[OSM-talk] Another argument for Open Street Data ?

[Richard Fairhurst]

D Tucny wrote:

> What I noticed in that story, different to other 'satnav ruined my life'
> type stories, is that there is no claim that anyone is doing anything
> wrong... No mention of signs being ignored because the satnav told them
> to... no mention of people doing overtly stupid things (well, except for
> hitting buildings trying to get big trucks through gaps that are too
> small)... So, through deduction, it would seem that there is a valid,
> shorter route through these villages, it may not be suitable for
> wide/long/heavy vehicles, but, it's not been signposted as such and the
> councils responsible for such signage are blaming satnav for people choosing
> a shorter route that they haven't bothered to sign as being unsuitable...

Up to a point.

I know the area passingly. There's no way that, using "traditional"  
navigation techniques, you would choose to drive an HGV via Llangadog.  
On a printed map, the A4069 is marked as red with countless  
switchbacks: it's immediately apparent that it's unlikely to be  
suitable for lorries. "The knowledge", passed on orally from one  
driver to another, would communicate the same information. But people  
blindly trust satnavs even though they have neither of these benefits.

(Here's the road in question on OSM:  
http://www.openstreetmap.org/?lat=51.8795&lon=-3.8782&zoom=12&layers=B0F .  
Interestingly, it's been tagged as "secondary" though it should  
actually be "primary", perhaps indicating the perceived low quality of  
the route.)

You can't signpost everything - it's a problem we have on the UK  
waterway system where the risk-averse culture has led to a profusion  
of ugly-looking signs warning anglers not to cast off there due to  
electricity lines overhead. If people use common sense there's no need  
for those signs. Conversely, if you _do_ signpost at every  
opportunity, people stop thinking for themselves.

> This isn't going to be something we can really easily fix either

Hm, maybe not easily, but we've got a better chance than a software  
author using pure TeleAtlas/Navteq data - because we have real GPS  
tracklog data from which we can extrapolate driving times, delays and  
problems.

cheers
Richard