue, but maybe for some future version:
If these stylesheets could be customized according to region, it would
speed up changes and be more suitable for individual regions. Then the
Japanese mailing list would have the responsibility about “their country”.


By the way, I did not know that the original stylesheet is so impossible
to change. I rather though that it would be relatively easy in a way
like “if node has property x and y, do this“, not even with programming
skills involved. Sorry everyone, this was my mistake.

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


Re: [OSM-talk] Crossroad names

2013-03-24 Thread Hans Schmidt
Am 24.03.2013 17:59, schrieb Kevin Peat:
 On 24 March 2013 16:38, Tom Hughes t...@compton.nu wrote:
 What makes you think money is the problem?

 Money could help to speed up the process by buying time which people
 may not be able to give as a volunteer. Crowd sourcing map data works
 great because it is fun (for OSMers at least!) but building
 stylesheets is fun for way less people so some inducement may be
 required.

This is exactly what I think. Although I happily map with JOSM, I have almost 
no idea about programming or anything more difficult. I rather invest my time 
into something which I am good at, and where I can earn money. I would really 
like to give something of that money to OSM so that other people who are good 
at programming can improve the page for me. Or that somebody might start to do 
some unpleasant work if he is compensated for that dull work with money, so 
that he can do something fun with it. 
Why should I invest so much time a) learning how to program and b) learning to 
understand the stuff behind OSM, if somebody who already is proficient in these 
things can do it so much more efficient than me? 

I wonder why the monetary aspect is not more prominent in open source projects. 
Imo, this could solve many many problems in projects, which are stuck in their 
development due to lack of manpower.
This divison of labour is the core of our economic system and it works very 
well. You pay a hairdresser to get your hair cut, because you do not want to or 
you cannot cut your hair yourself. You pay somebody to have a nice book from a 
foreign language translated, because it would cost you too much time to learn 
the language for yourself. 


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


Re: [OSM-talk] Crossroad names

2013-03-24 Thread Martin Koppenhoefer
2013/3/24 Hans Schmidt z0idb...@gmx.de:
 About this: Is is not possible to make some styles specific to one
 region, so that it would not wreak havoc in another? For example, if we
 have crossroad names only in Japan and Korea, would it not be possible
 to limit these changes only to these countries? So that the other
 countries’ stylesheets are not even affected? Then it would be possible
 to “play a little bit around” in one country, without affecting the
 entire OSM page.


AFAIK with the current implementation on OSM it is not possible, but
if you see the mapquest stack it would be feasible to distinguish by
country. IMHO it is not really necessary to cope on style sheet level
with this: either there is the data, or not, and if crossroads don't
have a name (Europe) nothing will show up even if there was a
rendering rule for named crossroads.


 This would not only be relevant for my issue here. Imagine that in one
 country, one type of shop would be so ubiquitous that it needed to be
 supported. In another country, that kind of shop is non-existant or not
 really relevant.


see above: the style renders data that is there, if a certain shop
type is non-existant in a country it won't be rendered, as it's not
there ;-)


 Or again concerning the crossroads: in Japan, they are usually displayed
 with a rectangle around the name. In Korea, this may be different (not
 sure about that, though). If we only have one stylesheet for the whole
 world, this would inevitably cause problems or create some kind of
 substandard “average”, where nobody is really content with.


this is a case where you would need the feature (e.g. also for custom
subway signs, road colors, highway shields etc. by
country/city/region/etc.)

cheers,
Martin

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


Re: [OSM-talk] Crossroad names

2013-03-24 Thread malenki
Am Sun, 24 Mar 2013 18:15:19 +0100
schrieb Hans Schmidt z0idb...@gmx.de:

 Am 24.03.2013 16:15, schrieb malenki:
  Since you didn't go into details in you OP, where from should I know
  this?
 
 Well, I mentioned it some weeks before. I didn’t want to write
 anything again, because this should have been just a running-up
 thread for the bug tracker.

A link to the post would have done.

 page. Imagine if a German person would like to see the map of Russia?
 Should he then first search for the Russian OSM page? Why not just
 use one map for everything?

If I have a look at Japanese or Korean regions with osm.org I still
wouldn't miss the names of the crossroads :)



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


Re: [OSM-talk] Area tags in Overpass API

2013-03-24 Thread Martin Koppenhoefer
2013/3/24 Tobias Knerr o...@tobias-knerr.de:
 On 24.03.2013 17:01, Martin Raifer wrote:
 * natural=cliff has onWay=true and onArea=true and defines a way as an
 area, if the way forms a closed loop.
 Well, the first example is the one that contradicts my expectations. I
 assumed the general rule was that only area-only tags (including
 area=yes) can turn a closed way into an area.


+1, IMHO this should be discussed and probably we'll have to adjust
the wiki after discussion.

cheers,
Martin

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


Re: [OSM-talk] Crossroad names

2013-03-24 Thread Hans Schmidt
Am 24.03.2013 19:28, schrieb Martin Koppenhoefer:
 this is a case where you would need the feature (e.g. also for custom
 subway signs, road colors, highway shields etc. by
 country/city/region/etc.)

Actually if I think about it, the feature is rather a must-have in some
future version:

Different signs for temples, schools, parking lots, cementaries etc.
Everything where something culturally or linguistical plays a role. I
wouldn’t even dare to guess how much stuff there is which is better to
distinguish from country to country.


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


Re: [OSM-talk] Area tags in Overpass API

2013-03-24 Thread Dave Sutter
There is an example of a line/area from indoor maps. In working in my
real job we model walls either as a line or an area, depending on the
resolution of the building source. A closed way can form an area-type
wall or it can form a line-type wall closed around a room or a
building.

I expect the same will be done in indoor maps in OSM, which I am working on.

Dave

On Sun, Mar 24, 2013 at 9:28 AM, Tobias Knerr o...@tobias-knerr.de wrote:
 On 24.03.2013 17:01, Martin Raifer wrote:
 The wiki (and thus also taginfo) has information about where a tag can
 be used on, but not whether the tag defines an area or not. See the
 following examples:

 * natural=cliff has onWay=true and onArea=true and defines a way as an
 area, if the way forms a closed loop.
 * name=* also has onWay=true and onArea=true, but does not define any
 way as an area.
 * highway=* has onArea=false, but if combined with area=yes (or any
 other area-defining tag) the way is still an area.

 Well, the first example is the one that contradicts my expectations. I
 assumed the general rule was that only area-only tags (including
 area=yes) can turn a closed way into an area.

 I did not know that there were actually features that can be linear, but
 where a closed way nevertheless represents an area by default and you
 need an area=no if you want it to be linear. If that is accepted
 practice, though, then the wiki does indeed not have enough data to make
 the distinction.

 Tobias

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

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


[Talk-de] Wie mappt man einen Hundekotbeutelspender-und-Mülleimer-in-Einem?

2013-03-24 Thread Wuzzy
Wie mappt man etwas, was gleichzeitig ein Hundekotbeutelspender _und_
ein Mülleimer ist?

Folgendes Taggingschema ginge ja leider nicht:
Ein node mit:
 * amenity=waste_basket
 * amenity=vending_machine
 * vending=excrement_bags

Es geht nicht, weil amenity=* ja nur einen Wert haben darf.

Deshalb bin ich vorerst auf diese suboptimale Lösung aufgesprungen:
Ein node mit:
 * amenity=waste_basket
Ein andere node an der gleichen Position mit:
 * amenity=vending_machine
 * vending=excement_bags


Was kann mir Talk-de dazu sagen?

-- 
Wuzzy
XMPP: wuz...@jabber.ccc.de
E-Mail: wuz...@mail.ru

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


Re: [Talk-de] Wie mappt man einen Hundekotbeutelspender-und-Mülleimer-in-Einem?

2013-03-24 Thread Martin Koppenhoefer
Am 24. März 2013 14:33 schrieb Wuzzy wuz...@mail.ru:
 Wie mappt man etwas, was gleichzeitig ein Hundekotbeutelspender _und_
 ein Mülleimer ist?

 Folgendes Taggingschema ginge ja leider nicht:
 Ein node mit:
  * amenity=waste_basket
  * amenity=vending_machine
  * vending=excrement_bags

 Es geht nicht, weil amenity=* ja nur einen Wert haben darf.

 Deshalb bin ich vorerst auf diese suboptimale Lösung aufgesprungen:
 Ein node mit:
  * amenity=waste_basket
 Ein andere node an der gleichen Position mit:
  * amenity=vending_machine
  * vending=excement_bags


 Was kann mir Talk-de dazu sagen?


ja, vermutlich ist das derzeit die beste Möglichkeit (2 nodes),
alternativ könntest Du höchstens einen Doppelwert
amenity=waste_basket;vending_machine taggen, mit dem Ergebnis, dass
höchstwahrscheinlich keiner der Werte mehr ausgewertet wird, daher
wird davon sehr abgeraten.

Gruß Martin

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


Re: [Talk-de] Wie mappt man einen Hundekotbeutelspender-und-Mülleimer-in-Einem?

2013-03-24 Thread Jo
2 Nodes ist hier die beste Lösung.

Jo

2013/3/24 Martin Koppenhoefer dieterdre...@gmail.com

 Am 24. März 2013 14:33 schrieb Wuzzy wuz...@mail.ru:
  Wie mappt man etwas, was gleichzeitig ein Hundekotbeutelspender _und_
  ein Mülleimer ist?
 
  Folgendes Taggingschema ginge ja leider nicht:
  Ein node mit:
   * amenity=waste_basket
   * amenity=vending_machine
   * vending=excrement_bags
 
  Es geht nicht, weil amenity=* ja nur einen Wert haben darf.
 
  Deshalb bin ich vorerst auf diese suboptimale Lösung aufgesprungen:
  Ein node mit:
   * amenity=waste_basket
  Ein andere node an der gleichen Position mit:
   * amenity=vending_machine
   * vending=excement_bags
 
 
  Was kann mir Talk-de dazu sagen?


 ja, vermutlich ist das derzeit die beste Möglichkeit (2 nodes),
 alternativ könntest Du höchstens einen Doppelwert
 amenity=waste_basket;vending_machine taggen, mit dem Ergebnis, dass
 höchstwahrscheinlich keiner der Werte mehr ausgewertet wird, daher
 wird davon sehr abgeraten.

 Gruß Martin

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

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


Re: [Talk-de] Wie Buchstaben-Ergänzung bei Hausnummern mappen?

2013-03-24 Thread Thorsten Alge


On 22.03.2013 13:02, bkmap wrote:
 Hallo,
 
 Dem Nutzer will ich in dieser Sache nur ungern eine Schreibweise
 vorschreiben. Das empfinde ich persönlich als Überregulierung und ist
 meist sowieso nicht durchsetzbar. Außerdem gibt es ja, wie bereits
 erwähnt, regionale Unterschiede ;)
 
 +1
 Ich stimme dem zu. Man soll es jedem selbst überlassen, ob er das Ei am
 breiten oder spitzen Ende aufschlägt ;)

Schön gesagt.

 
 Gruß
 Burkhard
 
 
 
 ___
 Talk-de mailing list
 Talk-de@openstreetmap.org
 http://lists.openstreetmap.org/listinfo/talk-de

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


Re: [Talk-it] Incertezza su autostrada CT-SR

2013-03-24 Thread Aury88
Ciao Luigi Toscano,
se hai dubbi in merito la strada e come essa viene indicata nella
cartellonistica ti consiglio di porre la questione sul sito  skyscrapercity
http://www.skyscrapercity.com/forumdisplay.php?f=169   
in particolare ti consiglio di riferirti al  topic sulla CT-SR
http://www.skyscrapercity.com/showthread.php?t=1042385  
personalmente ho trovato di estrema utilità l'essermi iscritto a questo
forum visto che è una fonte incredibile di informazioni utilissime per
mappare :)

Ps: se decidessi di scrivere sul forum scrivi che stai collaborando a OSM
così fai anche un po' di pubblicità al progetto...visto l'ambito del forum
direi che in moltissimi utenti potrebbero decidere di collaborare alla
nostra amata mappa  ;)




--
View this message in context: 
http://gis.19327.n5.nabble.com/Incertezza-su-autostrada-CT-SR-tp5754419p5754446.html
Sent from the Italy General mailing list archive at Nabble.com.

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


Re: [Talk-it] way avanti e indietro

2013-03-24 Thread Aury88
in realtà potrebbe essere presente una strada che si immette sul raccordo
in ingresso nella rotonda...mettiamo il caso fosse quella la destinazione e
noi provenissimo dalla rotonda, il navigatore potrebbe decidere che, essendo
effettivamente la via più breve e non venendo indicato alcun divieto di
svolta (inversione a U), la rotta da seguire debba essere uscire dalla
rotonda ed effettuare  poi una inversione ad U  che porti sul raccordo di
immissione alla rotonda dove imboccare (a destra) la strada di destinazione.
quindi direi che è una separazione della via + restrizione sarebbe un metodo
di mappatura auspicabile (purtroppo però nel wiki non si fa riferimento a
ciò)

mi è venuto un dubbio: è possibile *usare un bot* per identificare questi
raccordi, dividerli e imporre nel punto di divisione un divieto di svolta?



--
View this message in context: 
http://gis.19327.n5.nabble.com/way-avanti-e-indietro-tp5753797p5754448.html
Sent from the Italy General mailing list archive at Nabble.com.

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


Re: [Talk-it] way avanti e indietro

2013-03-24 Thread Martin Koppenhoefer




Am 24/mar/2013 um 08:34 schrieb Aury88 spacedrive...@gmail.com:

 in realtà potrebbe essere presente una strada che si immette sul raccordo
 in ingresso nella rotonda...mettiamo il caso fosse quella la destinazione e
 noi provenissimo dalla rotonda, il navigatore potrebbe decidere che, essendo
 effettivamente la via più breve e non venendo indicato alcun divieto di
 svolta (inversione a U), la rotta da seguire debba essere uscire dalla
 rotonda ed effettuare  poi una inversione ad U  che porti sul raccordo di
 immissione alla rotonda dove imboccare (a destra) la strada di destinazione.
 quindi direi che è una separazione della via + restrizione sarebbe un metodo
 di mappatura auspicabile


I divieti sono sempre utili da mappare, perché se fai per esempio una 
simulazione del traffico e si intoppa la via più breve devi prevenire all'uso 
di altre possibilità alternative (quali il router normalmente non prende perché 
sono poco più lunghi) se sono vietati per legge.

Lo stesso vale quando l'utente sbaglia strada: devi poi suggerire un 
alternativa legalmente possibile...

Quindi non ci possiamo mai limitare a mappare i divieti di svolta solo per i 
percorsi che pensiamo di essere i più brevi...

Ciao,
Martin
___
Talk-it mailing list
Talk-it@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk-it


Re: [Talk-it] way avanti e indietro

2013-03-24 Thread Martin Koppenhoefer




Am 24/mar/2013 um 08:34 schrieb Aury88 spacedrive...@gmail.com:

 mi è venuto un dubbio: è possibile *usare un bot* per identificare questi
 raccordi, dividerli e imporre nel punto di divisione un divieto di svolta?


no, non è possibile in OSM, dove ci affidiamo delle conoscenze locali dei 
mappatori invece di probabilità h euristiche ;-)

ciao,
Martin
___
Talk-it mailing list
Talk-it@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk-it


[Talk-it] Tag per indicare una villa

2013-03-24 Thread Leonardo

Ciao,

volevo mappare alcune ville della mia zona. Pensavo a questo gruppo di tag:

name = (nome)
building = yes
tourism = attraction
historic = manor

Per l'ultimo tag mi sono basato sulla descrizione della wikia inglese:

http://wiki.openstreetmap.org/wiki/Tag:historic%3Dmanor

Inoltre è abbastanza usato secondo tagwatch:

http://taginfo.openstreetmap.org/tags/?key=historicvalue=manor

Che dite, va bene? Suggerimenti?

Grazie!

Leonardo




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


Re: [Talk-it] Tag per indicare una villa

2013-03-24 Thread bruno
On dom, 2013-03-24 at 13:21 +0100, Leonardo wrote:
 Che dite, va bene? Suggerimenti?

«vai così che è una figata» :-)
/bruno
-- 
CONTACTS http://tracciabi.li/~bruno/contacts.html 2nd email br...@tracciabi.li
GNU/Linux registered user #121507 http://linuxcounter.net


signature.asc
Description: This is a digitally signed message part
___
Talk-it mailing list
Talk-it@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk-it


Re: [Talk-it] Tag per indicare una villa

2013-03-24 Thread Volker Schmidt
L'ho utilizzato in questo modo per ville venete.
tourism=attraction normalmente non ho messo, salvo per le ville più famose
e visitabili.
Ma forse dovrebbe essere utilizzato anche per ville che non si visitano,
perché anche solo dall'esterno possono essere di interesse.
Poi ho aggiunto link to wikipedia e il sito della villa stessa, dove
disponibili.

Non so invece come indicare in generale se una villa o altro posto non è
apèerto per visite al pubblico.

Volker

2013/3/24 Leonardo kinetocor...@gmail.com

 Ciao,

 volevo mappare alcune ville della mia zona. Pensavo a questo gruppo di tag:

 name = (nome)
 building = yes
 tourism = attraction
 historic = manor

 Per l'ultimo tag mi sono basato sulla descrizione della wikia inglese:

 http://wiki.openstreetmap.org/**wiki/Tag:historic%3Dmanorhttp://wiki.openstreetmap.org/wiki/Tag:historic%3Dmanor

 Inoltre è abbastanza usato secondo tagwatch:

 http://taginfo.openstreetmap.**org/tags/?key=historicvalue=**manorhttp://taginfo.openstreetmap.org/tags/?key=historicvalue=manor

 Che dite, va bene? Suggerimenti?

 Grazie!

 Leonardo




 __**_
 Talk-it mailing list
 Talk-it@openstreetmap.org
 http://lists.openstreetmap.**org/listinfo/talk-ithttp://lists.openstreetmap.org/listinfo/talk-it

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


Re: [Talk-it] Tag per indicare una villa

2013-03-24 Thread Martin Koppenhoefer
2013/3/24 Leonardo kinetocor...@gmail.com:
 Ciao,

 volevo mappare alcune ville della mia zona. Pensavo a questo gruppo di tag:

 name = (nome)
 building = yes
 tourism = attraction
 historic = manor

 Per l'ultimo tag mi sono basato sulla descrizione della wikia inglese:
 http://wiki.openstreetmap.org/wiki/Tag:historic%3Dmanor


è discutibile, un manor è simile ad una villa, ma non è proprio la
stessa cosa, farei un pensiero anche all'ipotesi di utilizzare un
nuovo tag come historic=villa
http://taginfo.openstreetmap.org/search?q=historic%3Dvilla (usato solo
8 volte). Forse lo puoi mettere anche su tutto il contorno della
proprietà e non solo ad un singolo edificio.

Per il tag building invece sarei decisamente a favore di usare
building=villa, è chiaramente un tipo architettonico.

In più potresti aggiungere tags come
architect
start_date
wikipedia

ciao,
Martin

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


Re: [Talk-it] Tag per indicare una villa

2013-03-24 Thread Martin Koppenhoefer
2013/3/24 Martin Koppenhoefer dieterdre...@gmail.com:
 In più potresti aggiungere tags come
 architect
 start_date
 wikipedia


e ovviamente ancora tanti altri come
operator
website
addr:housename, addr:housenumber, addr:street
opening_hours
fee
...

ciao,
Martin

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


Re: [Talk-it] Fwd: Dubbio tag name per chiesetta

2013-03-24 Thread Mario Pichetti

Il 21/03/2013 18:53, girarsi ha scritto:

Il 21/03/2013 07:04, Mario Pichetti ha scritto:

Il 18/03/2013 10:03, Simone Saviolo ha scritto:

Chiesa di San Valentino


Per aggiungere o modificare
http://wiki.openstreetmap.org/wiki/IT:How_to_map_a#C

Ciao Mario.


Sinceramente nel link che mi hai postato, non collima con quanto fatto 
da me, nel senso, il tag amenity=place_of_worship l'ho applicato ad 
un'area, non ad un nodo, per cui credo sia ammissibile mettere anche 
area nel suggerimento.

credi giusto:-)


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


Giusta osservazione:-)
Infatti nel link 
http://wiki.openstreetmap.org/wiki/Tag:amenity%3Dplace_of_worship dice:
che si possono usare gli elements node e area, ma non aveva fatto i 
conti con un copia e incolla veloce

da parte del sottoscritto.
Provvedo:-)


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


Re: [Talk-it] Fwd: Dubbio tag name per chiesetta

2013-03-24 Thread Martin Koppenhoefer
2013/3/24 Mario Pichetti mario.piche...@gmail.com:
 Sinceramente nel link che mi hai postato, non collima con quanto fatto da
 me, nel senso, il tag amenity=place_of_worship l'ho applicato ad un'area,
 non ad un nodo, per cui credo sia ammissibile mettere anche area nel
 suggerimento.


nel caso di una chiesa si aggiunge anche building=church all'area
(oltre ad amenity=place_of_worship). Molto spesso non viene fatto, e
il risultato visivo rimane inalterato, ma un luogo di culto non è
necessariamente un edificio.

ciao,
Martin

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


Re: [Talk-it] Tag per indicare una villa

2013-03-24 Thread Volker Schmidt
Non son tanto d'accordo con la creazione di un nuovo  tag building=villa (o
historic=villa, che mi sembrerebbe più logico in analogia con
historic=manor). Se si leggono gli articoli su Manor e Villa on Wikipedia,
viene chiaro che la maggior distintione dei due è in quale paese si
trovano.
L'uso di historic=manor è modesto e di historic=villa nullo, mi sembra.


2013/3/24 Martin Koppenhoefer dieterdre...@gmail.com

 2013/3/24 Martin Koppenhoefer dieterdre...@gmail.com:
  In più potresti aggiungere tags come
  architect
  start_date
  wikipedia


 e ovviamente ancora tanti altri come
 operator
 website
 addr:housename, addr:housenumber, addr:street
 opening_hours
 fee
 ...

 ciao,
 Martin

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

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


Re: [Talk-it] Tag per indicare una villa

2013-03-24 Thread Martin Koppenhoefer
2013/3/24 Volker Schmidt vosc...@gmail.com:
 Non son tanto d'accordo con la creazione di un nuovo  tag building=villa (o
 historic=villa, che mi sembrerebbe più logico in analogia con
 historic=manor). Se si leggono gli articoli su Manor e Villa on Wikipedia,
 viene chiaro che la maggior distintione dei due è in quale paese si trovano.
 L'uso di historic=manor è modesto e di historic=villa nullo, mi sembra.


il tag building=villa non è nuovo, la semantica del tag building è
building=building_type, con valore yes per uso preliminare.
Qualsiasi altro valore più specifico di yes viene gradito,
sopratutto quando si tratta di una tipologia così ben definita come
villa.
Il tag historic=villa invece non è usato molto, concordo. Comunque
faccio notare che wikipedia inglese distingue tra manor e villa,
non li tratta nello stesso articolo.

ciao,
Martin

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


Re: [Talk-it] Tag per indicare una villa

2013-03-24 Thread Leonardo
In realtà è il concetto stesso di villa che si è evoluto nel tempo: 
inizialmente erano edifici padronali circondati da tenute agricole, 
usati quindi per controllare la zona, successivamente sono diventate 
abitazioni di lusso, spesso in campagna. Quindi le ville più antiche 
ricadrebbero idealmente sotto il tag historic=manor mentre quelle più 
nuove potrebbero stare sotto historic=villa. Inoltre il concetto di 
villa è stato esportato anche all'estero (vedi articolo wiki in inglese 
sulle ville in inghilterra). A questo punto sarebbe meglio questa 
combinazione:


building=villa
historic=villa

Pareri?

Leonardo

Il 24/03/2013 15:38, Martin Koppenhoefer ha scritto:

2013/3/24 Volker Schmidt vosc...@gmail.com:

Non son tanto d'accordo con la creazione di un nuovo  tag building=villa (o
historic=villa, che mi sembrerebbe più logico in analogia con
historic=manor). Se si leggono gli articoli su Manor e Villa on Wikipedia,
viene chiaro che la maggior distintione dei due è in quale paese si trovano.
L'uso di historic=manor è modesto e di historic=villa nullo, mi sembra.


il tag building=villa non è nuovo, la semantica del tag building è
building=building_type, con valore yes per uso preliminare.
Qualsiasi altro valore più specifico di yes viene gradito,
sopratutto quando si tratta di una tipologia così ben definita come
villa.
Il tag historic=villa invece non è usato molto, concordo. Comunque
faccio notare che wikipedia inglese distingue tra manor e villa,
non li tratta nello stesso articolo.

ciao,
Martin

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



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


Re: [Talk-it] Fwd: Dubbio tag name per chiesetta

2013-03-24 Thread Mario Pichetti

Il 24/03/2013 15:18, Martin Koppenhoefer ha scritto:

2013/3/24 Mario Pichetti mario.piche...@gmail.com:

Sinceramente nel link che mi hai postato, non collima con quanto fatto da
me, nel senso, il tag amenity=place_of_worship l'ho applicato ad un'area,
non ad un nodo, per cui credo sia ammissibile mettere anche area nel
suggerimento.


nel caso di una chiesa si aggiunge anche building=church all'area
(oltre ad amenity=place_of_worship). Molto spesso non viene fatto, e
il risultato visivo rimane inalterato, ma un luogo di culto non è
necessariamente un edificio.

ciao,
Martin

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


Scusate, ma in fatto di religion, sono neutro:-)
Ho dato una sistematina 
http://wiki.openstreetmap.org/wiki/IT:How_to_map_a#Chiesa.2FChiesetta

Potete segnalare eventuali aggiunte o se volete provvedere personalmente.

Buona serata, Mario.


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


Re: [Talk-it] Fwd: Dubbio tag name per chiesetta

2013-03-24 Thread girarsi_liste

Il 24/03/2013 15:18, Martin Koppenhoefer ha scritto:
nel caso di una chiesa si aggiunge anche building=church all'area 
(oltre ad amenity=place_of_worship).


Avevo dimenticato quel tag in questa chiesetta, grazie di avermelo 
ricordato.



Molto spesso non viene fatto, e il risultato visivo rimane inalterato,
ma un luogo di culto non � necessariamente un edificio. ciao, Martin

In realtà compare una croce all'interno dell'edifici nel caso di un'area 
taggata con building=church






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


Re: [Talk-it] Fwd: Dubbio tag name per chiesetta

2013-03-24 Thread girarsi_liste

Il 24/03/2013 16:21, Mario Pichetti ha scritto:


Scusate, ma in fatto di religion, sono neutro:-)
Ho dato una sistematina 
http://wiki.openstreetmap.org/wiki/IT:How_to_map_a#Chiesa.2FChiesetta

Potete segnalare eventuali aggiunte o se volete provvedere personalmente.

Buona serata, Mario.


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

Direi di inserire anche il simbolo area insieme al simbolo del nodo.

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


Re: [Talk-it] Tag per indicare una villa

2013-03-24 Thread Paolo Monegato

Il 24/03/2013 15:44, Leonardo ha scritto:
In realtà è il concetto stesso di villa che si è evoluto nel tempo: 
inizialmente erano edifici padronali circondati da tenute agricole, 
usati quindi per controllare la zona, successivamente sono diventate 
abitazioni di lusso, spesso in campagna. Quindi le ville più antiche 
ricadrebbero idealmente sotto il tag historic=manor mentre quelle più 
nuove potrebbero stare sotto historic=villa. Inoltre il concetto di 
villa è stato esportato anche all'estero (vedi articolo wiki in 
inglese sulle ville in inghilterra). A questo punto sarebbe meglio 
questa combinazione:


building=villa
historic=villa

Pareri?


C'era già stata (molto) tempo fa una discussione [1]... vediamo se sta 
volta riusciamo a definire un tag. All'epoca avevo proposto proprio 
historic=villa (o forse addirittura historic=villa veneta, per 
distinguerla da altri tipi di villa che hanno uno scopo ed una 
conformazione differente), dunque non posso che appoggiare la proposta.


Il building=villa lo metteresti solo sulla casa dominicale o anche 
sulle barchesse?


ciao
Paolo M

ps: historic=manor da quanto capisco dal wiki è in conflitto con 
historic=castle + castle_type=manor (e pure castle_type=stately volendo)


[1] http://lists.openstreetmap.org/pipermail/talk-it/2009-June/008818.html

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


Re: [Talk-it] Tag per indicare una villa

2013-03-24 Thread Gian Mario Navillod
Se la villa veneta è un tipo architettonico (
http://en.wikipedia.org/wiki/Palladian_villas_of_Veneto) forse sarebbe più
chiaro un tag più specifico del tipo

building=veneto_villa

anche se su taginfo ci sono già 20
building=villa

Buon mapping.

Gian Mario.
___
Talk-it mailing list
Talk-it@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk-it


Re: [Talk-it] Fwd: Dubbio tag name per chiesetta

2013-03-24 Thread Martin Koppenhoefer
2013/3/24 girarsi_liste liste.gira...@gmail.com:
 Il 24/03/2013 15:18, Martin Koppenhoefer ha scritto:
nel caso di una chiesa si aggiunge anche building=church all'area (oltre ad
 amenity=place_of_worship).

 Avevo dimenticato quel tag in questa chiesetta, grazie di avermelo
 ricordato.



Molto spesso non viene fatto, e il risultato visivo rimane inalterato,
ma un luogo di culto non � necessariamente un edificio. ciao, Martin

 In realtà compare una croce all'interno dell'edifici nel caso di un'area
 taggata con building=church


quello sopratutto viene gestito dal tag religion=christian.
Al solito si aggiunge anche la denomination=catholic (o roman_catholic)
http://taginfo.openstreetmap.org/keys/denomination#values

ciao,
Martin

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


Re: [Talk-it] Tag per indicare una villa

2013-03-24 Thread Martin Koppenhoefer
2013/3/24 Gian Mario Navillod gian.mario.navil...@gmail.com:
 Se la villa veneta è un tipo architettonico
 (http://en.wikipedia.org/wiki/Palladian_villas_of_Veneto) forse sarebbe più
 chiaro un tag più specifico del tipo

 building=veneto_villa

 anche se su taginfo ci sono già 20
 building=villa


per me building=venetian_villa o simile sarebbe troppo specifico come
archetipo di edificio (credo, ammetto che non sono ne esperto di
Palladio ne del Veneto). Ammetto che non ho ne anche letto tutto
l'articolo inglese sulle ville veneziane del Palladio, potresti
indicare quale paragrafo ti fa pensare che un tipo proprio sarebbe più
opportuno?

ciao,
Martin

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


Re: [Talk-it] Consiglio tag shop

2013-03-24 Thread Martin Koppenhoefer
2013/3/24 Gianluca Boero gianlucabo...@alice.it:
 come sempre da uno scontrino che ho in mano vorrei fare la mappatura di un
 negozio.
 E' un negozio di scarpe però come molti ora vende anche abbigliamento.

 quindi shop=shoes.


+1, se si tratta di un negozio di scarpe è questo il tag ;-)


 Abbigliamento lo posso inserire come clothes=yes ? (non potendo abbinare due
 tag shop)


potresti usare vending:clothes=yes oppure sell:clothes=yes perché
altrimenti non è molto chiaro ne anche per un essere umano (potrebbe
anche significare che puoi essere vestito per entrare ecc.)

ciao,
Martin

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


Re: [Talk-se] Postnummer (igen)

2013-03-24 Thread Karl Wettin

http://osm.kodapan.se/postnummer

Nu med Voronoi där regioner med angränsande punkter i samma postnummer slås 
ihop till en polygon. Det senare buggar lite och skapar exempelvis lite 
oväntade streck här och där.

OSM-filer att hälla in i JSOM för varje postnummerområde (varje postnummer kan 
alltså bestå av flera stycken områden) finns för Halmstad (som i princip bygger 
på noderna av  vägar som finns i ett enda postnummer) och Lund (där nästan 
varje hus i centrala delarna har ett nummer och skapar en känd punkt för ett 
postnummer) här:

http://osm.kodapan.se/postnummer/lund.tar.gz
http://osm.kodapan.se/postnummer/halmstad.tar.gz

Lade även upp ett par screenshots från JOSM om man inte orkar:

http://osm.kodapan.se/postnummer/lund%20kommun.tiff
http://osm.kodapan.se/postnummer/lund.tiff
http://osm.kodapan.se/postnummer/lund%20centralt.tiff

http://osm.kodapan.se/postnummer/halmstad%20kommun.tiff
http://osm.kodapan.se/postnummer/halmstad.tiff
http://osm.kodapan.se/postnummer/halmstad%20centralt.tiff

Det kräver fortfarande ytterligare några helgers arbete innan det här är klart.

Om någon känner sig manad att kolla hur bra det fungerar i ett område de känner 
till är det bara att säga till.


Applikationen äter ruskigt mycket minne och är ganska långsam om man inte 
cachar alla anrop till Posten, vilket man naturligtvis inte vill göra i det 
långa loppet.  Vill någon leka med java-projektet (som också är en soppa) är 
det bara att höra av sig.

Det kommer bli en baggis att extrahera postortsgränser från det här. Portorter 
är inte samma sak som SCBs småorter och tätorter, men jag tycker nog att en 
postortspolygon i många fall är mycket bättre än en enda nod som pekar ut 
centroiden av en ort.

Ett stort problem är att jag inte vet till vilken postort en given punkt hör. 
Jag måste söka upp de närmsta städerna och slå upp om gatan är unik för de 
närmsta städerna. Är det inte unikt (vilket det ofta inte är) kan jag inte 
använda mig av datan. Man kan förvisso efter ett par iterationen använda sig av 
de framgruvade postortspolygonerna till detta, men jag oroar mig för att detta 
kommer skapa för mycket feedback med grus i systemet. En lösning är att komma 
över en öppen karta som visar två- eller tresiffrors postnummer. Annars vet jag 
inte riktigt. Om någon har förslag är det mycket välkommet.


Slutmålet är som tidigare nämnt att skriva en öppen svensk geocoder som 
inter/extrapolerar fram husnummer (Posten listar alla nummer som finns), fixar 
med svenska regler för hur man kan tänkas skriva fel, etc . Detta ligger dock 
långt fram (flera år?) om det över huvud taget blir av. Det stora problemet är 
att gruva fram all saknad data... Det borde dock gå att få till när man 
kombinerar Postens data med OSM. Det krävs även en hel del arbete med att 
extrahera data från Postgis så det bildar en hierarkisk struktur. Nominatim gör 
det, men det är en soppa av olika språk och verktyg och jag lyckas inte tyda 
vad som händer. Om någon har tankar kring det kan jag ganska omgående skriva en 
geocoder fungerar kanon i Kalmar och Lund, men det känns relativt meningslöst 
att lägga tid på det nu.

Har också kommit fram till att om något av detta (postnummerpolygonerna, 
postortspolygoner, gissa position för husnummer, etc)  någonsin blir klart 
kommer jag nog sätta upp en ny Postgis med Overpass API som enbart innehåller 
dessa data så det blir lättare att rendera om allt från början lite då och då. 
Slipper folk som ritar in saker i noderna, osv. Dessutom blir det jag 
personligen och inte OSM som är data-martyren om någon (läs: Posten, PTS, etc) 
skulle bli väldigt arga. Kanadensiska posten blev ju väldigt upprörda.




kalle


3 jan 2013 kl. 20:06 skrev Karl Wettin:

 Jag har sytt ihop en snurra som fixar postnummer via postens webtjänst 
 http://www.posten.se/soktjanst/postnummersok/.
 
 Den kan redan nu lägga till och rätta postnummer på postadressrelaterade 
 saker, tycker ni jag skall släppa loss den för att göra det? Skall slänga upp 
 koden på github eller något sånt.
 
 Jag hade dock tänkt mig att den skall skapa områdespolygoner vilket tyvärr 
 inte fungerar så bra ännu.
 
 Först samlar den upp massa punkter den kan koppla till ett postnummer:
 
 1. Nodes och ways med [addr:street=*] och [addr:housenumber=*].
 2. Ways med [highway=*] och [name=*] där  hela gatan befinner sig i samma 
 postnummer.
 
 Saknas [addr:city] väljer den det på följande sätt:
 
 1. Hämtar alla [place={hamlet, village, town, city}] med [ref:se:scb=*].
 2. Väljer ut de som har [name] som existerar som postort hos Posten.
 3. Sorterar till avståndet från positionen till alla postorter.
 4. Söker efter den första förekommande husnummer med rätt gatunamn bland 
 listan av möjliga postorter. (Bör nog göras om så den väljer bara om det 
 finns i en av postorterna)
 
 För att skapa polygoner har jag testkört lite över Lund (centrala) och 
 Halmstad (kommunen).
 
 Mitt första försök är att hälla alla punkter per 

Re: [Talk-es] LearnOSM em español

2013-03-24 Thread shiguera

Hola:
A mí me gustaría colaborar, pero no tengo muy claro como hacerlo.
He visto el repositorio de learnosm/valencia

https://github.com/geoinquietosvlc/learnosm

Pero a partir de ahí no se como hacer.
¿Alguna pista?
Gracias

Fdo.: Santiago Higuera

l sáb, 23-03-2013 a las 19:16 +0100, Jorge Gaspar Sanz Salinas escribió:
 2013/3/23 Jorge Gaspar Sanz Salinas js...@osgeo.org:
  2013/3/22 Alex Barth a...@mapbox.com:
  Acabodo de crear un branch (rama?) 'spanish' y un ticket para coordenar la
  traducción:
 
  https://github.com/hotosm/learnosm/issues/63
 
  Genial Alex, a ver si alguien se anima. Si no lo hace nadie antes, tal
  vez haga el fork en la cuenta de GeoinquietosVLC y desde ahí
  coordinamos la traducción para ir enviando pull requests.
 
  Por cierto, no sé si el material del taller que preparé el año
  pasado[1] puede ser útil. Dentro de un mes haremos otro taller aquí en
  Valencia y seguramente lo revisemos y nos apoyemos en materiales de
  LearnOSM.
 
  Saludos!
 
  [1] http://jsanz.github.com/slides-201205-osm-xirivella/
 
 
 
 Bueno ya lo he comentado en el ticket, pero si alguien se anima a
 echar una mano (una persona ya me ha contactado fuera de lista), que
 se pase por el ticket y nos coordinamos allí.
 
 Si no tiene ni la más remota idea de cómo usar GitHub, pues dificulta
 un poco el asunto pero lo podemos resolver entonces por aquí.
 
 Saludos
 




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


Re: [Talk-ro] Vandalizare Onesti, Bacau

2013-03-24 Thread Gabriel Sebastian Moise
Fratilor, va repar eu Onesti, cu tot cu cladiri, insa am o rugaminte mare de
a nu accesa nimeni Onestiul pentru modificari pana Marti Seara  pe data de
26.03.2013.

  



--
View this message in context: 
http://gis.19327.n5.nabble.com/Vandalizare-Onesti-Bacau-tp5749344p5754544.html
Sent from the Romania mailing list archive at Nabble.com.

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


Re: [Talk-ro] Vandalizare Onesti, Bacau

2013-03-24 Thread Strainu
Poate trebuie sa salvezi mai des, gen la 5-10 min si sa reintri in
editor? E mai lent, dar o sa vezi repede ce fac altii. Oricum, stai
linistit. Am desenat acu o saptamana vreo cateva strazi si nu se
atinsese nimeni de Onesti de cand a inceput threadul asta.

Strainu

În data de 24 martie 2013, 22:34, Gabriel Sebastian Moise
gabrielsebastianmo...@gmail.com a scris:
 Eddy, ca sa nu cumva sa ma trezesc ca mai reface vreunu si , punem de 2 ori
 un drum.  Am patit azi-noapte in Ploiesti Prahova .   Deja partea de nord
 vest e 90 % gata pana in strada principala , cu tot cu nume , cu nr de
 lane-uri, limita de viteza, tip de drum etc.



 --
 View this message in context: 
 http://gis.19327.n5.nabble.com/Vandalizare-Onesti-Bacau-tp5749344p5754553.html
 Sent from the Romania mailing list archive at Nabble.com.

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

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


Re: [Talk-ro] Vandalizare Onesti, Bacau

2013-03-24 Thread Gabriel Sebastian Moise
Eddy,  tot ce fac eu este by-hand, dupa imaginile din satelit. Denumirile le
iau de oriunde pot si vreau. 



--
View this message in context: 
http://gis.19327.n5.nabble.com/Vandalizare-Onesti-Bacau-tp5749344p5754561.html
Sent from the Romania mailing list archive at Nabble.com.

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


Re: [Talk-ro] Vandalizare Onesti, Bacau

2013-03-24 Thread Razvan
Dar nu are nimeni idee cam cine ar fi a vut creierul atat de ars incat 
sa faca asa ceva?

On 24.03.2013 22:40, Gabriel Sebastian Moise wrote:

Stiu, nu-mi fac problema de asta, dar imi place ca treaba sa fie facuta bine.
Adineauri am refacut si sensul giratoriu de pe principala care, era ceva ,
ce nu fapt nu era nimic.



--
View this message in context: 
http://gis.19327.n5.nabble.com/Vandalizare-Onesti-Bacau-tp5749344p5754560.html
Sent from the Romania mailing list archive at Nabble.com.

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



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


Re: [Talk-ro] O modalitate mai proasta de a te inregistra la acest asazis forum nici ca se putea. Facuta CU PICIOARELE

2013-03-24 Thread razor74
Pai tocmai ca am scris asazis forum, pt ca arata ca un forum pagina de
discutii...




--
View this message in context: 
http://gis.19327.n5.nabble.com/O-modalitate-mai-proasta-de-a-te-inregistra-la-acest-asazis-forum-nici-ca-se-putea-Facuta-CU-PICIOARE-tp5754508p5754570.html
Sent from the Romania mailing list archive at Nabble.com.

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


Re: [OSM-Talk-ZA] Lepelle, Motlatse, etc.

2013-03-24 Thread Gerhardus Geldenhuis
Hi
I think the alt name should be the old name. If we do it the other way
around we risk getting involved in politics. Having the old names as alt
names would however be important. More names means it gives flexibility to
have differently rendered maps. You could probably even go as far as having
the old province borders imported as historical data, not that it would be
of much use but might be interesting for historical purposes.

Regards


On 24 March 2013 10:22, David Schneider d.schnei...@tradermail.info wrote:

 Hi List!

 The OSM data still has the Olifants River, Blyde River, etc. listed under
 their old names.
 According to various sources, the official name changed in 2005 to Lepelle.

 http://www.krugerpark.co.za/krugerpark-times-2-15-olifants-river-lepelle-20557.html
 So did the names of a number of other places.

 Have these name changes never been adopted by the ZA public, and are
 therefore
 merely an act of the bureaucracy? I think, both names should be in our
 data.
 Either as name: Lepelle, old_name: Olifants River, or name: Olifants River,
 alt_name: Lepelle. What do you think?

 Best,
 David


 ___
 Talk-ZA mailing list
 Talk-ZA@openstreetmap.org
 http://lists.openstreetmap.org/listinfo/talk-za




-- 
Gerhardus Geldenhuis
___
Talk-ZA mailing list
Talk-ZA@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk-za


Re: [OSM-Talk-ZA] Lepelle, Motlatse, etc.

2013-03-24 Thread Glen Wilson
Before going too far into this, take a look at the name key 
http://wiki.openstreetmap.org/wiki/Name on the OSM wiki.  Also take a 
look at the on-the-ground rule 
http://wiki.openstreetmap.org/wiki/Disputes#On_the_Ground_Rule.  OSM 
records what is - there's little scope for what was and will be.


Basically name=* is what displays on most of the renderers etc. If most 
(local) people and (more importantly) road signs know it as the Blyde 
River then that's what should be set in the name=* tag.  In such a case 
official_name=Motlatse River is useful.


alt_name=* would probably apply if it were also called Blyde Canyon River.

Of course if the local government's got lots of money for sign changes 
(and the people want out with the old) then go ahead and use 
name=Motlatse River; old_name=Blyde River.


Language variations (name:xx=*) might be useful in cases where the 
official name is really just the same thing in a different language 
(which seems to be the case for Levuvhu vs Levubu, but I'm no language 
expert).  In these cases it's still a good idea to use 
name/official_name/old_name etc.


This naming stuff also applies to names of other things like places, 
parks, and streets.


Include the suffix (River/Stream/Spruit etc) in the name if known.  
It's quite possible to have Blyde River tagged as waterway=stream 
(waterway=* is a physical/functional tag and would be relevant even if 
it was called Motlatse Street).


Later

Glen

On 24-Mar-13 16:38, Gerhardus Geldenhuis wrote:

Hi
I think the alt name should be the old name. If we do it the other way 
around we risk getting involved in politics. Having the old names as 
alt names would however be important. More names means it gives 
flexibility to have differently rendered maps. You could probably even 
go as far as having the old province borders imported as historical 
data, not that it would be of much use but might be interesting for 
historical purposes.


Regards


On 24 March 2013 10:22, David Schneider d.schnei...@tradermail.info 
mailto:d.schnei...@tradermail.info wrote:


Hi List!

The OSM data still has the Olifants River, Blyde River, etc.
listed under
their old names.
According to various sources, the official name changed in 2005 to
Lepelle.

http://www.krugerpark.co.za/krugerpark-times-2-15-olifants-river-lepelle-20557.html
So did the names of a number of other places.

Have these name changes never been adopted by the ZA public, and
are therefore
merely an act of the bureaucracy? I think, both names should be in
our data.
Either as name: Lepelle, old_name: Olifants River, or name:
Olifants River,
alt_name: Lepelle. What do you think?

Best,
David


___
Talk-ZA mailing list
Talk-ZA@openstreetmap.org mailto:Talk-ZA@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk-za




--
Gerhardus Geldenhuis


___
Talk-ZA mailing list
Talk-ZA@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk-za


___
Talk-ZA mailing list
Talk-ZA@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk-za


Re: [Talk-cz] Konference GIS ve veřejné správě

2013-03-24 Thread Jachym Cepicky
Pro normální lidi ano - vstupné je 2500,-

předpokládám, že ve vložném je i nějaké to občerstvení (jak tak na
konferencích bývá)

členové CAGI mají drobnou slevu,

studenti za 600,-

Jáchym

Dne 22.3.2013 10:00, karpi.li...@email.cz napsal(a):
 
 ..no já jsem se do toho díval, letmo.
 ..a to Vložné 2500,- resp. 2900,-
 ...to znamená jako vstupné?
 
 
 
 
 -- Původní zpráva --
 Od: Jachym Cepicky jachym.cepi...@gmail.com
 Datum: 27. 2. 2013
 Předmět: [Talk-cz] Konference GIS ve veřejné správě
 
 Přátelé,
 
 rád bych vás pozval na šestý ročník konference, pořádané Českou asociací
 pro geoinformace (CAGI) [1] - GIS ve veřejné správě [2].
 
 Datum a místo: 27. – 28. 5. 2013 Novotného lávka 5, Praha
 
 Na GIS Ostrava jsme se shodli, že v rámci CAGI vznikne
 nová pracovní skupina pro Open Source a Open (Geo)Data. Jedna sekce v
 rámci konference je věnována právě této problematice (problematikám).
 Očekávám tak bohatou aktivní účast. Většina z nás má nějaké vazby na
 veřejnou správu, znáte problémy, na které narážíme při návrhu
 obsahujícím Open Source software. Opakují se některé problémy
 systematicky? Můžeme je odstranit? Nebo stačí jen stále vyvracet některé
 mýty?
 
 CUZK v poslední době provádí kroky směřující k otevřeným datům. Bude
 zajímavé slyšet, jak daleko se dostali v novinkách představených na GIS
 Ostrava. Na jejich službách staví hodně Open Street Map. V CUZK si
 určitě rádi poslechnou názory a připomínky ke svým službám od jejich
 uživatelů.
 
 První cirkulář najdete na [3].
 
 Přihlášky můžete posílat buď mě nebo se registrovat rovnou na
 givs2013-autor at cagi.cz
 
 Jáchym Čepický
 
 [1] http://cagi.cz
 [2] http://cagi.cz/akcevypis.php?id=519
 [3] http://cagi.cz/files/GIVS2013-Cirkular-01_01021391738.pdf
 
 
 
 ___
 Talk-cz mailing list
 Talk-cz@openstreetmap.org
 http://lists.openstreetmap.org/listinfo/talk-cz
 

-- 
Jachym Cepicky
Help Service - Remote Sensing s.r.o.
jachym.cepi...@gmail.com
HS-RS: jac...@hsrs.cz http://bnhelp.cz
http://les-ejk.cz



signature.asc
Description: OpenPGP digital signature
___
Talk-cz mailing list
Talk-cz@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk-cz


Re: [OSM-talk-fr] Pont de Ponte-Novu détruit

2013-03-24 Thread Philippe Verdy
Le 23 mars 2013 21:04, Francescu GAROBY windu...@gmail.com a écrit :
 Les piétons ne peuvent même pas aller jusqu'au au bout du morceau de pont
 restant : une barrière de sécurité en limite l'accès.

Ce qui est faux si on voit les photos. Il y a une première travée de
chaque côté (sur une arche), cette travée n'est pas fermée mais
interdite aux véhicules par des plots en béton. Au bout de cette
travée il y a une barrière. Ensuite il y a le tro où il manque deux
arches et où il reste un pilier.

On ne voit pas la même chose sur les photos. Ou alors c'est qu'il y a
eu une destruction très récente des travées qui restaient...

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


Re: [OSM-talk-fr] Pont de Ponte-Novu détruit

2013-03-24 Thread Ista Pouss
Il faudrait tenir compte de ce que certains personnages de bandes dessinées
peuvent marcher dans le vide tant qu'ils croient être sur le sol. Aussi je
suggère :

comics_access=true



Le 24 mars 2013 09:31, Philippe Verdy verd...@wanadoo.fr a écrit :

 Le 23 mars 2013 21:04, Francescu GAROBY windu...@gmail.com a écrit :
  Les piétons ne peuvent même pas aller jusqu'au au bout du morceau de pont
  restant : une barrière de sécurité en limite l'accès.

 Ce qui est faux si on voit les photos. Il y a une première travée de
 chaque côté (sur une arche), cette travée n'est pas fermée mais
 interdite aux véhicules par des plots en béton. Au bout de cette
 travée il y a une barrière. Ensuite il y a le tro où il manque deux
 arches et où il reste un pilier.

 On ne voit pas la même chose sur les photos. Ou alors c'est qu'il y a
 eu une destruction très récente des travées qui restaient...

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




-- 
Les dérives de rue :
Des textes à promenades http://drivrsdu.fr/des-textes-a-promenades/
___
Talk-fr mailing list
Talk-fr@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk-fr


Re: [OSM-talk-fr] Pont de Ponte-Novu détruit

2013-03-24 Thread Francescu GAROBY
Non, ce que je dis n'est pas faux : la barrière n'est pas en bout du
morceau de pont ! Elle est en bout de la partie praticable, mais derrière,
il reste encore 1 bon mètre, en plus mauvais état.
Cf. cette autre photo de Google Street
Viewhttps://maps.google.com/maps?q=ponte-novuhl=frll=42.485818,9.281731spn=0.006599,0.014312sll=49.184628,-0.372259sspn=0.093576,0.228996hnear=Ponte+Novu,+Castello-di-Rostino,+Haute-Corse,+Corse,+Francet=hlayer=ccbll=42.486226,9.281073panoid=RXbtP2lor7ddBZvFCM_IEwcbp=12,272.62,,0,10.87z=17,
où le pont est plus de profil, et où on voit bien un morceau (légèrement en
contrebas), après la barrière et le murets de sécurité.

Francescu


Le 24 mars 2013 09:31, Philippe Verdy verd...@wanadoo.fr a écrit :

 Le 23 mars 2013 21:04, Francescu GAROBY windu...@gmail.com a écrit :
  Les piétons ne peuvent même pas aller jusqu'au au bout du morceau de pont
  restant : une barrière de sécurité en limite l'accès.

 Ce qui est faux si on voit les photos. Il y a une première travée de
 chaque côté (sur une arche), cette travée n'est pas fermée mais
 interdite aux véhicules par des plots en béton. Au bout de cette
 travée il y a une barrière. Ensuite il y a le tro où il manque deux
 arches et où il reste un pilier.

 On ne voit pas la même chose sur les photos. Ou alors c'est qu'il y a
 eu une destruction très récente des travées qui restaient...




-- 
Cordialement,
Francescu GAROBY
___
Talk-fr mailing list
Talk-fr@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk-fr


Re: [OSM-talk-fr] Version française du projet LearnOSM ?

2013-03-24 Thread Pieren
2013/3/23 Yohan Boniface yohanbonif...@free.fr:
 J'ai aussi commencé un wiki avec les conventions de la traduction française,
 histoire d'avoir une traduction cohérente:
 https://github.com/hotosm/learnosm/wiki/FrenchTranslation . N'hésitez pas à
 éditer/compléter/corriger.

In order to keep all the translations persistent, here are some translations 
of common OSM specific
words. Feel free to suggest others words or better translations.
email = mail
happy mapping = ?
mailing list = liste diffusion
mapper = mappeur

Il y a des termes qu'on n'aime pas forcément mais qu'on devrait se
forcer à utiliser dans une traduction officielle:
email = courriel
happy mapping = bonne cartographie ?
mailing list = liste de diffusion
mapper = cartographe (mappeur me fait peur ;)

Pieren

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


Re: [OSM-talk-fr] Rendu des terrains de sport

2013-03-24 Thread Ista Pouss
Bonjour,

Je me suis un peu renseigné à propos d'un terrain de foot juste à coté de
chez moi et dont j'avais parlé à cause d'un rendu mal foutu pour le rendu
fr :

http://tile.openstreetmap.fr/?zoom=19lat=45.4207lon=4.4232layers=B0


Il apparaît que, s'il se joue bien des petites compétitions amateurs sur ce
terrain, ses dimensions sont complètement folkloriques, et notamment il est
beaucoup trop petit.

Il parait que la fédération française de foot homologue les terrains selon
les niveaux de compétition ; malheureusement ces infos sont introuvables
sur leur site (enfin... je ne les ai pas trouvées)... À quand le mouvement
open data dans le sport ?

Donc, si ce terrain mérite bien le terme de terrain de football (il s'y
joue au moins 30 compétitions par week-end ! ), il est vain de vouloir y
copier/coller un svg standard de terrain de football. Pour ce genre de cas
un logo serait je pense plus adapté.

Je ne sais pas comment tout cela pourrait se traduire en tags ?




Le 22 mars 2013 19:40, Philippe Verdy verd...@wanadoo.fr a écrit :

 Un gris (neutre vu que ce n'est pas une surface naturelle comme
 l'herbe ou la terre battue) peut-être un peu bleuté pour faire encore
 plus artificiel (les nouvelles surfaces complètement artificielles de
 compétition intérieures sont carrément d'un bleu bien prononcé pour
 mieux passer à la TV, mais je n'irais pas jusque là pour les surfaces
 exérieures, là ce serait braiment trop flashy).

 Le 22 mars 2013 16:44, Christian Quest cqu...@openstreetmap.fr a écrit :
  Le 22 mars 2013 14:37, Otourly Wiki otou...@yahoo.fr a écrit :
  Le tag surface pourrait-être pris en compte pour les stades de foot en
 gore
  ?
 
  Ouaille note.
 
  J'ai modifié la couleur pour surface=grass (y compris pour les tennis
  c'était trop flashy)
  Par contre, pour l'artificiel, je ne vois pas trop quoi mettre.

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




-- 
Les dérives de rue :
Des textes à promenades http://drivrsdu.fr/des-textes-a-promenades/
___
Talk-fr mailing list
Talk-fr@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk-fr


Re: [OSM-talk-fr] Rendu des terrains de sport

2013-03-24 Thread Bruno Besson
Petite aberration de rendu :
http://tile.openstreetmap.fr/?zoom=16lat=45.31181lon=5.87866layers=B0(terrain
de tennis géant uniquement au zoom 16)


2013/3/24 Ista Pouss ista...@gmail.com

 Bonjour,

 Je me suis un peu renseigné à propos d'un terrain de foot juste à coté de
 chez moi et dont j'avais parlé à cause d'un rendu mal foutu pour le rendu
 fr :

 http://tile.openstreetmap.fr/?zoom=19lat=45.4207lon=4.4232layers=B0


 Il apparaît que, s'il se joue bien des petites compétitions amateurs sur
 ce terrain, ses dimensions sont complètement folkloriques, et notamment il
 est beaucoup trop petit.

 Il parait que la fédération française de foot homologue les terrains selon
 les niveaux de compétition ; malheureusement ces infos sont introuvables
 sur leur site (enfin... je ne les ai pas trouvées)... À quand le mouvement
 open data dans le sport ?

 Donc, si ce terrain mérite bien le terme de terrain de football (il s'y
 joue au moins 30 compétitions par week-end ! ), il est vain de vouloir y
 copier/coller un svg standard de terrain de football. Pour ce genre de cas
 un logo serait je pense plus adapté.

 Je ne sais pas comment tout cela pourrait se traduire en tags ?




 Le 22 mars 2013 19:40, Philippe Verdy verd...@wanadoo.fr a écrit :

 Un gris (neutre vu que ce n'est pas une surface naturelle comme
 l'herbe ou la terre battue) peut-être un peu bleuté pour faire encore
 plus artificiel (les nouvelles surfaces complètement artificielles de
 compétition intérieures sont carrément d'un bleu bien prononcé pour
 mieux passer à la TV, mais je n'irais pas jusque là pour les surfaces
 exérieures, là ce serait braiment trop flashy).

 Le 22 mars 2013 16:44, Christian Quest cqu...@openstreetmap.fr a écrit
 :
  Le 22 mars 2013 14:37, Otourly Wiki otou...@yahoo.fr a écrit :
  Le tag surface pourrait-être pris en compte pour les stades de foot en
 gore
  ?
 
  Ouaille note.
 
  J'ai modifié la couleur pour surface=grass (y compris pour les tennis
  c'était trop flashy)
  Par contre, pour l'artificiel, je ne vois pas trop quoi mettre.

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




 --
 Les dérives de rue :
 Des textes à promenades http://drivrsdu.fr/des-textes-a-promenades/

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


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


Re: [OSM-talk-fr] Demande d'assistance cartographique sur le Cyclone Haruna à Madagascar

2013-03-24 Thread Eric SIBERT

Bonjour Cédric,

Merci pour ce retour.

 Le Bureau national de gestion des risques et des catastrophes, des ONG

et des habitants de Toliara nous ont félicité pour ce travail.


S'il était possible d'avoir des contacts directement à l'intérieur du 
Bureau national de gestion des risques et des catastrophes, ça serait 
très intéressant pour OSM en général et pour la gestion des crises en 
particulier: savoir où est une commune, un district...


Pour préciser, on trouve sur le site de l'OCHA les limites de toutes les 
communes de Madagascar. Pour des questions de licence, on ne peut pas 
intégrer les données de l'OCHA dans OSM. Sauf que les données en 
question viennent des services ministériels de Madagascar. Les 
tentatives précédentes de contact n'ont pas été fructueuses. Tout 
nouveau contact, serait le bienvenue.




Plus globalement, dans le cadre d'interventions d'urgence, il est
nécessaire de :
- donner une méthodologie simple aux acteurs de l'urgence pour compléter
les cartes que nous élaborons sans que leur mission de secours en soit
perturbée,
- intégrer dans les simulations d'intervention d'urgence et dans la
préparation des projets , la nécessaire collaboration avec les
contributeurs de la cartographie humanitaire


C'est sûr que la préparation aux réponses d'urgence est un domaine 
intéressant à développer, en particulier pour minimiser le travail et 
l'improvisation de ceux qui sont sur le terrain.


Dans le genre de détail tout bête, j'ai découvert que pas mal de 
sinistrés étaient relogés dans les écoles. Ça fait une petite motivation 
supplémentaire pour la petite fourmis escaladant le flanc de la grande 
montagne pour cartographier les établissements scolaire. Mais s'il y 
avait une liste plus importante de détails utiles en intervention 
d'urgence, ça pourrait être intégré dans la page wiki sur Madagascar.


Toujours dans une perspective de préparation, j'ai demandé à Bing de 
fournir des vues satellites détaillées de toutes les villes de 
Madagascar non encore couvertes : Mahajanga, Maintirano, Morombe (qui a 
bien souffert), Nosy Sainte Marie et Nosy Be. C'est dans la liste 
desire to acquire.




- et Eric pour son travail de répérage des POI et de rectification de
notre cartographie sur cette région.


Certes, j'ai repassé en revue toutes mes données sur Tuléar... mais j'ai 
fini il y a seulement trois jours.




Éric

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


Re: [OSM-talk-fr] Rendu des terrains de sport

2013-03-24 Thread Philippe Verdy
Dans les faits la fédé n'impose des dimensions aux terrains qu'à
partir d'un cetain niveau de compétition. En ligue 1 les dimensions
demandées sont supérieures, et pour obtenir un agréement dans une
compétition internationale, c'est encore plus. Il y a aussi des
exigences au niveau des places de tribunes.

Mais pour le foot amateur, chaque collectivité ou club se débrouille
avec ses moyens.

Il n'y a pas de dimension standard, juste des longueurs/largeurs
minimales et maximales, la règle du jeu imposant une ligne centrale,
une taille pour le ond central de mise en jeu, et des lignes de but et
de pénalité mesurées depuis le centre des buts. L'écartement des buts
n'est lui non plus pas standard (là aussi un minimum et un maximum,
les terrains amateurs ayant souvent des buts plus petits (parfois très
réduits pour la pratique du football par des enfants: selon les
catégories, cadets et minimes par exemple, ce ne sont pas les mêmes
compétitions avec les mêmes contraintes demandées aux clubs).

Et malgré cela, s'il y a accord entre les équipes devant entrer en
compétition, et si cela ne nuit pas à l'équité des championnats
auxquels ils participent, ce qui nécessaite l'accord aussi de la fédé
organisatrice de la compétition) les dimensions de terrains peuvent
passer outre les règles standards (par exemple si le terrain habituel
d'une équipe est temporairement inutilisable, inondé par exemple, il
peut y avoir accord pour jouer sur un autre terrain plus petit, même
si la capacité d'accueil du public est plus réduite, l'important étant
que pour l'équité de la compétition, le match puisse bien avoir lieu.

Bref pour le rendu des terrains de foot, je suggère plus de
flexibilité et ne PAS appliquer une dimension unique du SVG plaqué
dessus, mais plutôt d'utiliser les mesures effectivement indiquées
dans la base pour l'aire de jeu.

En aucun cas, ce rendu de SVG ne doit déborder de la surface
indiquée dans la base. Si ça ne tient pas, dans les dimensions
mesurées, il FAUT réduire la taille de ce SVG pour qu'il tienne dedans
! (contre certaines anomalies je suggère que cette réduction de taille
ne passe pas en dessous de 50%, sinon ne plus afficher ce SVG et
plaquer juste une mini-icône de ballon de foot au centroïde calculé
(et éventuellement sortir un log des dimensions insuffisantes,
suspectes, qui peuvent indiquer un problème de géométrie de polygones,
ou le fait que ce n'est plus un terrain de foot mais qu'il est
partiellement occupé par autre chose, un bâtiment par exemple, ou une
route).

Je ne sais pas pas si le rendu peut émettre ces anomalies détectées
vers Osmose, mais il peut sans doûte les émettre vers MapDust pour que
ce soit vérifié.


Le 24 mars 2013 12:07, Ista Pouss ista...@gmail.com a écrit :
 Bonjour,

 Je me suis un peu renseigné à propos d'un terrain de foot juste à coté de
 chez moi et dont j'avais parlé à cause d'un rendu mal foutu pour le rendu
 fr :

 http://tile.openstreetmap.fr/?zoom=19lat=45.4207lon=4.4232layers=B0


 Il apparaît que, s'il se joue bien des petites compétitions amateurs sur ce
 terrain, ses dimensions sont complètement folkloriques, et notamment il est
 beaucoup trop petit.

 Il parait que la fédération française de foot homologue les terrains selon
 les niveaux de compétition ; malheureusement ces infos sont introuvables sur
 leur site (enfin... je ne les ai pas trouvées)... À quand le mouvement open
 data dans le sport ?

 Donc, si ce terrain mérite bien le terme de terrain de football (il s'y
 joue au moins 30 compétitions par week-end ! ), il est vain de vouloir y
 copier/coller un svg standard de terrain de football. Pour ce genre de cas
 un logo serait je pense plus adapté.

 Je ne sais pas comment tout cela pourrait se traduire en tags ?




 Le 22 mars 2013 19:40, Philippe Verdy verd...@wanadoo.fr a écrit :

 Un gris (neutre vu que ce n'est pas une surface naturelle comme
 l'herbe ou la terre battue) peut-être un peu bleuté pour faire encore
 plus artificiel (les nouvelles surfaces complètement artificielles de
 compétition intérieures sont carrément d'un bleu bien prononcé pour
 mieux passer à la TV, mais je n'irais pas jusque là pour les surfaces
 exérieures, là ce serait braiment trop flashy).

 Le 22 mars 2013 16:44, Christian Quest cqu...@openstreetmap.fr a écrit :
  Le 22 mars 2013 14:37, Otourly Wiki otou...@yahoo.fr a écrit :
  Le tag surface pourrait-être pris en compte pour les stades de foot en
  gore
  ?
 
  Ouaille note.
 
  J'ai modifié la couleur pour surface=grass (y compris pour les tennis
  c'était trop flashy)
  Par contre, pour l'artificiel, je ne vois pas trop quoi mettre.

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




 --
 Les dérives de rue :
 Des textes à promenades

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


Re: [OSM-talk-fr] Rendu des terrains de sport

2013-03-24 Thread Philippe Verdy
Note: le rendu du terrain de foot peut très facilement calculer le
rectangle le mieux adapté contenant le terrain :
Déjà il peut obtenir au moins une diagonale en cherchant le couple de
noeuds les plus distants (méfiance toutefois : certains terrains
peuvent apparaître ovales en prenant la forme du stade qui l'entoure,
les diagonales du terrains pourraint ne pas être les longueurs les
plus longues, la plus longue étant alors l'axe longitudinal, passant
par le centre des buts et le centre du terrain et joignant le point
médian de chaque arc de cercle).

Ensuite il peut chercher la seconde diagonale en cherchant une autre
paire de points faisant un angle supérieur à 30 degrés avec la
première diagonale (le sinus de l'angle est supérieur à 1/2, autrement
dit la distance d'une extrémité de diagonale à la droite de l'autre
diagonale doit être supérieure à la moitié de la longueur de cette
seconde diagonale).

Cette seconde diagonale devrait avoir à peu près la même mesure que la
première diagonale (-/- quelques pourcents) : cela élimine le cas des
terrains de forme ovoïde dont il n'est pas évident de trouver les
dimensions du terrain de foot placé au milieu, qui n'est souvent alors
pas qu'un terrain de foot mais un terrain pour l'athlé aussi ou
d'autres sports (raison de plus pour ne pas tenter de dessiner un
terrain de foot : sinon on peut tracer au milieu du stade un rectangle
pour le terrain de foot) .

Avec cette seconde diagonale (plus courte, très rarement exactement de
même longueur que la première, mais assez proche toute de même), on
réduit de façon équitable la première diagonale à chaque extrémité
pour qu'elle ait la même mesure. On obtient alors un rectangle parfait
permettant de calculer les dimensions réelles du SVG à positionner,
proportionner, et orienter.

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


Re: [OSM-talk-fr] Pont de Ponte-Novu détruit

2013-03-24 Thread Philippe Verdy
Il y a beaucoup plus qu'1 mètre, tu as de mauvais yeux !

Déjà la première travée est complète, seule la seconde est en partie
barrée pour ne pas s'approcher trop près du second pillier (qui est
certainement alors instable), ce qui ampute environ un mètre (il y a
d'autres photos montrant que la partie fermée par la barrière est
assez réduite et ne dépasse pas un mètre, Google Street View en
référence quelques unes sur Panoromio, qui montrent d'autres angles)
mais cette barrière métallique laisse de l'ordre de 4 mètres
accessibles sur la seconde travée jusqu'à la barrière.

Cela en fait donc encore un pont sur les 2 premières travées même si
la seconde est en partie barrée.

Et l'état n'est pas si mauvais que ça, je dirais même qu'il est
excellent et visiblement bien entretenu (sauf bien entendu au delà de
la barrière qui coupe le dernier

J'estime à vu de nez que la longueur accessible sur la moitié sud
atteint un total d'environ 10-12 mètres (sur la première travée,
partie la plus large) plus 4 mètres sur la seconde travée (moins
large). On a à peu près la même longueur accessible sur l'autre moitié
nord restante du pont.

On a à peu près la même chose de l'autre côté.

Mais étant donné la forme générale de cette structure restante, il
s'agit bien d'un pont (en fait 2 ponts distincts) car il y a bien des
arches suspendues (et en bon état). Même si ces ponts sont sans issue
(noexit=yes), et accessibles uniquement aux piétons et vélos).

Bref je mettrais bien deux ponts, tous les deux en noexit=yes, tous
les deux inaccessibles aux véhicules (motorized_vehicule=no).
J'ajouterais sur le noeud à l'entrée un tag barrier pour les plots
en bétons, et une autre barrière fermant la fin de la seconde travée.

Entre les deux ponts il reste ensuite à placer un noeud ou polygone
pour la construction encore debout du pilier au pied de la rivière qui
ne soutient plus aucune arche, et là je mettrais le tag indiquant que
c'est une ruine, et le tag eindiquant sa valeur historique.

Le tout peut porter le nom Ancien pont de Ponto-Novu.

Le 24 mars 2013 10:29, Francescu GAROBY windu...@gmail.com a écrit :
 Non, ce que je dis n'est pas faux : la barrière n'est pas en bout du morceau
 de pont ! Elle est en bout de la partie praticable, mais derrière,  il reste
 encore 1 bon mètre, en plus mauvais état.
 Cf. cette autre photo de Google Street View, où le pont est plus de profil,
 et où on voit bien un morceau (légèrement en contrebas), après la barrière
 et le murets de sécurité.

 Francescu


 Le 24 mars 2013 09:31, Philippe Verdy verd...@wanadoo.fr a écrit :

 Le 23 mars 2013 21:04, Francescu GAROBY windu...@gmail.com a écrit :
  Les piétons ne peuvent même pas aller jusqu'au au bout du morceau de
  pont
  restant : une barrière de sécurité en limite l'accès.

 Ce qui est faux si on voit les photos. Il y a une première travée de
 chaque côté (sur une arche), cette travée n'est pas fermée mais
 interdite aux véhicules par des plots en béton. Au bout de cette
 travée il y a une barrière. Ensuite il y a le tro où il manque deux
 arches et où il reste un pilier.

 On ne voit pas la même chose sur les photos. Ou alors c'est qu'il y a
 eu une destruction très récente des travées qui restaient...




 --
 Cordialement,
 Francescu GAROBY

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


Re: [OSM-talk-fr] Pont de Ponte-Novu détruit

2013-03-24 Thread clansco
On Sun, 24 Mar 2013 10:09:54 +0100
Ista Pouss ista...@gmail.com wrote:

 comics_access=true
+1

-- 
Frédéric Falsetti
http://clansco.org

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


Re: [OSM-talk-fr] Pont de Ponte-Novu détruit

2013-03-24 Thread Jean-Francois Nifenecker

Le 24/03/2013 10:09, Ista Pouss a écrit :

Il faudrait tenir compte de ce que certains personnages de bandes
dessinées peuvent marcher dans le vide tant qu'ils croient être sur le
sol. Aussi je suggère :

comics_access=true



\o/

--
Jean-Francois Nifenecker, Bordeaux

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


Re: [OSM-talk-fr] Rendu des terrains de sport

2013-03-24 Thread Christian Quest
Pour info, voilà la requête utilisée pour le rendu des terrains de sport*:

select *, abs(a12-a23) as angle_diff, (a12+a23+90)/2 as angle from
(select way, sport, surface, access, way_area, st_npoints(way2) as nb,
osm_id, ST_Distance(st_pointn(way2,1),st_pointn(way2,2)) as d12,
ST_Distance(st_pointn(way2,3),st_pointn(way2,2)) as
d23,ST_Distance(st_pointn(way2,1),st_pointn(way2,3)) as d13,
degrees(st_azimuth(st_pointn(way2,1),st_pointn(way2,2))) as a12,
degrees(st_azimuth(st_pointn(way2,2),st_pointn(way2,3))) as a23 from
(select *, ST_ExteriorRing(ST_SimplifyPreserveTopology(way,100)) as
way2 from planet_osm_polygon where sport in
('tennis','soccer','basketball','rugby','rugby_union','rugby_league','american_football')
AND way  !bbox!) as simplified) as simplified2

Que fait-elle ?

1) elle cherche bien sûr les polygones avec les tags sports qui sont
pris en compte
2) elle calcule un polygone simplifié au maximum ce qui transforme les
patatoïdes en rectangle (ST_SimplifyPreverseTopology +
ST_ExteriorRing)
3) elle calcule les longueurs des segments 1-2 2-3 et la diagonale 1-3
(st_distance/st_pointn), ainsi que les orientations des segments 1-2
et 2-3.
4) elle sort le polygone d'origine + la différence d'orientation des 2
premiers segments + l'orientation moyenne + les longueurs et la
surface du terrain

Ensuite côté feuille de style, des contrôles sont faits pour
n'appliquer un SVG que dans certaines limites de dimensions et
d'angles, et pour ce qui est des terrains de foot de dimension
variable pour adapter le SVG aux dimensions en terrain (3 tailles
prises en compte).

Bien sûr, c'est un rendu approximatif, qui n'a qu'une vocation
d'information, c'est à dire que voir un tracé reconnaissable
correspond à un terrain de foot ou de tennis apporte une info
graphique sur le type de terrain, pas sur la position exacte des
marques.

Si les dimensions ou la forme ne permettent pas de caler le SVG, le
SVG n'est pas rendu.

Bien sûr, les terrains rectangulaires (touche Q dans JOSM) donnent les
meilleurs résultats, mais ça marche quand même pas mal pour des cas
comme:

http://tile.openstreetmap.fr/?zoom=18lat=48.78543lon=2.36074
http://tile.openstreetmap.fr/?zoom=18lat=48.85609lon=2.4128
http://tile.openstreetmap.fr/?zoom=18lat=48.88662lon=2.44061

* rappel, tout est sur github: https://github.com/cquest/osmfr-cartocss
-- 
Christian Quest - OpenStreetMap France
Synthèse du Week-end SOTM-FR à Lyon : http://openstreetmap.fr/synthese-sotmfr

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


[OSM-talk-fr] Carto-partie pendant la Semaine digitale à Bordeaux

2013-03-24 Thread Jean-Francois Nifenecker

Bonsoir,

Bordeaux va ouvrir demain 25 mars sa 3e semaine digitale. Le projet 
OpenStreetMap et l'Abul y tiendront un rôle puisque nous allons animer 
une après-midi de cartographie, samedi 30.


Pour plus d'informations, voir :
http://www.abul.org/La-3e-Semaine-digitale-a-Bordeaux.html

Également, tous les soirs au Village de l'innovation, à partir de mardi 
26 à 17 heures, présentation du projet OpenStreetMap et des outils qui 
l'accompagnent.


À bientôt,
--
Jean-Francois Nifenecker, Bordeaux

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


[OSM-talk-fr] De l'utilité des riverbank pour les streams ou drain ou ditch

2013-03-24 Thread PierreV
Bonsoir,

Je suis en train de travailler sur la simplification des nombreux
riverbank de mon marais poitevin.

Qu'est ce que vous en pensez de l'utilité de garder l'emprise des voies
d'eau en riverbank pour ce qui correspond a des petits canaux et fossés de
type stream, drain ou ditch?

Peut être que si vous êtes d'accord avec moi que de le garder uniquement
pour les river et canal on pourrait éventuellement faire une analyse
osmose pour proposer de supprimer les riverbank sur les stream, drain
ou ditch? qui serait complémentaire a Riverbank sans river?



--
View this message in context: 
http://gis.19327.n5.nabble.com/De-l-utilite-des-riverbank-pour-les-streams-ou-drain-ou-ditch-tp5754543.html
Sent from the France mailing list archive at Nabble.com.

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


Re: [OSM-talk-fr] De l'utilité des riverbank pour les streams ou drain ou ditch

2013-03-24 Thread Jean-Francois Nifenecker

Bonsoir,

Le 24/03/2013 21:00, PierreV a écrit :


Qu'est ce que vous en pensez de l'utilité de garder l'emprise des voies
d'eau en riverbank pour ce qui correspond a des petits canaux et fossés de
type stream, drain ou ditch?

Peut être que si vous êtes d'accord avec moi que de le garder uniquement
pour les river et canal on pourrait éventuellement faire une analyse
osmose pour proposer de supprimer les riverbank sur les stream, drain
ou ditch? qui serait complémentaire a Riverbank sans river?



+1

J'en ai ma claque de voir des foultitudes de riverbank autour de rien. 
Pour l'instant je ne supprime pas parce qu'il y a un gros boulot à faire 
en reconstituant le waterway et que je n'ai pas le temps.


--
Jean-Francois Nifenecker, Bordeaux

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


Re: [OSM-talk-fr] De l'utilité des riverbank pour les streams ou drain ou ditch

2013-03-24 Thread PierreV
En tout cas je ne veut pas que l'on supprimer a la volée les riverbank sans
waterway... ils sont par chez moi très utiles... mais propose de les
supprimer seulement après que les petits cours d'eau sont en waterway de
supprimer les riverbank.



--
View this message in context: 
http://gis.19327.n5.nabble.com/De-l-utilite-des-riverbank-pour-les-streams-ou-drain-ou-ditch-tp5754543p5754548.html
Sent from the France mailing list archive at Nabble.com.

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


Re: [OSM-talk-fr] De l'utilité des riverbank pour les streams ou drain ou ditch

2013-03-24 Thread Frédéric Rodrigo

Le 24/03/2013 21:00, PierreV a écrit :

Bonsoir,

Je suis en train de travailler sur la simplification des nombreux
riverbank de mon marais poitevin.

Qu'est ce que vous en pensez de l'utilité de garder l'emprise des voies
d'eau en riverbank pour ce qui correspond a des petits canaux et fossés de
type stream, drain ou ditch?

Peut être que si vous êtes d'accord avec moi que de le garder uniquement
pour les river et canal on pourrait éventuellement faire une analyse
osmose pour proposer de supprimer les riverbank sur les stream, drain
ou ditch? qui serait complémentaire a Riverbank sans river?


Le principe d'Osmose n'est pas de diriger le mapping vers une façon de 
faire, mais de l'accompagner.


Je reconnais que le cadastre est fantaisiste la dessus, mais il n'a pas 
de notion de filaire.


Frédéric.


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


Re: [OSM-talk-fr] De l'utilité des riverbank pour les streams ou drain ou ditch

2013-03-24 Thread Jo.
Pareil de mon coté... C'est un gros boulot de retouche. J'ai l'impression
que c'est une importation du cadastre et parfois les fossé n'existe plus.
Surtout on n'a aucune information sur le sens d'écoulement.

J'ai tenté plusieurs fois de scinder les waterbank pour en conserver une
partie que je convertissait en stream, drain ou ditch mais dès qu'il y a de
nombreux bras ça demande beaucoup de concentration et d'agilité cérébrale
pour ne pas perdre de donnée par erreur.

Pour l'instant avec JOSM j'utilise le greffon FastDraw dans un nouveau
calque où je décalque approximativement le centre des waterbank. Après
vérification, je fusionne les calque et supprime les waterbank concerné.
Lors de l'utilisation du greffon FastDraw, le touche « Q » permet de régler
la simplification du traçage.

C'est pour l'instant la méthode la plus simple que j'ai trouvé.





Le 24 mars 2013 21:17, Jean-Francois Nifenecker 
jean-francois.nifenec...@laposte.net a écrit :

 Bonsoir,

 Le 24/03/2013 21:00, PierreV a écrit :


 Qu'est ce que vous en pensez de l'utilité de garder l'emprise des voies
 d'eau en riverbank pour ce qui correspond a des petits canaux et fossés
 de
 type stream, drain ou ditch?

 Peut être que si vous êtes d'accord avec moi que de le garder uniquement
 pour les river et canal on pourrait éventuellement faire une analyse
 osmose pour proposer de supprimer les riverbank sur les stream,
 drain
 ou ditch? qui serait complémentaire a Riverbank sans river?


 +1

 J'en ai ma claque de voir des foultitudes de riverbank autour de rien.
 Pour l'instant je ne supprime pas parce qu'il y a un gros boulot à faire en
 reconstituant le waterway et que je n'ai pas le temps.

 --
 Jean-Francois Nifenecker, Bordeaux


 __**_
 Talk-fr mailing list
 Talk-fr@openstreetmap.org
 http://lists.openstreetmap.**org/listinfo/talk-frhttp://lists.openstreetmap.org/listinfo/talk-fr

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


[OSM-talk-fr] Chute de Bangui : participation à la création de données pour la réponse humanitaire

2013-03-24 Thread Severin MENARD
Bonjour,

En réponse à cette crise politique qui risque fort de devenir une crise
humanitaire, HOT relance le processus de cartographie et de consolidation
des données sur la Centrafrique. Voici un extrait de la page
wikihttp://wiki.openstreetmap.org/wiki/FR:WikiProject_Central_African_Republic
que
je viens de mettre à jour:

*A partir de la mi-octobre 2012, HOThttp://wiki.openstreetmap.org/wiki/HOT a
mené un projet de cartographie en République Centrafricaine dans le cadre
du projet EUROSHA http://hot.openstreetmap.org/projects/eurosha_0 avec
France Volontaires. L'objectif initial est la cartographie de base de
Bangui et les régions où les besoins sont les plus cruciaux. Toutes ces
activités ont pour but de soutenir la croissance d'un groupe local OSM en
Centrafrique et la croissance du projet OSM.*

*En décembre 2012, suite à l'avancée des rebelles depuis le Nord du pays,
le projet EUROSHA sur place a été suspendu et
HOThttp://wiki.openstreetmap.org/wiki/HOT a
décidé d’activer la communauté OSM afin de cartographier des Aires
d'Intérêt identifiés par les acteurs humanitaires en RCA (voir Activités
actuelles) à l'aide de plusieurs tâches du Tasking Manager (voir plus bas).
Les zones habitées des localités considérées comme de haute importance ont
été cartographiées entre 80 et 100%, à l'exception de Bambari (20%),
seulement couverte par une imagerie Orbview3.*

Pour ceux intéressés, il est donc possible de participer à la cartographie
des principales localités du pays, de son réseau routier ou de faire
progresser la qualité des données OSM qui seront utilisées par les acteurs
humanitaires.


Bien cordialement,


Severin MENARD

membre de HOT et de OSM France
___
Talk-fr mailing list
Talk-fr@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk-fr


Re: [OSM-talk-fr] De l'utilité des riverbank pour les streams ou drain ou ditch

2013-03-24 Thread PierreV
Frédéric Rodrigo-2 wrote
 Le principe d'Osmose n'est pas de diriger le mapping vers une façon de 
 faire, mais de l'accompagner.
 
 Je reconnais que le cadastre est fantaisiste la dessus, mais il n'a pas 
 de notion de filaire.
 
 Frédéric.

Concrètement qu'est ce que tu veut dire par rapport a ce que je propose?

Actuellement il y a une analyse qui propose qu'il faut mapper en filaire
lorsqu'il y a du surfacique
Je propose qu'il soit crée une analyse complémentaire qui propose de
supprimer du surfacique lorsque le fillaire est identifié comme petite
voie d'eau...

Alors soit y'a les deux (si mon idée est acceptée par tout le monde) soit
aucune des deux doivent être présentes sur osmose?
En tout cas c'est la deuxième solution que je comprend si je suis bien ta
réponse?




--
View this message in context: 
http://gis.19327.n5.nabble.com/De-l-utilite-des-riverbank-pour-les-streams-ou-drain-ou-ditch-tp5754543p5754556.html
Sent from the France mailing list archive at Nabble.com.

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


Re: [OSM-talk-fr] De l'utilité des riverbank pour les streams ou drain ou ditch

2013-03-24 Thread Frédéric Rodrigo

Le 24/03/2013 21:38, PierreV a écrit :

Frédéric Rodrigo-2 wrote

Le principe d'Osmose n'est pas de diriger le mapping vers une façon de
faire, mais de l'accompagner.

Je reconnais que le cadastre est fantaisiste la dessus, mais il n'a pas
de notion de filaire.

Frédéric.


Concrètement qu'est ce que tu veut dire par rapport a ce que je propose?

Actuellement il y a une analyse qui propose qu'il faut mapper en filaire
lorsqu'il y a du surfacique


Il faut toujours du filaire, non pas à la place mais en premier lieu.


Je propose qu'il soit crée une analyse complémentaire qui propose de
supprimer du surfacique lorsque le fillaire est identifié comme petite
voie d'eau...


Avoir du surfacique n'est pas en soit un problème. Le problème c'est la 
qualité du surfacique du cadastre.



Alors soit y'a les deux (si mon idée est acceptée par tout le monde) soit
aucune des deux doivent être présentes sur osmose?
En tout cas c'est la deuxième solution que je comprend si je suis bien ta
réponse?


Et non, il n'y aura pas d'analyse Osmose pour proposer de supprimer de 
la données.


Frédéric.


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


Re: [OSM-talk-fr] De l'utilité des riverbank pour les streams ou drain ou ditch

2013-03-24 Thread Philippe Verdy
sur les streams c'est discutable, certains ont des lits très larges
même si la plupart du temps leurs cours est assez réduit et qu'on les
considère comme des ruisseaux. D'autre part des rivières sont nommées
ruisseau alors qu'il méritnet largement d'avoir des lits assez
larges.

L'analyse debrait donc se faire sur la largeur moyenne mesurée autour
du cours d'eau central.

Le 24 mars 2013 21:25, PierreV belett...@hotmail.fr a écrit :
 En tout cas je ne veut pas que l'on supprimer a la volée les riverbank sans
 waterway... ils sont par chez moi très utiles... mais propose de les
 supprimer seulement après que les petits cours d'eau sont en waterway de
 supprimer les riverbank.

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


Re: [OSM-talk-fr] De l'utilité des riverbank pour les streams ou drain ou ditch

2013-03-24 Thread DH

Le 24/03/2013 21:00, PierreV a écrit :

Bonsoir,

Je suis en train de travailler sur la simplification des nombreux
riverbank de mon marais poitevin.

Qu'est ce que vous en pensez de l'utilité de garder l'emprise des voies
d'eau en riverbank pour ce qui correspond a des petits canaux et fossés de
type stream, drain ou ditch?

Peut être que si vous êtes d'accord avec moi que de le garder uniquement
pour les river et canal on pourrait éventuellement faire une analyse
osmose pour proposer de supprimer les riverbank sur les stream, drain
ou ditch? qui serait complémentaire a Riverbank sans river?


En dessous de 5 m de large, le riverbank est superfétatoire car 
illusoire en terme de maintenabilité.


mes 0,02€
Denis

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


Re: [OSM-talk-fr] De l'utilité des riverbank pour les streams ou drain ou ditch

2013-03-24 Thread Philippe Verdy
Le 24 mars 2013 21:00, PierreV belett...@hotmail.fr a écrit :
 Qu'est ce que vous en pensez de l'utilité de garder l'emprise des voies
 d'eau en riverbank pour ce qui correspond a des petits canaux et fossés de
 type stream, drain ou ditch?

En revanche OK pour les drain et ditch ce sont des fossés artificiels,
leur tracé est normalement très peu large, souvent très régulier sans
méandres, leur origine étant humaine et non naturelle.

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


Re: [OSM-talk-fr] Rendu des terrains de sport

2013-03-24 Thread Philippe Verdy
Le 24 mars 2013 17:09, Christian Quest cqu...@openstreetmap.fr a écrit :
where sport in
 ('tennis','soccer','basketball','rugby','rugby_union','rugby_league','american_football')
 AND way  !bbox!) as simplified) as simplified2

Est-ce que pour les stades patatoïdes, cela ne vaudrait pas le coup
de tracer effectivement la zone rectangulaire de jeu, seule celle-ci
indiquant le type de sport pratiqué (football:soccer, football
américain: american_footlbal, tennis,
rugby:rugby_union/rugby_league, handball, basketball) sans
indiquer le/les sports au niveau du stade lui-même (qui est souvent
multisport) ?

Autre problème: il y a souvent des demi-terrains de basketball (avec
un seul panneau de but, destiné à l'entrainement, mais qui fait aussi
l'objet de règles de jeu spécifiques avec des équipes réduites). Ils
sont courants en milieu urbain.

Au fait il ne semble y avoir aucun traitement pour le volleyball. Ni
non plus su le stade est marqué pour l'athlétisme et dispose d'une
piste circulaire autour du terrain (cela inclut pratiquement tous les
grands stades, surtout ceux avec des tribunes, même s'ils sont
utilisés toute l'année pour le football mais pas tous les jours).

De plus les stades à tribune couverte pourraient avoir une surface de
jeu partiellement couverte, même si pour le football la partie
centrale du jeu (et souvent aussi la piste d'athlétisme autour) ne
l'est pas (ce qui permet de tracer effectivement le terrain : là un
rectangle unique suffit (qu'on force en rectangle avec la touche Q
dans JOSM), et ça évite de tout taguer pour le football uniquement, en
laissant les détails autour apparaître sans être surchargé par le
tracé du foot..

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


Re: [OSM-talk-fr] De l'utilité des riverbank pour les streams ou drain ou ditch

2013-03-24 Thread PierreV
verdy_p wrote
 En revanche OK pour les drain et ditch ce sont des fossés artificiels,
 leur tracé est normalement très peu large, souvent très régulier sans
 méandres, leur origine étant humaine et non naturelle.

!je ne serais pas aussi catégorique: il existe quand même des fossés
naturels... qui par leur taille ne peuvent etre tagué en stream...



--
View this message in context: 
http://gis.19327.n5.nabble.com/De-l-utilite-des-riverbank-pour-les-streams-ou-drain-ou-ditch-tp5754543p5754577.html
Sent from the France mailing list archive at Nabble.com.

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


Re: [OSM-talk-fr] De l'utilité des riverbank pour les streams ou drain ou ditch

2013-03-24 Thread Philippe Verdy
Le 24 mars 2013 22:22, PierreV belett...@hotmail.fr a écrit :
 verdy_p wrote
 En revanche OK pour les drain et ditch ce sont des fossés artificiels,
 leur tracé est normalement très peu large, souvent très régulier sans
 méandres, leur origine étant humaine et non naturelle.

 !je ne serais pas aussi catégorique: il existe quand même des fossés
 naturels... qui par leur taille ne peuvent etre tagué en stream...

J'ai précisé les drain et ditch, et cela ne s'applique pas aux fossés
naturels. J'avais aussi précisé fossés artificiels (entre drain et
ditch l'écart sémantique est très faible, pour ne pas dire impossible
à distinguer assez souvent).

D'autre part parmi les canaux :
- les rigoles d'irrigation ou d'adjonction d'eau vers un moulin ou un
lavoir ne sont ni des drain ni des ditch, mais bien des canaux,
même s'ils ne sont pas navigables.
- même chose pour les aqueducs (enterrés dans une buse ou dans un
tunnel, couverts, ou à ciel ouvert). Certains sont assez larges pour
mériter un rendu de leur surface, d'autres non.

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


Re: [OSM-talk-fr] cadastre.openstreetmap.fr HS

2013-03-24 Thread Jean-Claude Repetto

On 04/03/2013 21:46, Jocelyn Jaubert wrote:


Je viens de redémarrer le serveur, parce que je n'arrivais plus à y
accéder par ssh.

Ça a l'air de remarcher correctement maintenant.




Il semble être de nouveau en panne.

Jean-Claude


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


[OSM-ja] GPSログをOSMに簡単アップロードできるiPhoneアプリありますか?

2013-03-24 Thread togawa manabu
はじめまして。最近OSM入力しはじめたばかりの 十河と申します。

iPhoneアプリのPushpinが初心者でも簡単に入力できてとても気に入っています。
同じようにGPSトレースログも簡単にOSMにアップロードできればと思うのですが、スタートボタン-移動-ストップ-アップロードという感じで簡単にトレースログをアップロードできるような、よいアプリはありますでしょうか?

どうぞよろしくおねがいいたします。
___
Talk-ja mailing list
Talk-ja@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk-ja


[Talk-GB] Windermere and Rydal Water - missing

2013-03-24 Thread Nick Whitelegg

Any reason why these seem to have disappeared or is it just an incorrect edit?

If it's just a mistake I'll add Rydal Water back later, whether I have the time 
to do Windermere is another matter though unfortunately...

Thanks,
Nick
___
Talk-GB mailing list
Talk-GB@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk-gb


Re: [Talk-GB] Windermere and Rydal Water - missing

2013-03-24 Thread Dave F.

What do you think is wrong with them?

They're visible in mapnik,  most ways haven't been edited since 
January. There does appear to be a bit of pointless overuse of 
multipolygons to represent riverbanks, but nothing that makes them 
'missing'.


Dave F.



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


Re: [Talk-GB] Windermere and Rydal Water - missing

2013-03-24 Thread Nick Whitelegg

1. Didn't appear on Freemap when I updated the data earlier today (first update 
since the 64-bit ID issue in Feb)

2. Went onto Potlatch and the natural=water polygon seemed to be missing in 
both cases.

Other lakes in the Lake District all ok.


-Dave F. dave...@madasafish.com wrote: -
From: Dave F. dave...@madasafish.com
Date: 24/03/2013 11:11AM
Cc: talk-gb@openstreetmap.org
Subject: Re: [Talk-GB] Windermere and Rydal Water - missing

What do you think is wrong with them?

They're visible in mapnik,  most ways haven't been edited since 
January. There does appear to be a bit of pointless overuse of 
multipolygons to represent riverbanks, but nothing that makes them 
'missing'.

Dave F.



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

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


  1   2   >