Re: [Talk-us] US highway classification

[Alex Mauer]

On 05/27/2011 09:06 AM, Richard Welty wrote:
> if you peruse the wiki, and make a reasonably through search
> for definitions of trunk in the US, you will find an extensive
> complex of contradictions and inconsistencies.

Maybe someone should find all these and bring it up on the list so that
a definition can be determined and the inconsistencies can be fixed?

Just saying “the definition is inconsistent so I’ll just use my own
interpretation” isn’t very constructive.

—Alex Mauer “hawke”


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

Re: [Talk-us] Creating relations for abandoned railway lines

[Alex Mauer]

On 01/10/2011 11:05 AM, Nathan Edgars II wrote:

“&” is not an abbreviation, so it should be “Elgin&  Belvidere Electric
Company”


Say what? http://books.google.com/books?id=FI0pYAAJ&pg=PA390


Ah, indeed.  The company is in fact called “Elgin and Belvidere Electric 
Company”.


I had assumed that the name was correctly “Elgin & Belvidere”.  Were 
that the case, it would be wrong to replace “&” with “and” on the 
mistaken idea that “&” is an abbreviation.


—Alex Mauer “hawke”


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

Re: [Talk-us] Creating relations for abandoned railway lines

[Alex Mauer]

On 01/10/2011 10:23 AM, Nathan Edgars II wrote:

On Mon, Jan 10, 2011 at 11:11 AM, Kristian M Zoerhoff

operator = Elgin&  Belvidere Electric Co.

This should be unabbreviated: Elgin and Belvidere Electric Company.


“&” is not an abbreviation, so it should be “Elgin & Belvidere Electric 
Company”


—Alex Mauer “hawke”


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

Re: [Talk-us] US highway tagging (was Re: highway shields: "get your kicks, where?")

[Alex Mauer]

On 01/05/2011 09:50 AM, Richard Weait wrote:

On Tue, Jan 4, 2011 at 4:09 PM, Alex Mauer  wrote:


Not sure if you’re looking for commentary on the shield overlay in general,
but it seems like it has some problems.  Take a look at I-39/US-51 here[1].
  Only one shield for I-39 until you scroll all the way south to Bloomington,
IL.

I don’t know if that’s because it’s prioritizing US shields over interstate
shields or what, but it should show both at equal frequency.  It also seems
like there are way too many US-51 shields.


When more than one relation is shared on a way, the shield placement
is sensitive to relative way-length, and starting points.  Zoom in a
bit and you get alternating shields.


You have to zoom in quite a bit (z11) to start seeing I-39 regularly. 
It makes it quite a bit harder to follow the route.



The correct way to do this will be to find co-incident relations, and
build a combined shield to place at each shield location, rather then
alternating positions.  This scales better for multiple co-incident
relations.  And it looks great.


It seems to me that the correct way is to actually alternate 
positions…at zoom 9 you see one I-39 shield near Wausau, and then 
bunches of US-51 shields as you go south.


I’m sure that building a combined shield would also do the job though, 
as long as it doesn’t end up too wide.



I've added state routes in several states.  Check your favorite places
in CA, CO, NH, NY, OH, MA, and a few others (so far).  I also added
some shields in Australia the other day.


Ah, I only looked in WI.


I’d give the shields a black outline rather than putting them on a solid
black box.


I understand that others will make different rendering choices when
they build their styles.  ;-)

I have nothing against the "die-cut" style of US shield, but the ones
I see posted on the roads 'round here have the black rectangular
background.  I think CA still uses the die-cut shield.  Do others?


As NE2 said, the black background is common on standalone signs but the 
die-cut style is used on the green guide signs [1].  I find the black 
corners distracting, and it loses the distinctive shape of the US sign. 
 It also looks sort of like the artifacts you see when text has the 
wrong-color background or a colored background instead of a transparent one.


Maybe try just a black outline on the shaped shield, or even just 
beveling the corners of the black outline?


—Alex Mauer “hawke”

1. http://www.aaroads.com/delaware/delaware010/us-040_eb_at_de-001_sb.jpg


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

Re: [Talk-us] US highway tagging (was Re: highway shields: "get your kicks, where?")

[Alex Mauer]

On 01/04/2011 10:49 AM, Richard Weait wrote:

Then let me make this point absolutely clear.  Don't look at the
background layer.  It doesn't matter at all.  Look at the shield
overlay.  The shield overlay could be added to any rendering layer.


Not sure if you’re looking for commentary on the shield overlay in 
general, but it seems like it has some problems.  Take a look at 
I-39/US-51 here[1].  Only one shield for I-39 until you scroll all the 
way south to Bloomington, IL.


I don’t know if that’s because it’s prioritizing US shields over 
interstate shields or what, but it should show both at equal frequency. 
 It also seems like there are way too many US-51 shields.


I assume it’s not expected to display state routes, at least not yet.

I’d give the shields a black outline rather than putting them on a solid 
black box.


Other than that it looks great.

—Alex Mauer “hawke”

1. 
http://weait.com:8080/map/shield2.html?zoom=9&lat=44.90346&lon=-89.61928&layers=0BTT



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

[Talk-us] Divided diamond interchanges in the US

[Alex Mauer]

I recently stumbled upon an article[1] about the new use of the divided 
diamond interchange design in the US.


It seems that the first one[2] is here[3] in Missouri and as yet unmapped.

A second one in the same city is here[4], and it appears that the old 
interchange hasn’t been mapped either.


Aerial photos on Google and Yahoo are both out of date, so I can’t map 
them myself.


Is anyone in that area and able to update these interchanges?

—Alex Mauer “hawke”

1. 
http://www.core77.com/blog/technology/video_visualization_of_a_new_type_of_traffic-improving_intersection_17734.asp


2. 
http://en.wikipedia.org/wiki/Diverging_diamond_interchange#Use_in_North_America


3. http://www.openstreetmap.org/?lat=37.24975&lon=-93.31073&zoom=15&layers=M

4. http://www.openstreetmap.org/?lat=37.24975&lon=-93.31073&zoom=15&layers=M


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

Re: [Talk-us] Highway Tagging Consensus to Improve OSM (and address some of 41 latitude's concerns)

[Alex Mauer]

On 10/26/2010 12:42 PM, Anthony wrote:

As for the question of tagging, basically you can use relations, or
you can hack something up to simulate relations (specifically, to
handle the very common situation where there is more than one route
using the same way) without actually using relations.


I haven’t seen anyone say that route relations are not the way to go. 
Have you?



Then there's the question of how to render it.  Probably something to
discuss on a different list, like a mapnik-specific one.


Why?  It’s a mostly US-specific problem, though I hear that at least 
shield rendering is also desired in Australia.


—Alex Mauer “hawke”


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

Re: [Talk-us] Highway Tagging Consensus to Improve OSM (and address some of 41 latitude's concerns)

[Alex Mauer]

On 10/26/2010 12:15 PM, Richard Weait wrote:

These rendering decisions are completely unrelated to the discussion
of how shields might best be tagged.


This portion of the thread clearly moved on to a different but related 
topic as soon as someone said “Some shields are poorly-designed for

display in a limited number of pixels”.  Thanks though.

—Alex Mauer “hawke”


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

Re: [Talk-us] Highway Tagging Consensus to Improve OSM (and address some of 41 latitude's concerns)

[Alex Mauer]

On 10/26/2010 10:50 AM, Nathan Edgars II wrote:

The actual size of a circular 7 shield generated by Mapnik.


Yeah, but is it set in stone that it Cannot Be Larger Than It Is Now?  I 
doubt it.  And I feel that gaining the ability to have state-specific 
shields is worth giving up a tiny bit of space.



Is it not possible to render an icon at any size we want?  Yes, if it’s too
big it will not work on the map, but I think 20×20 is quite reasonable (and
readable).


It's a tradeoff where bigger shields reduce the space for other features.


Sure, but that doesn’t mean that we can’t adjust to give a little more 
space to highway shields.


Not that we have to or should follow the lead of other map sites, but it 
seems like 17×17 is a pretty bare minimum.


Mapquest US highway: 24×22
Mapquest Interstate: 22×23
Mapquest generic state: 22×18

Google US highway: 22×22
Google Interstate: 20×21
Google generic state: 24×16

Yahoo US highway: 17×17
Yahoo Interstate: 17×18
Yahoo generic state: 22×13


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

Re: [Talk-us] Highway Tagging Consensus to Improve OSM (and address some of 41 latitude's concerns)

[Alex Mauer]

On 10/26/2010 09:11 AM, Nathan Edgars II wrote:

Attached are the bitmaps of the shield that is "poorly-designed for
display in a limited number of pixels."  The first one is 39x39 pixels,
and the second is 20x20 pixels.  Both are quite readable.


It's actually 17x17 that you want:
http://upload.wikimedia.org/wikipedia/commons/thumb/b/b8/Colorado_7.svg/17px-Colorado_7.svg.png


Where does this number come from?

Is it not possible to render an icon at any size we want?  Yes, if it’s 
too big it will not work on the map, but I think 20×20 is quite 
reasonable (and readable).



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

Re: [Talk-us] Highway Tagging Consensus to Improve OSM (and address some of 41 latitude's concerns)

[Alex Mauer]

On 10/25/2010 04:31 PM, Nathan Edgars II wrote:

I totally agree.  My point is just that some people and some states
(Michigan, Kansas) feel that the prefix itself is an important part of the
reference number: “The M in the state highway numbers is an integral part of
the designation…Michigan highways are properly referred to using the M and
never as ‘Route 28’ or ‘Highway 28’”.


It's part of the name when you're talking about the route, just like
one would say "I-95" or "US 1". It's not part of the designation as
shown in shields, either on the ground or on maps.


Oh, really?

http://en.wikipedia.org/wiki/File:M-28.svg

http://en.wikipedia.org/wiki/Michigan_Highway_System


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

Re: [Talk-us] Highway Tagging Consensus to Improve OSM (and address some of 41 latitude's concerns)

[Alex Mauer]

On 10/25/2010 02:44 PM, Phil! Gold wrote:

* Alex Mauer  [2010-10-25 12:44 -0500]:

So dealing with having a prefix in the ref is pretty much guaranteed
to be a requirement no matter what.


Not strictly.  Having a prefix in the rendering is important, but that can
be synthesized from the other tags in every suggestion that's been made.


I totally agree.  My point is just that some people and some states 
(Michigan, Kansas) feel that the prefix itself is an important part of 
the reference number: “The M in the state highway numbers is an integral 
part of the designation…Michigan highways are properly referred to using 
the M and never as ‘Route 28’ or ‘Highway 28’”.


Personally, I think it’s a bit silly, but then I’m not a resident of 
either of those states.  (I can imagine similar objections being raised 
if someone proposed removing M-, A- and B- in Britain, and simply 
inferring them from the highway=* type.)


—Alex Mauer “hawke”


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

Re: [Talk-us] Highway Tagging Consensus to Improve OSM (and address some of 41 latitude's concerns)

[Alex Mauer]

On 10/23/2010 10:46 AM, Ian Dees wrote:

On Sat, Oct 23, 2010 at 10:42 AM, Nathan Edgars IIwrote:

Because we're not in Europe? The common way to visually specify the
difference between our roads is with shields. Every single nav product I've
interacted with (Google Maps, MapQuest, Bing, Garmin, TomTom, and Google
Navigation to name a few) display the blue/red shields for interstates,
white with black outline shields for US routes, round white circles for
county roads, etc. They don't display the prefix (but they may use them in
routing).


Having the prefix in the ref does not preclude displaying it as a shield.

Much as it’s easier to store the long form and render an abbreviation, 
it’s easier to strip the prefix than to add it later.


Also, some people insist that certain states do absolutely need the 
prefix; in particular Michigan [1].


So dealing with having a prefix in the ref is pretty much guaranteed to 
be a requirement no matter what.


—Alex Mauer “hawke”

1. http://en.wikipedia.org/wiki/Michigan_Highway_System#Usage


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

Re: [Talk-us] [Tagging] how to tag US townships?

[Alex Mauer]

On 10/21/2010 07:12 PM, Anthony wrote:

On Thu, Oct 21, 2010 at 12:24 PM, Alex Mauer  wrote:

On 10/21/2010 08:06 AM, Anthony wrote:

On Thu, Oct 21, 2010 at 8:32 AM, Greg Troxelwrote:


So if we have whole-multiple-counties=5 (eg
NYC) county=6 township=7 city/town=8 then it would make sense
everywhere.


What would be an example of a township that would be at admin_level=7?


I don’t have a specific, named example, but see pages 8-32–8-35 of 
http://www.census.gov/geo/www/GARM/Ch8GARM.pdf



It’s not about whether they do that much; it’s about whether they’re
administered by a government.  School boards are a part of the government
yes, but they’re don’t govern the districts that they cover.


Absolutely they do.


In some (most) cases, you’re right, although it does vary from state to 
state.  They are special-purpose governments though.



The point of admin_level is *not* primarily to record which governments are
above another.  It’s to indicate which governments across different
countries and states are (approximately) equivalent.


Then we shouldn't use numbers, or if we're going to use numbers we
should assign those numbers in random order.


Huh?  Why?  What do you propose instead?  Please don’t say “use the name 
of the entity”, because we already have a key that does that 
(border_type) and it would make it a nightmare to make a consistent 
international map.


—Alex Mauer “hawke”


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

Re: [Talk-us] [Tagging] how to tag US townships?

[Alex Mauer]

On 10/20/2010 03:59 PM, Anthony wrote:

On Wed, Oct 20, 2010 at 4:47 PM, Anthony  wrote:

At the very least it would be nice to have a table outlining exactly
what "municipality" or "minor civil division" means for each state.
Is there one somewhere already?  Should I start one?


http://wiki.openstreetmap.org/wiki/United_States_municipalities

Can someone please turn off my need to constantly enter a capatcha
(User:User_5528)?


Is it because you’re adding external references?  That always triggers a 
CAPTCHA…


—Alex Mauer “hawke”


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

Re: [Talk-us] [Tagging] how to tag US townships?

[Alex Mauer]

On 10/20/2010 03:59 PM, Anthony wrote:

On Wed, Oct 20, 2010 at 4:47 PM, Anthony  wrote:

At the very least it would be nice to have a table outlining exactly
what "municipality" or "minor civil division" means for each state.
Is there one somewhere already?  Should I start one?


http://wiki.openstreetmap.org/wiki/United_States_municipalities

Can someone please turn off my need to constantly enter a capatcha
(User:User_5528)?


You can stop this by using interwiki links like [[wikipedia:Page on 
wikipedia]] instead of external links like 
[http://en.wikipedia.org/wiki/Page on wikipedia].  You’ll still need to 
use the CAPTCHA for other external links though.


—Alex Mauer “hawke”


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

Re: [Talk-us] [Tagging] how to tag US townships?

[Alex Mauer]

On 10/20/2010 03:47 PM, Anthony wrote:

1. http://en.wikipedia.org/wiki/Minor_civil_division


So far all (three) of the states I've checked fit fine with
admin_level=6 for county equivalent, and admin_level=8 for
municipality.


I’ve recorded what I’ve found at 
http://wiki.openstreetmap.org/wiki/United_States_admin_level along with 
some relevant references.


—Alex Mauer “hawke”


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

Re: [Talk-us] [Tagging] how to tag US townships?

[Alex Mauer]

On 10/20/2010 05:51 PM, Anthony wrote:

On Wed, Oct 20, 2010 at 6:34 PM, Alex Mauer  wrote:

I’d put town at 7, city and village at 8, based on
http://en.wikipedia.org/wiki/Administrative_divisions_of_New_York#Town and
http://en.wikipedia.org/wiki/Administrative_divisions_of_New_York#Village

Specifically, "Villages are a third layer of government, which are usually
overlaid inside a town, and co-administer with the town, county, and state."
and "To be incorporated, the area of the proposed village must have at least
500 inhabitants and not be part of an existing city or village."


I don't know.  "Cities are neither part of nor subordinate to towns
except for the city of Sherrill, which for some purposes is treated as
if it were a village of the town of Vernon."

If you're going to use 7 and 8, wouldn't 7 be city/town, and 8 be
village?  (Or, IMO more consistent with the rest of the US,
8=city/town and 9=village)?


I would, but villages are (by my understanding) closer to what any other 
municipality is in the rest of the US.


I don’t think/know that a lower level (higher number) admin_level 
necessarily implies that it must be, or can be, within a higher level 
(lower number) admin_level.  They’re just there to give an indication of 
equivalence.


—Alex Mauer “hawke”


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

Re: [Talk-us] [Tagging] how to tag US townships?

[Alex Mauer]

On 10/20/2010 05:37 PM, Nathan Edgars II wrote:

Why can't something with admin_level=x cross a border with admin_level
less than x? There are a lot of cities that are in more than one
county.


Agreed, though I think New York City is a special case since it actually 
encompasses several counties rather than simply being a part of several 
counties.



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

Re: [Talk-us] [Tagging] how to tag US townships?

[Alex Mauer]

On 10/20/2010 05:13 PM, Anthony wrote:

On Wed, Oct 20, 2010 at 5:49 PM, Anthony  wrote:

On Wed, Oct 20, 2010 at 5:21 PM, Anthony  wrote:

Oh, I'm only planning on doing PA, NJ, and FL.  Also maybe NY, as it's
the other one of the four states I've lived in.


Okay, well, I started New York, and concluded that it doesn't fit into
the design of admin_levels.


Hmm, maybe this would work:

2=national border
4=state border
5=New York City border
6=county/borough border
8=city/town border (other than NYC and Sherrill)
villages - don't use admin_level


I’d put town at 7, city and village at 8, based on 
http://en.wikipedia.org/wiki/Administrative_divisions_of_New_York#Town 
and 
http://en.wikipedia.org/wiki/Administrative_divisions_of_New_York#Village


Specifically, "Villages are a third layer of government, which are 
usually overlaid inside a town, and co-administer with the town, county, 
and state." and "To be incorporated, the area of the proposed village 
must have at least 500 inhabitants and not be part of an existing city 
or village."



City of Sherrill - is treated like a village of the town of Vernon
City of Geneva - is located within both the counties of Ontario and Seneca,
although the section in Seneca county has no population and is all water.




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

Re: [Talk-us] [Tagging] how to tag US townships?

[Alex Mauer]

On 10/20/2010 02:42 PM, Nathan Edgars II wrote:

On Wed, Oct 20, 2010 at 3:30 PM, Alex Mauer  wrote:

Perhaps we need to shift the discussion to actually figuring out a better
replacement for place=*?


place=incorporated?


I’d try to find something that wouldn’t exclude unincorporated 
communities ( 
http://en.wikipedia.org/wiki/Unincorporated_area#United_States )


—Alex Mauer “hawke”


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

Re: [Talk-us] Highway Tagging Consensus to Improve OSM (and address some of 41 latitude's concerns)

[Alex Mauer]

On 10/19/2010 04:11 PM, Anthony wrote:

Well, presumably you’d want to start your own.  That way it can always be a
perfect system in the future, never actually producing a map with the tools
that you have in the present.


What would be the point of that?


I don’t know, it’s what you seem to want to do.

I’ll take Ian’s advice and stop here.

—Alex Mauer “hawke”


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

Re: [Talk-us] Highway Tagging Consensus to Improve OSM (and address some of 41 latitude's concerns)

[Alex Mauer]

On 10/19/2010 05:24 PM, Peter Budny wrote:

Alex Mauer  writes:

You could also add a link to an SVG icon for the shield rendering into
the county boundary relation, so it would only be need to be changed
in once place.  (I know linking to such things is a little iffy
though)


I'd support this, too... for generic shields that look like "put numbers
inside an outline of the state" we could do this, and then handle the
exceptions as such, but I'd be quite happy right now just to have
ordinary symbol= tags rendered.


There was (is) some work done (I think by JohnSmith) to get this sort of 
thing done.  Wikipedia has some blank SVGs with placeholder digits which 
can be substituted; it’s not hard at all with the appropriate fonts to 
make that bit work, or to modify an existing numbered SVG.  Only problem 
is that you generally need a different sign for 3-digit vs. 2-digit 
signs (and sometimes 1-digit signs as well).



However, there are many stretches of road that are designated "Col. John
Q Public Memorial Highway" or something like that.  It only applies to
part of the route (the whole route through a state, or maybe just a
bridge or an intersection).  In that case, it belongs on the ways, not
the route.


Yup, or on another route relation.


What, so make the route relation contain sub-relations for each distinct
stretch of road, recursively, until it gets down to single ways that
can't be combined (e.g. due to different bridge/tunnel tags, speed
limits, etc)?


It could be done that way, but I was thinking of a more single-level 
approach:



 
 
 
  
  
  
  
  
  
 


  
  


  


  
  
  




  



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

Re: [Talk-us] Highway Tagging Consensus to Improve OSM (and address some of 41 latitude's concerns)

[Alex Mauer]

On 10/19/2010 03:58 PM, Peter Budny wrote:

For example, Kansas highway 18:
type = route
route = road
network = US:KS
ref = 18
(optional?) symbol=* tag


Also an optional wikipedia link.


There does seem to be some debate about county roads. I would probably
throw my vote in with something like "network=US:KS:Riley"


The county name needs to be in there, otherwise you can't tell two
county roads apart which use the same number.  (Analogously, you
wouldn't put US:STATE... how would you know which state?)


A relation with the boundary relation.  This could be done with US:STATE 
as well, but I think the use of the postal abbreviation for states is 
well-established while this is not the case for counties.


You could also add a link to an SVG icon for the shield rendering into 
the county boundary relation, so it would only be need to be changed in 
once place.  (I know linking to such things is a little iffy though)



I don't see any advantage to abbreviating the county name... that just
seems like more effort for mappers, with no real payback.  (I certainly
don't know abbreviations for all 159 counties in Georgia.)


+1.


However, there are many stretches of road that are designated "Col. John
Q Public Memorial Highway" or something like that.  It only applies to
part of the route (the whole route through a state, or maybe just a
bridge or an intersection).  In that case, it belongs on the ways, not
the route.


Yup, or on another route relation.


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

Re: [Talk-us] Highway Tagging Consensus to Improve OSM (and address some of 41 latitude's concerns)

[Alex Mauer]

On 10/19/2010 04:00 PM, Anthony wrote:

What project would you recommend?  I'm looking for a project that
creates and provides free geographic data such as street maps to
anyone who wants them.  Not one that makes maps in the present, using
the tools we have now.


Well, presumably you’d want to start your own.  That way it can always 
be a perfect system in the future, never actually producing a map with 
the tools that you have in the present.


—Alex Mauer “hawke”


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

Re: [Talk-us] Highway Tagging Consensus to Improve OSM (and address some of 41 latitude's concerns)

[Alex Mauer]

On 10/19/2010 03:27 PM, Toby Murray wrote:

So to get back to the basics of this thread... I think we can all
agree that we should (and are) using relations to represent highway
routes and that we need to get renderer support for route relations
ASAP.


+1



So then the question is what tags to use on relations.


All documented long ago at 
http://wiki.openstreetmap.org/wiki/Relation:route (especially 
http://wiki.openstreetmap.org/wiki/Relation:route#Tags )



There does seem to be some debate about county roads. I would probably
throw my vote in with something like "network=US:KS:Riley"


Yup, there’s debate about that.

I’d prefer something like US:KS:CTH or US:KS:COUNTY.  Or even US:KS:CR, 
though I don’t like the two-character code as it looks just like a state 
abbreviation.


IMO, connecting the road to the county should be done with a relation 
(super-relation actually) between the route and the boundary of the 
applicable county.



I do have one question: Is it acceptable/proper to have a name=* tag
on a relation? I have seen it on some and have actually used it a
couple of times - for example "name=KS 18"


It is incorrect on a route relation unless it does have a name, like 
“The Joe Q. Bloggs Memorial Parkway” or something like that.



The only advantage I see is that it makes things easier to read in
editors and when browsing data since the name tag is used when
displaying relations in lists or listing what relations a way is part
of instead of just showing the numeric ID. But this is a case of
tagging for tools so I could see reasonable objections to it.


Exactly.  The tools should be improved.  The interface for relations in 
josm (not sure about potlatch) is atrocious.



Let's focus on getting a concrete system in place that we
can go beat the rendering people over the head with. I think if we as
a US community come out with a solid plan and say "we need this now"
people will listen.


We’ve had one for a long time.  What’s needed is for someone to do the 
hard (“trivial” as Anthony would say) work of actually making use of the 
plan.


—Alex Mauer “hawke”


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

Re: [Talk-us] Highway Tagging Consensus to Improve OSM (and address some of 41 latitude's concerns)

[Alex Mauer]

On 10/19/2010 03:09 PM, Anthony wrote:
>> Agreed, but that does us little good when we’re trying to make a map 
in the

>> present, using the tools we have now.
>
> That's not what I'm trying to do, because I don't see the point in
> trying to do that.

…you may want to consider some other project, then.

>>> It would
>>> be trivial to write a preparser
>>
>> Sounds good.  Why hasn’t it been done, then?
>
> Because it's unnecessary, because no one has removed the ref tags 
from the ways.


Sure, it’s unnecessary…unless you want people to stop applying the ref 
tags to ways.


> It's also more difficult to write the preparser when you have
> contradictory information on the ways.  Part of the process of
> removing the information from the ways would be to reconcile
> inconsistencies and decide which of the two pieces of information is
> correct and which is incorrect.

Not necessary.  Use the route relation and ignore the way ref data.

Or if you’re particularly ambitious, just combine the two, ignoring 
duplicates and you’re good.  So a way which was tagged WI-66 and a 
member of a relation tagged with network=US:WI + ref=66 would end up 
with two final-rendering ref values (One WI-66, one US:WI 66).  It’s not 
the end of the world, and I am quite certain that it’d get fixed PDQ.


—Alex Mauer “hawke”


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

Re: [Talk-us] Highway Tagging Consensus to Improve OSM (and address some of 41 latitude's concerns)

[Alex Mauer]

On 10/19/2010 02:37 PM, Anthony wrote:

On Tue, Oct 19, 2010 at 3:21 PM, Alex Mauer  wrote:

And I agree that street relations are a better option in the long run, if a
little silly for the majority of cases where a street consists of a single
way (and also a usability nightmare in editors).  But I also don’t think
that removing the names from every way in the hopes that someone will notice
the problem and fix the renderer would be the right way to go.


I certainly wouldn't recommend removing the names from the ways until
you have the names in the relations.  At the point where you do, sure,
they should be removed.  The idea that no one will ever create a
renderer which uses the names in the relations is ludicrous.


Agreed, but that does us little good when we’re trying to make a map in 
the present, using the tools we have now.



It would
be trivial to write a preparser 


Sounds good.  Why hasn’t it been done, then?


No matter how much you may wish it were otherwise, part of the current
standard system is to apply ref=* to the ways which make up the route. Once
the route relation is better, I’m sure people will start using that instead,
and stop using the current system.


What about the route relation needs to be improved?


Renderer support, and a decision about how to handle mixed 
dual/single-carriageway roads. Should it be one relation per direction 
plus a super-relation, or one relation with roles? In either case, 
validator support needs improvement.


—Alex Mauer “hawke”


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

Re: [Talk-us] Highway Tagging Consensus to Improve OSM (and address some of 41 latitude's concerns)

[Alex Mauer]

On 10/18/2010 09:53 PM, Peter Budny wrote:

Ian Dees  writes:

On Mon, Oct 18, 2010 at 3:22 PM, Alex Mauer  wrote:
 For relations I agree, but for ways this doesn’t work.  And as renderers
 can only handle ways for now…

This is a data project, not a renderer project.


It’s actually kinda both.  Without the renderers, the data is useless 
(at least for making a map—analysis is still useful)



If the renderers aren't doing
the right thing then we need to make them do the right thing.


+1


+1 from me as well.


Continuing to use ref= tags at all when we have relations that represent
a much cleaner way to tag roads is a terrible case of tagging for the
renderer.  I think it's premature to remove ref tags, but I don't see
any point in adding them to new ways, rather than just creating a
relation.


If you want them to actually appear on this map we’re making, you kind 
of need to add them to new ways until renderers support the new system. 
 If you don’t care whether or not they appear on the map, what’s the 
point of adding them?


—Alex Mauer “hawke”


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

Re: [Talk-us] Highway Tagging Consensus to Improve OSM (and address some of 41 latitude's concerns)

[Alex Mauer]

On 10/19/2010 02:06 PM, Anthony wrote:

On Tue, Oct 19, 2010 at 2:59 PM, Alex Mauer  wrote:

On 10/18/2010 04:54 PM, Anthony wrote:

First of all, the ref tags aren't valid.  The numbers are references
of *routes*, not of *ways*.

[snip]

You could equally say “the name tags aren’t valid; the names are references
of *streets*, not of *ways*”.


I could, and I have, actually.


And I agree that street relations are a better option in the long run, 
if a little silly for the majority of cases where a street consists of a 
single way (and also a usability nightmare in editors).  But I also 
don’t think that removing the names from every way in the hopes that 
someone will notice the problem and fix the renderer would be the right 
way to go.  Same for ref tags.



 until we have something better we have to
live with it.


In terms of routes, we do have something better.  Route relations.


We don’t have something better.  We have the *start* of something better.


Fix the renderers first.


Don't tag for the renderer.


That’s not tagging for the renderer.  “Tagging for the renderer” would 
be if I wanted my fenceline to show up as a blue line at a low zoom 
level, so I might it highway=motorway.  That’s wrong.  Tagging something 
accurately, but also applying something which is not your pet schema, is 
not wrong, and is not “tagging for the renderer”.


No matter how much you may wish it were otherwise, part of the current 
standard system is to apply ref=* to the ways which make up the route. 
Once the route relation is better, I’m sure people will start using that 
instead, and stop using the current system.


—Alex Mauer “hawke”.


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

Re: [Talk-us] Highway Tagging Consensus to Improve OSM (and address some of 41 latitude's concerns)

[Alex Mauer]

On 10/18/2010 04:54 PM, Anthony wrote:

On Mon, Oct 18, 2010 at 5:48 PM, Alex Mauer  wrote:

On 10/18/2010 04:41 PM, Anthony wrote:

And, in fact, that attitude is exactly why the maps currently suck.


And having no shields at all is a big improvent.  Oh, wait, it’s not.


No, it's a step toward fixing the current mess.


No, fixing the renderers is what’s needed to fix the current “mess”.


In what strange alternate universe do you live where deleting valid
information which is stored following the current documented system, is not
vandalism?


First of all, the ref tags aren't valid.  The numbers are references
of *routes*, not of *ways*.


The numbers are references of neither.  “ways” is a concept built by 
openstreetmap, and has no true analogue in the real world.


You could equally say “the name tags aren’t valid; the names are 
references of *streets*, not of *ways*”.  But that’s both silly and 
irrelevant.  We have to apply the tags we have to the elements we have.


So just like applying a name= tag to a way to say “this way is part of 
the street named foo”, we must apply a ref= tag to a way to say “this 
way is part of the route with reference foo” in order to get it to show 
up on the map.  It’s what we’ve got for now; until we have something 
better we have to live with it.



Secondly, they are redundant.  In what strange alternative universe do
you live in where deleting redundant information is vandalism?


The world we actually live in, where sometimes you need redundant data 
in order to be able to make use of it.  At some point (hopefully) the 
renderers will be able to handle ref tags on route relations.  At that 
point, the documentation can be updated to note that applying ref tags 
to ways is deprecated (at least for routes, which AFAIK is the only 
current use of ref tags on ways).  Only then do ref tags on ways becomes 
*extraneous* as well as redundant, and they can reasonably be removed. 
And at that time I’ll be happy to be among the first to start deleting them.


Fix the renderers first.

—Alex Mauer “hawke”


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

Re: [Talk-us] Highway Tagging Consensus to Improve OSM (and address some of 41 latitude's concerns)

[Alex Mauer]

On 10/18/2010 04:41 PM, Anthony wrote:

On Mon, Oct 18, 2010 at 5:37 PM, Anthony  wrote:

On Mon, Oct 18, 2010 at 5:23 PM, Alex Mauer  wrote:

On 10/18/2010 04:16 PM, Anthony wrote:


I guess renderers are going to be wrong or now.

"For now" shouldn't last too long, though.  Just remove the ref info
from the ways, and the renderers will likely get their act together
rather quickly.


I for one would consider that to be vandalism.


Perhaps, but you'd be wrong.


And, in fact, that attitude is exactly why the maps currently suck.


And having no shields at all is a big improvent.  Oh, wait, it’s not.

In what strange alternate universe do you live where deleting valid 
information which is stored following the current documented system, is 
not vandalism?


Fix the renderers, don’t just delete valid data and hope someone else 
fixes them.


I’m sure patches would be welcome:
http://trac.openstreetmap.org/ticket/667
http://trac.openstreetmap.org/ticket/1666
http://trac.openstreetmap.org/ticket/2610
http://trac.openstreetmap.org/ticket/2864

—Alex Mauer “hawke”


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

Re: [Talk-us] Highway Tagging Consensus to Improve OSM (and address some of 41 latitude's concerns)

[Alex Mauer]

On 10/18/2010 04:16 PM, Anthony wrote:

I guess renderers are going to be wrong or now.

"For now" shouldn't last too long, though.  Just remove the ref info
from the ways, and the renderers will likely get their act together
rather quickly.


I for one would consider that to be vandalism.  I also doubt its 
efficacy, as the maintainers of the renderers have no vested interest in 
having relations render as we might like.


—Alex Mauer “hawke”


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

Re: [Talk-us] Highway Tagging Consensus to Improve OSM (and address some of 41 latitude's concerns)

[Alex Mauer]

On 10/18/2010 03:31 PM, Nathan Edgars II wrote:

On Mon, Oct 18, 2010 at 4:22 PM, Alex Mauer  wrote:

On 10/15/2010 09:44 PM, Richard Welty wrote:

i've seen an argument that the correct network value for a county
route involves using the actual county name, e.g.


I wouldn’t say it’s wrong.  “Unnecessary” probably, since county roads /
highways / trunk highways don’t, as far as I know, have different signs
within a state.


In most states they at least mention the name of the county (though we
obviously wouldn't do this on maps); Wisconsin may be alone in leaving
it off. Some counties (usually those that started signing routes
before the now-standard blue pentagon was created) have very different
designs, especially in New York (example:
http://en.wikipedia.org/wiki/File:Erie_County_Route_30_NY.svg - for


Good to know.  If they used the same pattern everywhere, just changing 
the name, I would still be in favor of simply recording that it was a 
county highway, but since different counties within a state use entirely 
different signs I stand corrected.  It is useful to have the county name 
in the network tag.


Perhaps it would be useful to make a wiki page documenting which states 
and counties have “non-standard” signs?


—Alex Mauer “hawke”


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

Re: [Talk-us] Highway Tagging Consensus to Improve OSM (and address some of 41 latitude's concerns)

[Alex Mauer]

On 10/15/2010 09:44 PM, Richard Welty wrote:

I don't think we should be storing any prefix as part of the network=*
or ref=* tags (thus my suggestion for
network=us_route/state_route/county_route or similar). For example the
"I-x" denotation shouldn't show up anywhere in our tags. If it's an
interstate it should be tagged as such (I suggest network=interstate
but I think there's a precedent on the wiki) and the renderer can add
the "I-" if it wants to.


i agree, it's a rendering prefix for a ref tag value and deserves
its own, separate tag.


For relations I agree, but for ways this doesn’t work.  And as renderers 
can only handle ways for now…


Sans prefices, the highway=motorway where US Highway 10, Wisconsin 
Highway 66, and Interstate Highway 39 run together would have 
ref=10;66;39.  Not very useful for determining which is which.



i've seen an argument that the correct network value for a county
route involves using the actual county name, e.g.


I wouldn’t say it’s wrong.  “Unnecessary” probably, since county roads / 
highways / trunk highways don’t, as far as I know, have different signs 
within a state.


For the curious, I documented CTH for the county highway ref= prefix 
partly because I’m from Wisconsin, but mostly because the 
three-character designation is not likely to be mistaken for a state or 
country designation.  (I know it’s possible to have the renderer look at 
where the way is, but it's a hell of a lot simpler to just read the prefix)


—Alex Mauer “hawke”


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

Re: [Talk-us] Abbreviation Police

[Alex Mauer]

On 08/04/2010 07:09 AM, Richard Welty wrote:
> otherwise, i'd go with local usage. some places use Service Road,
> others use Frontage Road, and i'm sure there are other usages.

Either way though, that’s not the actual name of the road.  It’s a
description  of the road’s function.  (though sometimes they are
actually named that by the local municipality as well, YMMV)

—Alex Mauer “hawke”



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

Re: [Talk-us] United States Roadway Classification Guidelines

[Alex Mauer]

On 07/27/2010 08:00 AM, Nathan Edgars II wrote:
> We have those tags: lanes=*, width=*, etc. But there's no "on the
> ground" definition of importance, and there's nothing wrong with
> tagging correctly for the renderers. Classification has been
> subjective from the beginning in the US, because there is no
> consistent government-assigned classification.

I’ve found that, when available, the HFCS (Highway Functional
Classification System—not to be confused with High Fructose Corn Syrup)
is quite consistent.  Unfortunately, it’s not available for every area.
 I am of the opinion that it should be followed when possible.  The
system described at the wiki page under discussion seems like a good way
to do it where HFCS is not available (with the addition of /trunk/ as
described below, though trunks are not always limited-access.)

> (By the way, the four lane limited access highway would still be
> highway=motorway (or highway=trunk if it has at-grade intersections).)




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

Re: [Talk-us] Fwd: Re: [OSM-talk] Mapquest launches site based on OSM!

[Alex Mauer]

On 07/19/2010 02:52 PM, Phil! Gold wrote:
> The wiki also explicitly says that you should use the two-character postal
> abbreviation for the state the road is in, but that seems to have been
> disregarded in states where a different prefix (like SR) is normally
> used.  (To be fair, there's a lot of other stuff people have done with the
> ref tags on state roads, including not having any network identifiers and
> putting the number in parentheses.)

Yeah, if people put in bad data you can’t expect to get good data out.
It just means the bad data needs to be corrected.

—Alex Mauer “hawke”



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

Re: [Talk-us] Fwd: Re: [OSM-talk] Mapquest launches site based on OSM!

[Alex Mauer]

On 07/19/2010 01:13 PM, Phil! Gold wrote:
> The problem with using refs to render state shields is that it can be
> difficult to get the right shield.  Some states use the state abbreviation
> in the road reference (so Maryland route 26 is generally written "MD 26"),
> but a bunch just use SR (so "SR 10" could be in Alabama, California,
> Florida, Georgia, Indiana, Maine, Nevada, Ohio, Tennessee, Utah, Virginia,
> or Washington state route 10).  At least the route relations have the
> state encoded specifically so the specific shield can be used.  I think
> there's more than one that uses "SH", too, and so on.

The wiki says[1] that refs should include US:[two-letter state
abbreviation].  That should make it pretty simple.

—Alex Mauer “hawke”

1.
http://wiki.openstreetmap.org/wiki/United_States_roads_tagging#State_Highways



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

Re: [Talk-us] Trail Route Relationships

[Alex Mauer]

On 05/11/2010 10:25 PM, Val Kartchner wrote:
> On Mon, 2010-05-10 at 12:07 -0500, Alex Mauer wrote:
>> On 05/07/2010 03:04 PM, Val Kartchner wrote:
>>> Which levels should be assigned as the "name" for each trail segment?
>>> The rest will be used in the "name" for the route relationships?  Once I
>>> get an answer, I'll change the trails for consistency.
>>
>> Basically, follow the directions[1] for working with route
>> relationships.  Create a relation for each trail section as you have,
>> give it the name of the section.  Then create a super-relation
>> containing all the trail sections, and give it the name of the trail
>> network.  Assign each of them an appropriate route type[2].  Add any
>> extra stuff, e.g. foot=*, bicycle=*, horse=*.
>>
>> I’ll admit to doing a bit of tagging for the renderer where I’ve had to
>> map similar trails, by using the ref=* tag to hold the section name, and
>> name=* to hold the global trail name.
> 
> The responses are for the words that I asked, but were not really for
> the question that I should have asked.  Let me try to phrase the
> question a different way.
> 
> Here is a simplified diagram of part of the trails that I need to map:
> 
>   A  B  C
> -+---+-
>  |   |
>   \_/
>  D
> 
> Sections A, B and C are part of the Ogden River Parkway.  Sections B and
> D are part of the 21st Street Pond Trail.  All of these are part of the
> Ogden Trail Network.  The Ogden River Parkway is part of the Centennial
> Trail.  The Ogden Trail Network and the Centennial Trail will obviously
> be done with relationships.
> 
> The big question is about section B: Which trail should be used for the
> "name" for this section?  Should B be part of the longer Ogden River
> Parkway then B & D be done as a relationship?  Should it be done the
> other way around?  Should B attributes be "name=Ogden River Parkway" and
> "name_1=21st Street Pond Trail"?
> 
> Which name should take priority?

If you're talking about the lowest-level name (i.e. the one on the way),
I think it would be acceptable to either use the one as signed assuming
there's only one, or otherwise just leave it empty and leave it to
renderers to pick up the name(s) from the relations.

I don't think it's a problem for B to be in the two relations...

This may be obvious, but you'd presumably need the following relations:

Relation 1, containing A, B, C
name=Ogden River Parkway

Relation 2, containing B, D
name=21st Street Pond Trail

Relation 3, containing relation 1 and 2
name=Ogden Trail Network

Relation 4, containing relation 1 (and presumably other relations)
name=Centennial Trail



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

Re: [Talk-us] Trail Route Relationships

[Alex Mauer]

On 05/07/2010 03:04 PM, Val Kartchner wrote:
> Which levels should be assigned as the "name" for each trail segment?
> The rest will be used in the "name" for the route relationships?  Once I
> get an answer, I'll change the trails for consistency.

Basically, follow the directions[1] for working with route
relationships.  Create a relation for each trail section as you have,
give it the name of the section.  Then create a super-relation
containing all the trail sections, and give it the name of the trail
network.  Assign each of them an appropriate route type[2].  Add any
extra stuff, e.g. foot=*, bicycle=*, horse=*.

I’ll admit to doing a bit of tagging for the renderer where I’ve had to
map similar trails, by using the ref=* tag to hold the section name, and
name=* to hold the global trail name.

1. http://wiki.openstreetmap.org/wiki/Relation:route#Mapping_practice
2.
http://wiki.openstreetmap.org/wiki/Relation:route#List_of_route_types_in_use



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

Re: [Talk-us] Street Naming Conventions

[Alex Mauer]

On 04/08/2010 10:32 PM, Val Kartchner wrote:
> 6) Should the direction prefix even be part of the street name since it
> (mostly) isn't on the sign?

That’s not true in all areas.  I’m in Wisconsin, and in most cities I’ve
been to, if the street has a direction prefix it’s on the sign
(abbreviated of course).

—Alex Mauer “hawke”



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

Re: [Talk-us] Questions about tagging local streets as US and state highways

[Alex Mauer]

On 01/30/2010 12:30 PM, Leroy E Leonard wrote:
> Some specific questions:
> 1. Do we need the ref tag for the way when we have the two relations
> with the same info?

Mapnik renders shields from the ref tag on the way, so yes.

> 2. Should the ref tag in the US 278 relation really be "US 278" or simply 
> "278"?

In the relation, it should be 278.  On the ways the ref tag should be
“US 278; US:GA 10”

> Philosophical questions:
> 3. This road never renders with shields. Should it? Is it a problem of
> tagging, rendering, or false expectations?

Only mapnik renders the shields, and then it uses the ref tag from the
way.  It also generally renders shields at a closer zoom than one might
expect.  So probably a mixture of all 3.

> 4. (Related to question 1) when we have duplicate information in ways
> and relations, is there a reason to having it duplicated or should it
> be moved to the relation?

Yes.  Renderers may pull their information from one source or another,
so it’s useful to have it in all places that a renderer might want.

Hope this helps.
-Alex Mauer “hawke”



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

Re: [Talk-us] Daily Updates to CloudMade Routing

[Alex Mauer]

On 01/15/2010 12:07 PM, Nick Black wrote:
> Hi Alex,
> 
> We just updated the tile set today with data from Tue 15th.  There are
> minutely tiles available at mapzen.cloudmade.com - but they aren't
> available through the API at the moment.
> 
> We're in the process of upgrading all services to daily updates - I
> know the timelyness of data has been a problem in the past so I'm sure
> this batch of updates will help.

Thanks, Nick.

I’m seeing the changes in question on mapzen.cloudmade.com, but not on
maps.cloudmade.com yet.  Is that expected?

-Alex Mauer “hawke”




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

Re: [Talk-us] Daily Updates to CloudMade Routing

[Alex Mauer]

On 01/14/2010 05:10 AM, Nick Black wrote:
> Hi guys,
> 
> Some of you use CloudMade's routing to debug your mapping.  We've had
> a lot of feedback that though useful, there are two problems:
> 
> 1 - The routing is out of date, often lagging many days behind
> 2 - There are bugs with turn restrictions
> 
> As of this week, we are updating routing every 24 hours which should
> help with checking mapping and finding errors.  Routing should update
> by early evening GMT (16.00 is the target time) with data from the
> previous day.  So at 16:00 on Friday 15th Dec, routing will have data
> up to 00:00 on Friday 15th Dec.   This is a new service, so exact
> completion times for the import will probably vary.

Is this already happening?  Does it affect rendering, or just routing?
Is it worldwide?  I have a road which I updated more than a week ago
which has not updated on cloudmade’s maps yet.

-Alex Mauer “hawke”



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

Re: [Talk-us] georgia road classifications

[Alex Mauer]

On 09/09/2009 03:17 PM, Kevin Samples wrote:
> it looks like classifying "Urban Freeways and Expressways" as motorway 
> is the way to go.  I double checked GA400, US78, and GA316 and they fall 
> in the "Urban Freeways and Expressways" where they are exclusively 
> limited access.  So after going back through the definitions from the 
> FHWA documents, here is a revised crosswalk. let me know what you think
> 
> Rural Interstate Principal Arterial - highway:motorway
> Rural Principal Arterial - highway:trunk
> Rural Minor Arterial - highway:primary
> Rural Major Collector - highway:secondary
> Rural Minor Collector - highway:tertiary
> Rural Local Road - highway:residential
> Urban Interstate Principal Arterial - highway:motorway
> Urban Freeways and Expressways - highway:motorway
> Urban Principal Arterial - highway:primary
> Urban Minor Arterial - highway:secondary
> Urban Collector Street - highway:tertiary
> Urban Local Street - highway:residential

I'd consider setting “Rural Local road” as highway=unclassified.
Otherwise, it looks great to me.

-Alex mauer “hawke”



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

Re: [Talk-us] Unpaved streets

[Alex Mauer]

On 09/09/2009 12:05 PM, Chris Hunter wrote:
> yes, yes, yes - this is the most sane summary of the road/track debate I've
> seen in quite a while.  Of course, it also reopens the track/path argument,
> but I'll leave that to others to battle out.

How is there even an argument there?  track is for cars and always has
been, while path is for not-cars and always has been.

-Alex Mauer “hawke”



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

Re: [Talk-us] Unpaved streets

[Alex Mauer]

On 09/09/2009 11:59 AM, Dale Puch wrote:
> highway= is for the type of road regardless of condition.  A paved surface
> is assumed unless otherwise stated.

...except for highway=track, (as you mention below) sure.

>   If it does not have a name (and thus no addresses), it should be a track.

Tracks can be named and have addresses, though this is uncommon.

> This would typically be a 1 lane (total) or possible wide enough for 2.
> This probably assumes an unpaved surface by default.

Yes.

-Alex Mauer “hawke”



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

Re: [Talk-us] Unpaved streets

[Alex Mauer]

On 09/08/2009 02:19 PM, Paul Fox wrote:
>  > and that
>  > most true dirt roads are unnamed.
> 
> perhaps.  but i'd say that's mostly only true if they're not
> publicly accessible.  any sort of public right-of-way usually
> comes with at least a locally-assigned number:  "Forest Route NN",
> or "Fire Road NN", or "County Road NNN".

in OSM that's ref=* not name=*

>  > And finally I would agree with you that regardless of their relative
>  > numbers, true dirt roads (not gravel) as described at
>  > http://en.wikipedia.org/wiki/Dirt_road should indeed be highway=track.
> 
> no, i don't agree.  as greg troxel (i think) said earlier, the
> term "track" implies a private right-of-way.  

That’s not correct, and is not what he said.  He said, ”highway=track,
on the other hand, seems definitely second-class ... if someone lives on
a track, their address will be a value on the real road the track
connects to.”

While generally the case, this is not a defining characteristic of a
track, and says nothing at all about whether it’s access=private.

> there are many many
> dirt roads in my travels that are better described
> "highway=residential surface=unpaved", due both to their public nature,
> and the presence of multiple residences.

Then describe them as such (though surface=dirt might be better).  But
their public nature has nothing to do with it.  We have access=* to
describe that. was only saying that *in general*, gravel roads are not
highway=track, while *in general* dirt roads are.  It’s a rule of thumb,
not an absolute.

Obviously there are exceptions.
http://en.wikipedia.org/wiki/File:DirtRoadCows.jpg should probably be
highway=unclassified, and
http://en.wikipedia.org/wiki/File:Seymour_Logging_Road.JPG should
probably be highway=track.

-Alex Mauer "hawke"



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

Re: [Talk-us] Unpaved streets

[Alex Mauer]

On 09/08/2009 01:18 PM, Paul Fox wrote:
> alex wrote:
>  > On 09/06/2009 05:56 PM, Paul Johnson wrote:
>  > > I would tend to go with highway=track unless the street in question is a
>  > > gravelled over macadam or some other semi-paved surface mostly because
>  > 
>  > I would expect that this applies to at least the majority of named roads...
>  > 
> 
> i don't understand.  "the majority of named roads" are, of
> course, paved.  

You're right there.  I should have said “The majority of named roads
which are not paved with some form of concrete”

> i would say that the next most numerous are
> simply dirt roads (at least here in new england, and in most of
> the US that i've traveled).  roads which are partly paved, or
> which are "gravel over macadam" (i'm not entirely clear on what
> that means) would be a small minority.

You can find more on macadam here: http://en.wikipedia.org/wiki/Macadam

I would consider a gravel-paved road
(http://en.wikipedia.org/wiki/Gravel_road) to be not a highway=track in
OSM terms, in most cases.

Further, I would expect that gravel-paved roads are the most common of
non-concrete-paved (including asphalt concrete) named roads, and that
most true dirt roads are unnamed.

And finally I would agree with you that regardless of their relative
numbers, true dirt roads (not gravel) as described at
http://en.wikipedia.org/wiki/Dirt_road should indeed be highway=track.

-Alex Mauer "hawke"



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

Re: [Talk-us] Unpaved streets

[Alex Mauer]

On 09/06/2009 05:56 PM, Paul Johnson wrote:
> I would tend to go with highway=track unless the street in question is a
> gravelled over macadam or some other semi-paved surface mostly because

I would expect that this applies to at least the majority of named roads...

> most routing engines and all renderers at this point are more likely to
> use the highway tag to determine and render such objects correctly.

“Tagging for the renderer” is generally discouraged.  It's probably a
better idea to tag what's actually on the ground, and put in a trac
ticket if you feel that it should be rendered differently.

-Alex Mauer “hawke”



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

Re: [Talk-us] Unpaved streets

[Alex Mauer]

On 09/03/2009 09:17 AM, Ian Dees wrote:
> No tag should ever imply any other tag. It's always better to be more
> verbose than not.

No it's not.  Are you seriously putting oneway=no (just to name one
example) on every street you tag?

-Alex Mauer "hawke"



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

Re: [Talk-us] Labeling community gardens

[Alex Mauer]

On 08/20/2009 03:12 PM, Cameron Adamez wrote:
> I was unsure what to use as a tag so some plots are tagged by  
> landuse=community_garden but I'm not sure if that is the best tag to  
> use.

That sounds like a good tag to me.

-Alex Mauer "hawke"



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

Re: [Talk-us] bike rail trail as built vs as proposed and imported

[Alex Mauer]

On 08/12/2009 06:55 AM, Greg Troxel wrote:
>> Probably not; however, it is accepted practice in OSM.  As you say,
>> someone with more familiarity with railroad procedures and how they
>> differ between the US and elsewhere might be able to answer that.
> 
> I think one of the bugs in OSM is failure to follow established
> practice.  But that's not worth worrying about in this case.

Often this is because there aren't necessarily people in the OSM
community who know where to find the established practice.  This is
presumably less true now, but I don't see anyone chiming in with "I has
the solution!!!" on this list

>> I think that's a big understatement. I would go so far as to say that
>> it's nearly impossible to take an arbitrary piece of railroad track and
>> determine whether it's abandoned or out of service (in the US legal
>> sense) -- or indeed, whether it's in fact still in service.
> 
> For tracks that are not abandoned, in service or out of service is
> determined by the railroad.

Sure, but you can't tell by looking at it.  Do they publish this somewhere?

> Yes, it is hard to find out, but that does not make it unverifiable.  I
> suspect one can write to the STB and ask and find out, although I
> haven't tried.
> 
> You are applying an unreasonable standard.  

I'm not applying a standard at all; I'm just asking if it is feasible
for someone to actually find out this information.

Not saying you can't tag this information, just that it's IMO unlikely
that a tagging scheme for this will see much use, because it's so
difficult to get the information.

Incidentally, there are other problems with the railway=abandoned /
railway=disused.  Specifically, using them loses information about what
type of railway it is ("standard" rail, tram, narrow_gauge, subway...)

See http://wiki.openstreetmap.org/wiki/Comparison_of_life_cycle_concepts
and http://wiki.openstreetmap.org/wiki/Proposed_Features/Status

-Alex Mauer "hawke"



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

Re: [Talk-us] bike rail trail as built vs as proposed and imported

[Alex Mauer]

On 08/11/2009 06:10 AM, Greg Troxel wrote:

> But, is "abandoned" really in use in other countries to mean what in the
> US we call "old railroad grade"?  (Here I am taking USGS norms to be
> established practice in the US.)  

Probably not; however, it is accepted practice in OSM.  As you say,
someone with more familiarity with railroad procedures and how they
differ between the US and elsewhere might be able to answer that.

> The Surface Transportation Board of the ICC makes abandonment decisions,
> and they are published by the federal government.  An example:
> 
> http://regulations.vlex.com/vid/railroad-abandonment-lamoille-valley-22682301
> 
> I'm not saying this is trivial to find,

I think that's a big understatement. I would go so far as to say that
it's nearly impossible to take an arbitrary piece of railroad track and
determine whether it's abandoned or out of service (in the US legal
sense) -- or indeed, whether it's in fact still in service.

If my understanding is correct:
* This several page document describes just one section of track.  So
there are many, many of these documents.
* This document just lists an intent to abandon a section of railroad.
It may or may not have been accepted by the relevant authority (although
it probably was)

Can you provide an example of the steps one would have to go through to
actually find this out for a specific piece of track?  As far as I can
tell it would involve trawling through
http://www.stb.dot.gov/decisions/readingroom.nsf/DailyReleases?OpenView
or http://www.gpoaccess.gov/fr/advanced.html (the latter of which only
goes back to 1995, and the former of which goes back to 1996)

So you might be able to find out if it *is* abandoned (If you're really
lucky it's on your other link at http://www.trainweather.com/aban.html)
but even that's extremely difficult, and it's even less possible to
determine that it's not abandoned.  It seems that the only way to do so
is to go through every single abandonment notice, and if it's not on any
of them, then it's probably not abandoned after 1995 -- though it would
be easy to miss it among the huge number of documents.  And if it is on
one of those abandonment notices, then you have to somehow figure out if
the abandonment was approved.

Do I have it right?

-Alex Mauer "hawke



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

Re: [Talk-us] bike rail trail as built vs as proposed and imported

[Alex Mauer]

On 08/08/2009 07:31 PM, Greg Troxel wrote:
> The current tag definition is awkward:
> 
>   http://wiki.openstreetmap.org/wiki/Key:railway
> 
> because tags mean something unintuitive:

It's not unintuitive, it's just not the same as US legal definitions.

disused = no longer used
abandoned = track/infrastructure removed

Is there somewhere that describes the difference between "abandoned" and
"out of service" railways, preferably something which is verifiable (in
the OSM sense, see http://wiki.openstreetmap.org/wiki/Verifiability)?

-Alex Mauer "hawke"



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

Re: [Talk-us] bike rail trail as built vs as proposed and imported

[Alex Mauer]

Greg Troxel wrote:
> Someone said railway=abandoned, but there are three separate things --
> my opinion is partly from USGS topo maps:
> 
> * out of service (disused=yes above)
>[tracks present, no trains, but not abandoned]
> 
> * abandoned (this is a legal distinction in the US)
> 
> * old railroad grade (beyond abandoned, where there are no longer tracks)

What's the difference between "out of service" and "abandoned"?  I
assume the distinction is in how often the railway is used.  I know of
some rail lines near me that are classified in the TIGER import as
abandoned (not even out-of-service) , but which still have trains run on
them occasionally (or even fairly often)

-Alex Mauer "hawke"



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

Re: [Talk-us] TIGER Data in Wrong Place

[Alex Mauer]

Dave Hansen wrote:
>> Also if you look at the diagonal (NW-SE) part of the Arizona-Mexico 
>> border there are streets in the US that appear in Mexico - the whole 
>> border area appears to be a mess. How do I know what needs to be done to 
>> fix this? The error looks larger than 100ft.
> 
> You really need some good GPS traces of the streets, or you can try to
> align it on yahoo imagery.

...which is itself off in places.  You really need some good GPS traces,
enough to figure out how far it's off and in what direction.  And
probably to determine how large the incorrect area is.

-Alex Mauer "hawke"



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

Re: [Talk-us] dirt road classification?

[Alex Mauer]

Paul Fox wrote:
> is there a way to distinguish pavement type separately from
> road classification?  is there a standard tag for doing so?
> in new england (and i assume in other parts of the country)
> it's not uncommon to have unpaved state numbered routes.  or
> unpaved residential areas. 
> 
> i turned to the wiki, but i find i need some interpretive help.  the
> third sentence at:
> http://wiki.openstreetmap.org/wiki/United_States_roads_tagging#Other_Roads
> is unparseable.  but even if i assume that the first three words
> should be dropped, i'm left with:
> "Unpaved roads, dirt track roads, forest development roads,
> jeep trails, and roads not passable by all vehicles merit
> highway=track instead."
> which seems like far too big a bucket for the roads i'm thinking
> of.  in my mind there's a big difference between "unpaved" and "track".
> 
> and finally, if i _did_ want to use "highway=track", shouldn't
> potlatch give that option in the dropdown menu?


You want the surface key.

http://wiki.openstreetmap.org/wiki/Key:surface

-Alex Mauer "hawke"



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

Re: [Talk-us] Blame me for JOSM yellowness

[Alex Mauer]

Dave Hansen wrote:
> On Fri, 2009-04-24 at 09:40 -0700, Alan Millar wrote:
>> If you don't like the tag, you don't have to use it.  But I have been
>> waiting for this highlighting feature for a long time, but never got
>> around to figuring out enough in JOSM, so I am happy to see it.  To
>> each his own; there is room for all of us in this project.
> 
> I'd also be very happy if JOSM flipped the tag for me when I edit a
> tiger object.  It seems reasonable enough that if I'm editing something
> that we can consider it reviewed.  I certainly don't want to have to go
> flip it manually every time I go fixing some minor road details.
> 
> -- Dave

Shouldn't it be removed rather than flipped?  That's what I've been doing...

-Alex Mauer "hawke"



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

Re: [Talk-us] Interstate Highways Relations List

[Alex Mauer]

Chris Lawrence wrote:
> I'd suggest that "CTH" is a little too specific to Wisconsin (where
> they are called County Trunk Highways, hence CTH); county highways are
> typically marked as "County Road x" in most states, so CR x would be
> more appropriate in most states.

It honestly doesn't matter to me what the specific code indicating a
county highway is.  Wisconsin also uses "County Road" pretty extensively.

That said, there's an advantage to using three letters: lack of conflict
with the ISO country codes used on national road systems (CR is Costa Rica)

-Alex Mauer "hawke"



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

Re: [Talk-us] Interstate Highways Relations List

[Alex Mauer]

Apollinaris Schoell wrote:
>>
>> Why make this more complicated than it has to be?  Leave the names on
>> the underlying way, not the relations; leave the refs on the  
>> relations,
>> not the underlying ways.  Then it's a matter of fixing mapnik and  
>> t...@h to
>> do the right thing, since relations are set up better to handle things
>> like route symbols.
>>
> 
> fully agreed.
> what is the best solution for another problem I have seen.
> navigation systems should use the name/ref on the signs. the names are  
> never (rarely?) used for interstate and us routes. but commonly used  
> for county routes ( there the names is used and the ref is used rarely)

Correct the data to put the ref in on county highways.  It should be CTH
 xxx, where xxx is the reference value for the county highway.

-Alex Mauer 'hawke'



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

Re: [Talk-us] Talk-us Digest, Vol 16, Issue 14

[Alex Mauer]

Alan Millar wrote:
>> Other discussion regarding the GNIS import suggests that the area is
>> preferred if available, and the tags (especially gnis:feature_id) should
>> be copied from the node to the area (and corrected where applicable.)
> 
> Is there an easy way to copy the tags without retyping them?  I can't seem
> to get the Potlatch repeat tags or the JOSM paste tags to copy from the
> node to the way.

There's a bug open for josm about that.

http://josm.openstreetmap.de/ticket/1187

-Alex Mauer "hawke"



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

Re: [Talk-us] Node versus area for school?

[Alex Mauer]

Alan Millar wrote:
> In my area I have some schools with both a node and an area.  Most schools
> are just the node with the GNIS ID, but some have the school grounds
> mapped out as an area also.
> 
> Does anyone have any thoughts or opinions on best practices for the node
> versus area?

Other discussion regarding the GNIS import suggests that the area is
preferred if available, and the tags (especially gnis:feature_id) should
be copied from the node to the area (and corrected where applicable.)

-Alex Mauer "hawke"



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

Re: [Talk-us] Tags for US Forest Service GIS Trail Data

[Alex Mauer]

Matt Maxon wrote:
> Spencer Riddile wrote:
>> I'm working on figuring out what tags to use for the fields/columns 
>> that are included in the USFS GIS trail data that I am going to import 
>> into OSM.  Has anyone set a precedent for this already?  Would it make 
>> sense for me to add a section to the wiki page for the USFS 
>> (http://wiki.openstreetmap.org/wiki/US_Forest_Service_Data) in an 
>> effort to standardize?  This would be for trails specifically and 
>> would follow what has been established on the Hiking wiki page 
>> (http://wiki.openstreetmap.org/wiki/Hiking).
>>
> I do think a US specific trails page (s) would is needed and would be useful
> 
>> National Forest Trails
>> tags proposed:
>>
>> * owner = national_forest, private_on_nf, private_noton_nf,
>>   county, state, city_town
>> * closure_status = open, closed, restricted, decommissioned
>>
> I'd like to see some US specific (relevant) tags
> 
> Generally here in California and my experience with the west is trails 
> have a series of designations
> 
> Horse
> Foot
> Bicycle
> OHV
> Handicapped
> 4WD
> Cross Country (informal) aka - XC or CC
> 
> There are NOT recommended designations, while not specifically 
> prohibiting an activity you'd be wise to heed it or at least exercise 
> caution

Is this not covered by the access key?  horse=designated,
ski:nordic=designated, etc.?

Not sure what OHV is (off-highway vehicle?), but it seems that a new
access type might be needed.

> There seems be to a  lack of discussion  about  other  agencies  BLM, 
> Park  Service  are  two biggies that come to mind.
> 
> The OHV tag would need to cover recommendations about vehicle type, 
> Motorcycle, ATV, Jeep(4wd) trail width etc



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

Re: [Talk-us] Railway spur

[Alex Mauer]

Russ Nelson wrote:
> Joseph Scanlan writes:
>  > enough, I found "railway=rail" and "service=spur" but no "railway=spur". 
> 
> Perhaps we should do an automated conversion of railway=spur into
> railway=rail+service=spur?  It would mean no editing in the case of an
> actual spur, and for anything else, it would mean changing only one
> tag rather than having to change one tag and add another.
> 
> I appreciate the that people want hand edits to be respected, but if
> railway=spur is never correct, even if somebody left it that way for a
> spur, changing it wouldn't hurt their edit.

Cue the complaints from the crowd who believes that map features should
document current usage, and that a tag can never be incorrect.

But seriously, that sounds great to me.

-Alex Mauer "hawke"



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

Re: [Talk-us] Railway spur

[Alex Mauer]

Ian Dees wrote:
> In my experience, TIGER ways that show up as railway are almost always not
> actually there (the right of way has been purchased by or leased to another
> entity and developed), so be extra careful when retagging TIGER railway.

Yep, that's common too, though I'm sure it varies a lot depending on the
area of the country and such.  I wouldn't trust anything other than
"on-the-ground" mapping for that, either -- rail lines are hard to see
from the aerial photos.

Once it's mapped of course, it becomes easier to determine what service
tag a rail line will need.

Of course, it doesn't help that most (all?) rail lines are posted no
trespassing...

-Alex Mauer "hawke"



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

Re: [Talk-us] Railway spur

[Alex Mauer]

Joseph Scanlan wrote:
> The airport runways, taxiways, and aprons now show on the slippy map but 
> the railways are missing.  I admit I fall firmly in the "don't tag for 
> the renderer" camp but I did get curious and checked the wiki.  Sure 
> enough, I found "railway=rail" and "service=spur" but no "railway=spur". 
> I suspect this is an artifact of the TIGER conversion but wanted to 
> check on this list before I started changing "railway=spur" tags when I 
> occasionally come across them.

That's exactly correct.  TIGER had the distinction between spur and
non-spur railways before service=spur was created.  So at the time it
was unclear how it would be handled.  They should be changed as you
described.  But be cautious, TIGER's classification of railway is very
inaccurate.  There are many routes which are marked as spur which
shouldn't be, and many which are marked as abandoned which shouldn't be.

-Alex Mauer "hawke"



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

Re: [Talk-us] Bicycle facility tags (Class III "bike route")

[Alex Mauer]

Karl Newman wrote:
> 
> I think your suggestions line up with how I would tag it, too. I'm curious
> where you heard about the Class I, II, III designations, though. I'm a
> California native and occasional bike rider and I've never heard of these.

I've never heard of them either here in wisconsin (I'm a regular bike
rider, formerly in Milwaukee and now in central WI)

It seems that the designations are not specific to california; a Google
search for "Class II" or "Class III" and "bike" suggests that the
terminology is in use throughout the US.

-Alex Mauer "hawke"



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

Re: [Talk-us] Changes to TIGER data

[Alex Mauer]

> On Tue, Aug 12, 2008 at 8:57 PM, Nathan West <[EMAIL PROTECTED]> wrote:
> 
>> I have been using potlatch to make all of my edits. Is there a way to
>> delete sections of ways and their points? This needs to be done for some
>> ancient pieces of road that do not exist anyone in my area and for small
>> sections of the old railway.
>>

I don't know of a way to do this in potlatch.  In JOSM it can be done by
splitting the way and deleting whichever section you need to.  In
theory, deleting should clear out all the unused nodes; however, all the
TIGER data has its nodes tagged, so unfortunately this won't happen and
you'll need to delete the nodes manually.

-Alex Mauer "hawke"


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

Re: [Talk-us] Changes to TIGER data

[Alex Mauer]

Adam Killian wrote:
> highway=cycleway or footway.  I don't think there's an established
> standard for multi-use paths.

highway=path + (bicycle|foot|horse|ski|etc]=(designated|yes)

A larger list of access methods can be found on the OSM wiki page 
"Key:access"

-Alex Mauer "hawke"


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

Re: [Talk-us] Changes to TIGER data

[Alex Mauer]

Nathan West wrote:
> So the final question is: If there are fairly large outdated data sets that
> may be off in location, is it easiest to just delete the old way and create
> a new route? (I can get the actual route easily)

I've actually run into exactly the same situation, and that's what I did.

It may be *easier* however, to get the actual route, compare it to the
"railroad" way, and modify tags accordingly.  For my case, the railroad
actually does exist on part of the route, and the trail goes alongside
for that portion.

-Alex Mauer "hawke"


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

Re: [Talk-us] Request for comment on a section of Interstate 40

[Alex Mauer]

Victor Snesarev wrote:
> The thought of splitting the interstate into two ways has occurred to me,
> but I thought I'd ask before undertaking such a fundamental edit. When I
> split the interstate, should I copy TIGER tags to the new way?

No, just leave them on whichever interstate direction you keep.  If you
don't keep either direction, don't worry too much about keeping the
TIGER tags at all.

> BTW, is there an area of the world that you consider to be a good example of
> how to do things?

I did most of Stevens Point and Plover in Wisconsin before the TIGER
import.  I think it makes for a pretty good example, as it uses most of
the features of OSM.

http://www.openstreetmap.org/?lat=44.5054&lon=-89.5516&zoom=12&layers=B00FTF

Hope this helps

-Alex Mauer "hawke"


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

Re: [Talk-us] Should I realign TIGER data to Yahoo areal imagery?

[Alex Mauer]

SteveC wrote:
> short answer - YES :-)

Slightly longer answer: Yes, but be careful.  In many places the Yahoo
aerial imagery is offset from real life by some amount (I've heard 30
meters up to 1km in some cases).  So if an area looks totally different
from what's on the ground, correcting to match Yahoo! is OK.   But if
it's just shifted in one direction, then don't correct it without first
verifying with a GPS track that the Yahoo! imagery is correct.

-Alex Mauer "hawke"


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

Re: [Talk-us] Problems with Tiger railroad data

[Alex Mauer]

Russ Nelson wrote:
> Sometimes you have parallel tracks owned by different railroads
> running through the same right-of-way.  Those should definitely not be
> merged.

Sorry, but at some (low) zoom levels it's just not possible to avoid
merging them.  You tell me how to render this clearly:
http://www.informationfreeway.org/?lat=44.51313782651155&lon=-89.55389522707162&zoom=17&layers=B000F000F
at zoom 13 (or zoom 17 for that matter) without merging some of them.

Again, I agree that the tracks in the data itself should not be merged.
 But for rendering purposes, there's just some times that you're not
going to be able to render every track individually. Even with a
single-pixel line per track, with a single pixel line between tracks.

-Alex Mauer "hawke"


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

Re: [Talk-us] Problems with Tiger railroad data

[Alex Mauer]

Peter Miller wrote:
> Also of note in the first example are the two parallel railway lines coming
> from SSE (one a spur and one a rail that are visible in Potlatch in edit
> more). Further to the east there are three tracks running parallel, all
> railway=rail. From the aerial photography there appears to be only one
> railway line (although there may be twin tracks close together but there
> certainly aren't two lines that far apart). I have seen this before and
> deleted the parallel one but was not sure if I was doing the rights thing. 

Probably not.  Railroad tends to be very hard to see from aerial
photography, and TIGER railroad data has been very good, in my
experience much better than the regular road data.

> I
> don't see any OSM tag to identify the number of parallel tracks which would
> be useful rather that running twin tracks very close, or indeed running four
> tracks parallel to each other but possibly this has been discussed in the
> past.

I think we should leave merging of parallel tracks up to the renderer,
rather than removing them from the data.  IMO the data should reflect
what's on the ground.

-Alex Mauer "hawke"


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

Re: [Talk-us] Counties Data Import

[Alex Mauer]

Ian Dees wrote:
> Hi everyone,
> 
> While I was trying to figure out how to divide the massive NHD dataset 
> into more management pieces, I found a county boundary dataset and 
> converted it to OSM.
> 
> I uploaded Wisconsin and Minnesota county boundaries and submit them for 
> your review. One minor issue:
> 
> http://tah.openstreetmap.org/Browse/?x=1036&y=1489&z=12&layer=tile 
> 
> ... shows that each county is a separate, closed way in OSM. This means 
> that at most boundaries, the dashed lines that are used to render 
> political boundaries are overlapping and look odd.

I don't see a problem with this.  I already had done Portage County, WI 
(and its townships) manually though...


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

Re: [Talk-us] braided streets

[Alex Mauer]

Russ Nelson wrote:
> Dave Hansen writes:
>  > The only thing I might do differently is not have the node shared
>  > between the railway and the streets.  That's what I did for the TIGER
>  > upload: created two nodes at the same location.  One for the street, one
>  > for the railway.
> 
> If it's an at-grade crossing, it should be a shared node.  Otherwise
> there should be separate nodes.

This is what I have done as well.

-Alex Mauer "hawke"


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

Re: [Talk-us] Fwd: Maps of State of Wisconsin, USA

[Alex Mauer]

Alan Millar wrote:
> I haven't looked at the Wisconsin data, but I did do a little
> experimentation with PDFs.  If the PDF contains a bitmap, you can extract
> it with any number of tools, but you're stuck with tracing.
> 
> If it contains vector data, you can start to do intersting things.  You
> can convert it to SVG using "pstoedit".  I experimented a little using
> pstoedit to create svg, and then started on a perl script to convert the
> SVG to an OSM file for JOSM.  It could be made to work for small areas
> where map projection would not be a problem.

It appears to contain vector data.


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

Re: [Talk-us] Fwd: Maps of State of Wisconsin, USA

[Alex Mauer]

Ian Dees wrote:
> I'll check to see what the data looks like. Thanks for the forward, Steve.
>

It looks like a lot of useful information can be gleaned from this, 
especially county, township, park, and municipality borders.

I'm not sure if, or how, the maps could be used directly without some 
sort of PDF overlay method in JOSM, or maybe a highly speciallized PDF 
-> OSM converter.

-Alex Mauer "hawke"


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

[Talk-us] county highways: tertiary?

[Alex Mauer]

I've been doing some more driving on rural county highways, and I've 
found that they mostly would fit better as tertiary, rather than 
secondary.  What do others think?

-Alex Mauer "hawke"


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

Re: [Talk-us] "highway" tags in the US

[Alex Mauer]

James Cloos wrote:

> Oh, and don't specify only 65+ MPH highways as equivilent to the
> Brittish Motorway; some states are holding fast to the double nickle

hmm, Interstates are not state highways, so the fact that some state 
highways (for what states?) have a 55-mph speed limit is not really 
relevant.

There are some Interstates which have lower speed limits as well, but I 
think the OSM descriptions basically say: "if it meets the spec, *or* is 
a designated Interstate", then mark it as a motorway.

-Alex Mauer "hawke"


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

Re: [Talk-us] "highway" tags in the US

[Alex Mauer]

Karl Newman wrote:
> On Tue, Mar 4, 2008 at 10:26 AM, Alex Mauer 
> 
> Hmm... Then why are all the generic roads from TIGER marked as residential?

DaveH answered this one.

> 
> Other roads change speed as they enter urban places (e.g. primary roads
> may go from 55 down to 25 to pass through a hamlet)  Why shouldn't
> tertiary be the same?  It's all about the relative status of a road to
> the roads around it.
> 
> But the residential road is slow for its entire length, whereas your 
> hamlet example is only slow while in the town.

Regardless of speed limit (for which we have a separate tag anyway) 
there are some residential/commercial/industrial/rural roads which are 
more significant than the ones around them.  I see no reason to simply 
discard the existence of the tertiary tag just because it's a 
residential area which is being mapped, and because the speed limit is 
the same.  The speed limit is not the only distinguishing factor.

You appear to be suggesting that a road with a 25 mph speed limit cannot 
be tertiary, but a primary road can be 25 mph for part of its length. 
That doesn't make sense to me.

> Hmm...it seems to me that perhaps you're writing this from a perspective
> of "what a driver using the road might expect", while I'm coming from a
> perspective of "how should this road be tagged".  Is that the case?
> 
> 
> "How should this road be tagged" needs to consider the expectations of a 
> driver/bicyclist/pedestrian. I've always been uncomfortable with the "if 
> it looks like a motorway, then tag it as such" for a couple reasons. 
> First, does it meet the standards for a motorway in terms of height 
> restrictions, turn radius, restricted access, etc. 

Yep, that's the question, "does it meet the standards?".  It's a hard 
question to answer at a glance.

> Second, over what 
> distance are you going to consider the road characteristics? Part of my 
> commute route follows a state highway that looks like a motorway in 
> isolated sections--dual carriageway, 60 MPH, restricted access, etc., 
> but it has stoplights at grade intersections every mile or so, sometimes 
> more often. Also, if you continue on the road, it peters out into a 
> 2-lane highway wit h slow tourists and grinds through pedestrian-clogged 
> tourist-trap towns. Not a motorway.

I would consider the characteristics over any portion of the roadway. 
To use your example, and assuming that by "restricted access" you mean 
"bicycles/pedestrians prohibited", I'd tag it as trunk for most of the 
length (the dual carriageway/high speed section) and secondary once it 
drops to two lanes.

There's no reason that a road must be tagged the same throughout its 
length; the state/county -> secondary, US->primary guidelines just give 
a minimum for those administrative classifications.  If a highway meets 
the relevant specifications it should absolutely be tagged as a 
motorway, as long as you know it meets them.  If it doesn't meet them, 
then the physical layout of the road (dual-carriageway, possibly ramps, 
etc.) as well as the speed limit tags and access restrictions give a 
hint to the map user how closely this will resemble a motorway.

-Alex Mauer "hawke"


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

Re: [Talk-us] "highway" tags in the US

[Alex Mauer]

[EMAIL PROTECTED] wrote:
>> Maybe (by way of compromise) we should use two different tags, one that
>> describes the physical characteristics of the road, and one that
>> describes the administrative status of the road?
> 
>> highway=motorway
>> admin=interstate_highway
>> name=Interstate 880
> 
> Yes, excellent idea!  That is a great way to handle both aspects of it.

There's a proposal to do just that:
http://wiki.openstreetmap.org/index.php/Proposed_features/Highway_administrative_and_physical_descriptions


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

Re: [Talk-us] "highway" tags in the US

[Alex Mauer]

Bone Killian wrote:
> On Mon, 2008-03-03 at 19:55 -0600, Alex Mauer wrote:
> 
> IMHO the only thing that qualifies a road as a motorway (in the US) is
> the big red and blue signage identifying it as part of the interstate
> highway system.

Meeting the Interstate highway system standards also qualify it as a 
motorway, for the US.  The sign is, for the most part, an indicator that 
it meets these standards.  Where it doesn't meet standards but is still 
designated, OSM can make an exception just as the US gov't does.

> If we tag things based on what it looks like (subjective) instead of
> what it *is* (objective), the map will be a big mess of different
> colors, without any consistency.
> 
> I think we need an 'official' one-to-one mapping of tags to real
> road-types.  Maybe we should arrange some sort of a vote or something?

One problem is that the signs have little to do with the real road 
types.  One potential solution to this is my proposal here:
http://wiki.openstreetmap.org/index.php/Proposed_features/Highway_administrative_and_physical_descriptions

It separates the signs from the real road types.

Unfortunately, reality is complicated and so are the definitions there.

-Alex Mauer "hawke"


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

Re: [Talk-us] "highway" tags in the US

[Alex Mauer]

Karl Newman wrote:
> 
> Karl Newman wrote:
>  > Okay, for trunk, how about *mainly* ramp access, with grade-level
>  > crossing access to side roads permitted, but without traffic
> controls on
>  > the highway where they meet?
> 
> Hmm, I don't think the ramp access should be a requirement.  Agreed on
> "without traffic controls on the trunk highway at intersections" though.
> 
> I said mainly, not all. Otherwise, why wouldn't you just make it primary 
> in that case?

Dual carriageway, higher speeds.  Built similarly to a motorway, but 
without the restricted access.

> 
> Around here the county roads vary widely in speed limit, anywhere from
> 30 to 55 mph.  They all seem to be two-lanes (total, not per-direction),
> with fairly narrow shoulders.
> 
> 
> My guidelines are thinking about expectations, as well. What you say 
> about county roads basically says that you have to lower your 
> expectations about them, because you don't know what the road is going 
> to be like or how fast you'll be able to travel it.

Right.

>  > So you would place unclassified above tertiary? I haven't spent much
>  > time looking at road classifications when I've been in Europe (I
>  > probably would now since I've joined OSM), but that seems to go
> against
>  > the existing guidelines.
> 
> No...I use tertiary as you describe "missing_tag":
> "residential branch roads which are main roads
> through subdivisions..." though, I don't see why it should be exclusive
> to residential/subdivision areas.
> 
> I noticed now there's also a difference in your usage of unclassified;
> As I use them, these are ordinary, unremarkable roads; if they were in a
> residential area, they'd be highway=residential.  So I put
> "unclassified" down a level from where you do, in between your
> missing_tag and residential.
> 
> Then why should we even distinguish between unclassified and residential?

As far as I know from the history of OSM unclassified and residential 
are basically the same, and they render the same in pretty much every 
renderer.  I believe it's a historical artifact where "residential" 
pretty much meant "unclassified+abutters=residential".

> The first part of your description of unclassified seems to fit there,
> as well: "urban commercial district or rural low-density housing..."
> 
> I suppose we differ in that I feel that tertiary can scale: in a
> residential area, it's the "main roads with fewer driveways", in a
> commercial or industrial area, it's the main roads as well, and in rural
> areas it's main roads which are not county highways.
> 
> I think it's not so good to make the tertiary tag cover all those cases. 
> As you've described it, in a residential neighborhood, that road is 
> generally 25 MPH, in a commercial or industrial area it's usually 35 
> MPH, and in a rural area, 40-55 MPH. I think only last one makes sense.

Other roads change speed as they enter urban places (e.g. primary roads 
may go from 55 down to 25 to pass through a hamlet)  Why shouldn't 
tertiary be the same?  It's all about the relative status of a road to 
the roads around it.

> I would not use "direct driveway access" as a factor for distinguishing
> highways.  Roads of all classifications except motorway may have direct
> driveway access, especially in rural areas.
> 
> It's just a guideline, and it goes toward expectations. Driving down a 
> residential street, I expect to have to watch carefully for kids playing 
> or for drivers backing out. I realize that almost any road can have 
> direct driveway access, but it's very infrequent (more like every 1/4 
> mile at least, not every 100 feet as in a residential neighborhood). On 
> a more heavily-traveled road with a higher speed limit, the driver 
> pulling out has to be more careful than the through traffic.

Hmm...it seems to me that perhaps you're writing this from a perspective 
of "what a driver using the road might expect", while I'm coming from a 
perspective of "how should this road be tagged".  Is that the case?

-Alex Mauer "hawke"


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

Re: [Talk-us] "highway" tags in the US

[Alex Mauer]

Karl Newman wrote:
> Okay, for trunk, how about *mainly* ramp access, with grade-level 
> crossing access to side roads permitted, but without traffic controls on 
> the highway where they meet?

Hmm, I don't think the ramp access should be a requirement.  Agreed on 
"without traffic controls on the trunk highway at intersections" though.

I can think of two highways in my area (central Wisconsin) that fall 
under this; both are state highways.

> Re: County roads as secondary, I was thinking mostly of speed limits as 
> a guideline for the classifications. There aren't a whole lot of 
> numbered county roads here in California, so I don't have much to judge 
> them by.

Around here the county roads vary widely in speed limit, anywhere from 
30 to 55 mph.  They all seem to be two-lanes (total, not per-direction), 
with fairly narrow shoulders.

> So you would place unclassified above tertiary? I haven't spent much 
> time looking at road classifications when I've been in Europe (I 
> probably would now since I've joined OSM), but that seems to go against 
> the existing guidelines.

No...I use tertiary as you describe "missing_tag":
"residential branch roads which are main roads
through subdivisions..." though, I don't see why it should be exclusive 
to residential/subdivision areas.

I noticed now there's also a difference in your usage of unclassified; 
As I use them, these are ordinary, unremarkable roads; if they were in a 
residential area, they'd be highway=residential.  So I put 
"unclassified" down a level from where you do, in between your 
missing_tag and residential.

The first part of your description of unclassified seems to fit there, 
as well: "urban commercial district or rural low-density housing..."

I suppose we differ in that I feel that tertiary can scale: in a 
residential area, it's the "main roads with fewer driveways", in a 
commercial or industrial area, it's the main roads as well, and in rural 
areas it's main roads which are not county highways.

I would not use "direct driveway access" as a factor for distinguishing 
highways.  Roads of all classifications except motorway may have direct 
driveway access, especially in rural areas.

-Alex Mauer "hawke"


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

Re: [Talk-us] "highway" tags in the US

[Alex Mauer]

Karl Newman wrote:
>
> Agreed. The criteria listed on the Wiki page promote too many highways
> to "motorways". It's too hard to distinguish between them; in dense
> urban areas you could end up with a lot of "motorways". It seems to me
> the "motorway" tag should be reserved for interstates, with some
> exceptions for major US highways. You left out "tertiary" from your
> descriptions. I would see "tertiary" as an important thoroughfare road
> through a town--higher speeds and less traffic controls than
> "unclassified". How about these guidelines, based on speed limits and
> lanes:
>
> * motorway: Interstate, 2+ travel lanes, ramp access only, speed
>   limit 65 MPH+
> * trunk: US highway, 2+ travel lanes, ramp access only, speed limit
>   60-70 MPH
> * primary: US highway, 1-2 travel lanes, or State highway, 2 travel
>   lanes, speed limit 55-65 MPH, can have occasional
>   stoplights/traffic controls
> * secondary: State highway, 1-2 travel lanes, or larger county
>   highway, speed limit 45-55 MPH
> * tertiary: County highway, other unnumbered thoroughfare, speed
>   limit 40-50 MPH
> * unclassified: urban commercial district or rural low-density
>   housing, normally no direct driveway access to housing in urban or
>   suburban areas, speed limit 30-40 MPH
> * missing_tag: It seems like there needs to be another
>   classification for residential branch roads which are main roads
>   through subdivisions but still have direct driveway access to
>   housing.
> * residential: urban or suburban roads primarily for providing
>   access to housing, speed limit 15-25 MPH
>

I agree, with the modification that trunk doesn't need to be ramp access 
only, and that county highways are secondary.

I've used tertiary for the missing_tag you describe, as this seems to be 
in line with the European tertiary roads.  (these comments are also on 
the wiki, I believe.

-Alex Mauer "hawke"


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

Re: [Talk-us] "highway" tags in the US

[Alex Mauer]

[EMAIL PROTECTED] wrote:
>>> For instance, OR217 and US26 west of Portland are both divided,
>>> multi-lane, limited access highways.  Despite not being interstates,
>>> they *ARE* motorways.
>> Hmm, are you sure?  maybe they just look like motorways; there are a lot
>> of almost-motorway roads
>
> Hmm.  I'm confused by that.  I don't quite understand what it would mean
> for something to just look like a motorway but not be one.
>
> As far as I can tell, this is an open community project, and terms have to
> be defined by the group to be useful.  There is no US Department of
> Transportation statement declaring what kind of road qualifies as a
> "motorway", nor what kind doesn't.

No, but there's a standard for Interstates [1], which we've taken to be 
equivalent to the extent that there's not a need for a separate tag for 
UK motorways vs. US Interstates (vs. the Autobahn, etc.)

> So we need to muddle through it
> ourselves.
>
> There are an awful lot of people in the US who recognize what a freeway
> is, regardless of whether they know which bureaucrat signs the check for
> the little bumpy dots between the lanes.
>
> If it looks like a duck, walks like a duck, and quacks like a duck, we
> should call it a duck.  If it is high-speed multi-lane restricted access
> way with cloverleaf entrances, we should call it a motorway.

What about other types of interchange, e.g. diamond, parclo, etc.[2]?

Regardless
> of whether it is part of the US Interstate Highway system or not.

Here I agree.  My only point of contention is what qualifies as 
duck-quacking for our purposes...certainly a highway that meets the 
standards without being designated should be tagged as motorway, but as 
the link should make clear, it's not all obvious whether a given piece 
of the road meets those standards.

> I hope I don't sound confrontational.  Are we saying the same thing?  I've
> definitely been pondering this myself as to how to classify or judge these
> tags.  Good discussion.

I hope the same for myself.  I'd really like to get this hashed out, 
just so the group can come to some agreement on an unambiguous 
definition for a motorway.  If that means "limited access, minimum 
median width, minimum lane width, minimum vertical/horizontal 
clearance", even if this means that some non-standard roads are marked 
as motorways, I don't have a problem with it -- as long as there's 
agreement on what exactly qualifies.

[1] http://en.wikipedia.org/wiki/Interstate_Highway_standards describes 
it, and references standards documents regarding it.

[2] http://en.wikipedia.org/wiki/Interchange_%28road%29

-Alex Mauer "hawke"


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

Re: [Talk-us] "highway" tags in the US

[Alex Mauer]

Dave Hansen wrote:
> I'm pretty sure I know one when I see one these days.
>
> http://wiki.openstreetmap.org/index.php/Map_Features
>> A restricted access major divided highway, normally with 2 or more
>> running lanes plus emergency hard shoulder. Equivalent to the Freeway,
>> Autobahn etc..
>
> Bingo.  It's truly restricted access.  No access except from onramps.
> It's divided.  It also has emergency shoulders.  Is is effectively an
> interstate.

There are many roads that fit those definitions, but have, for example, 
too-sharp curves, too-low bridges, perhaps too-narrow shoulders, etc. to 
meet the legal definition of an interstate.

Of course, it's very possible that they've simply not been designated as 
Interstate, but my point is that it's not easy to tell at a glance.

-Alex Mauer "hawke"


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

Re: [Talk-us] "highway" tags in the US

[Alex Mauer]

Dave Hansen wrote:

> For instance, OR217 and US26 west of Portland are both divided,
> multi-lane, limited access highways.  Despite not being interstates,
> they *ARE* motorways.  

Hmm, are you sure?  maybe they just look like motorways; there are a lot 
of almost-motorway roads

> 
> As for primary/secondary/tertiary.  I basically think that primaries are
> those really big roads that happen once every mile or two.  They're the
> ones where WalMart is. :)

I think those are secondary at best, especially in very urbanized areas.

If we want to tag the physical attributes, something more like 
http://wiki.openstreetmap.org/index.php/Proposed_features/Highway_administrative_and_physical_descriptions
 
is more appropriate.


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

Re: [Talk-us] "highway" tags in the US

[Alex Mauer]

Karl Newman wrote:
> http://wiki.openstreetmap.org/index.php/United_States_roads_tagging or 

This one can be disregarded, I wrote it as one of the first guides for 
the purpose.  Perhaps someone should update it to match the others.

> here: http://wiki.openstreetmap.org/index.php/Mapping/Features/Road or 

This is newer, also initially written by me, and reflects how I tag as 
well as the method used on the TIGER import.

> here: http://wiki.openstreetmap.org/index.php/Key:highway or here: 

This one is fairly consistent with that as well.

> http://wiki.openstreetmap.org/index.php/Highway_tag_usage or here: 

This one includes the fantasy that "the highway tag represents physical 
attributes", so I'd be disinclined to follow it.  However, below that 
(In the "International equivalence" section) the portion on the US is 
consistent with the others.

> http://wiki.openstreetmap.org/index.php/Talk:Highway_tag_usage (that's 
> where I put my proposal, username Si liconFiend)

This is consistent with the others as well.

In conclusion, the first link is crap, and all the others are fine.

-Alex Mauer "hawke"


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

Re: [Talk-us] Islands in Lakes/Rivers

[Alex Mauer]

Ian Dees wrote:
> Thanks Bone. I tried tagging the island as natural=land to no avail. Any 
> other ideas?

use a multipolygon relation between the river and its islands?

-Alex Mauer "hawke"


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

Re: [Talk-us] How accurate is Tiger data vs Yahoo Aerial?

[Alex Mauer]

[EMAIL PROTECTED] wrote:
> Any advice or suggestions?  Thanks.

In my area, the Yahoo! photos are wildly inaccurate (shifted east by at 
least 30 meters).

TIGER data is highly variable, getting some roads perfectly correct in 
some places, being completely wrong in others, missing old roads in 
others, and just being out of date in still others.

IMO, a GPS track (or better yet, several)is the only way to fly with 
respect to having a sane base reference.  Once you have that, you can 
align the yahoo images to it and work from there.  Unfortunately, the 
alignment is not retained, and for JOSM, newly downloaded chunks of 
yahoo imagery have to be painstakingly realigned with the existing data, 
which is quite hard to do consistently over large areas.  For this 
reason, I try to only use yahoo data for things that I can't map with a 
GPS (bodies of water primarily)

-Alex Mauer "hawke"


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