[OSM-talk] Highway tags and other junk

[guy at graviles-reynolds.org]

Quoting Ben Robbins <ben_robbins_ at hotmail.com>:

> >Travelling around the globe, I have driven on many roads which have been 
> >designated as 'Trunk Roads' but which ranged from 8 lane highways with hard
> 
> >shoulders, through dual carriage ways, single lane roads with passing 
> >points and on down to a line of poles marking a route across a desert.
> 
> Local definitions of the word Trunk dont dictate how something is to be 
> tagged in osm.  That is local thinking!.  If a road thats called a trunk is
> 
> a small single windy road then it should be tagged with the appropriate osm
> 
> tag.
The roads all still fall within the OSM definintion of a trunk road: "Important 
roads that aren't motorways. Typically maintained by central, not local 
government. Need not necessarily be a divided highway. In the UK, many 
2digit "A" roads (eg A34) were formally designated Trunk roads although that 
definition is no longer absolute...". 

It maybe a small and windy single lane unpaved road with passing points but it 
is still a trunk road by the definition. Thus the highway key will have the 
value trunk, and physical actual definition of road will be in the property 
key/value pairs. To do otherwise is to enforce UK/western centric thinking that 
assumes that all trunk roads have at least two lanes and are paved, and results 
in large chunks of the world being mapped as little more than tracks. 

For the purposes of navigation we neeed to keep to the political/usage highway 
types, regardless of the physical characteristic of the roads as they give us a 
guide to priority in which we should be using the roads, particulalry since 
they are often based on local traffic management schemes. I will sight Baldock 
again here, where following the bypass two roads; the A505 (primary route) and 
the A6141(trunk route), have been downgraded to to seconadary routes; B656 and 
B197 with no diminution in there physical characteristics. It is then 
underlying properties of the route that give the guidance to the 
navigator/navigation software as to the average speeds that can be attained 
along each section of the way. 

> 
> As I said in one of the posts above, Im not suggesting people cant add 
> speicifc tags if nessesery, but from much researching I have found that 
> every track can fall into one of these types.   After all We all live in the
> 
> same universe with the same properties.  Please read threw the wiki, and 
> mailing list again.

I have read through the tracktype proposal again and I am still of the view 
that it is centred in a region of the UK and potentially at the point in the 
farming calender that the survey was taken. The definitions refer to hedges and 
grass; hedges are not universaly used as boundaries within the UK let alone in 
other parts of the world, and whilst grass is common in the UK it is not 
elsewhere. Also the difference between a Grade 4 and a Grade 5 track may only 
be seasonal, a track which was a Grade 4 in the early part of the year when the 
weather is damp and there is alot of activity in the fields may by the end of 
the summer be a Grade 5. 

As real life example take the road between Cabanataun City and Gabaldon, in 
the Philippines for a large part of its length it is a 'primary' route, both in 
terms of the local and OSM definition. However for many years until it was 
fully paved, and recently after several thyphoons in many places it was/is 
little more than a rutted track. So the question is do you tag it using the  
political/usage key/value pair 'highway=primary' with the property key/value 
pair 'surface=unpaved' or do you tag it as 'highway=track' with in your 
tag 'tracktype=grade4'? Though grade4 does not fit becuase the road is in 
regular use by cars, and is bounded by irrigation ditches not hedges, but it 
cannot be a grade1,2,3 because it is neither paved nor does it contain 
gravel/hardcore and it can't be a grade5 because outside of the rainy season it 
is to hot and dry for grass to grow, and if any did it would be worn away by 
the traffic.

> You seem to think Im arguing the exact oposite of what Ive said in all 
> aspects.  Please read again, and dont send rather aggressive emails without
> 
> it being constructive.   There really is no point in a flame war unless what
> 
> is said is 1) fully read, and 2) fully understood, preir to debate, in wich
> 
> the outcome is garanteed to be productive.
> 

My intention is not to start a flame war (it is probably just my writing style) 
but to take us back to basics, in a global context.

Guy