[OSM-talk] Observed vs. official designation [was: Oxford High Street classification (Was: Bus sluice/gate)]

[Mike Collinson]

At 11:53 AM 2/12/2007, David Earl wrote:
>On 01/12/2007 10:01, Gervase Markham wrote:
>> But the open question is, should 
>> highway tags be defined for the benefit of routing algorithms and/or map 
>> rendering (one possible view) or should they be defined according to 
>> legal classifications, and additional auxiliary clarification tags used 
>> in the occasional situation where the legal classification would mislead 
>> rendering/routing software.
>> 
>> Granted, I'm new to OSM, but I'd argue for the latter.
>
>I disagree.
>
>(1) It is singularly unhelpful for a user of a map to find that the 
>situation on the ground is different from what the map says. It is 
>unlikely OSM would find its primary market in officialdom where such 
>distinctions do matter.
>
>(2) It is very hard to get the "official" information. Most of us only 
>have what we see in front of us to work from. It is impractical.
>
>(3) The vast amount of data already there is tagged by observation. 
>While no doubt we could change a certain amount automatically, there 
>would always be loose ends - especially for the cases affected by the 
>reason the change was being made. And even having changed, some people 
>will continue to tag things the way they want rather than following a 
>consensus (though that already applies the other way round too).
>
>Having said that, I see no problem in having some additional information 
>which gives the "official" designation if different. For example - 
>"ref:official=A120" or "highway:official=primary".
>
>David

I absolutely agree with David from an international perspective.  The hang-up with legal definitions of roads seem a peculiarly British preoccupation :-).  In most place I've been this information does not exist or is not well known.  A road is primary 'cos it looks like one and can be used as one. The roadsign, if it exists, may be pink or blue according to the predilictions of the local mayor or what decade it was put up in. 

And I'd reiterate that the general OSM principle of having a general easy to tag, easy to use tag at the expense of rigorous preciseness is a good workable one. The example I use is waterways. It is easy to look and judge whether it is a stream or a river within local context.  It is not easy to measure the depth or come back in 6 months and check if it is still there.  But canoeists, boat owners, fishermen need other information and these can be provided as precise specialist tags over time. Same with roads.

Mike