[Tagging] access=avoid

[Felix Hartmann]

On 15.06.2011 16:29, Pieren wrote:
> On Wed, Jun 15, 2011 at 3:59 PM, David Murn <davey at incanberra.com.au 
> <mailto:davey at incanberra.com.au>> wrote:
>
>
>     Out of interest, as tracktype=gradeX is subjective also, should we
>     avoid
>     using that too?
>
>
> A scale is not necessarily subjective if it is well defined. The key 
> "sac_scale" is a good example.
Dooooh, the sac_scale is like the worst example of how a good scale 
turned completely useless in OSM.
The scale itself is pristine yes, but then 95% of the users have NO CLUE 
at all about what the description means. They don't even know what 
alpine means for a start...
Hence in my eyes about 90% of all sac_scale=T3 or higher is wrongly 
graded (in 99% of all cases real difficulty is at least one value lower).
Of course JOSM Presets are mainly responsible for the sac_scale mess, 
but then if the sac_scale had been entered with T1 to T6 as values 
instead of the subjective descriptors, the situation would be much 
better (cause people would be forced to at least read once the sac_scale 
and what it means, though I would still doubt a majority could 
understand it).
The sac_scale bull**it is only topped in OSM by the completly unusable 
visibility usage (before I added pictures to describe the visibility, 
the situation was even worse, in so far that you basically needed to be 
an alpineer to know what the description meant, and still many people 
misuse it as they think it is about route markings and not the 
visibility of the way itself).

So even if there are well established scales (like both sac_scale and 
the visibility scale originating from the Swiss Alpine Club) - if the 
wiki and editors get it wrong, the outcome will be utter rubbish (as 
previously the sac_scale was only used by alpineers or authors who did 
indeed study the scale well, so the description did not need to cater 
for the casual mountain tourist or even worse OSM mapper who never sets 
foot into alpine regions, but wants to put them into OSM).

I actually intended to put an effort to have tags to describe everything 
topological that you can see in a Top10 Swiss Raster Map, but after 
thinking about it for a while, I decided it's not worth the effort, as 
even stripped down entering precise topological alpine info (ridges, 
chutes, ....) will not work with only a tiny minority inside OSM that is 
able to actually understand 90% of what such a map carries in 
information and able to interprete it. (e.g. if a mountain ridge is 
drawn inprecise into OSM, the usefulness tends to 0, as one cannot rely 
on it to assess the avalanche exposure of a certain place). So If I ever 
make that effort, it will be outside of OSM with a compatible license, 
but making sure that the people entering that data, are able to assess 
the consequences of wrong information (I rather have no topological info 
at all, than 95% correct (as the last 5% that are wrong are rendering 
the 95% that are correct to rubbish) -- that would be like hosting OSM 
on a Server with 5% downtime, simply a showstopper when those 5% of time 
happen when most users are online (which is likely for hosting), so 95% 
uptime would mean like 50% of all users are affected by downtime)

Also I see smoothness as in general of higher quality than tracktype. 
However it's great we have both of them. This enables crosschecking and 
filtering out wrong entries. Both gradings are really subjective, but in 
general of good quality.
Class:bicycle may be complicated, but it's only needed for borderline 
cases (where the normal information one can interprete from looking at 
other tags does not suffice), so actually taking it into account does 
improve autorouting quite substantially for cyclists. The fact that for 
90% of users the tag is too complicated actually improves the 
qualitative outcome. And as detailed above, for many things in OSM 
quantity means nothing, if quality is poor.
> "tracktype" would probably require a better description but it is, 
> imho, less subjective than the "smoothness" values.
> Coming back to the OP, a dangerousness scale for bikers has been many 
> times discussed in my country. The problem is that even outside OSM 
> itself, there is no real standard and many local initiatives (at 
> municipalities level) do not have a common agreement about such scale. 
> But it is clear for me that access=avoid is a mistake and should be 
> replaced by the legal access value. If it is not clear in the wiki 
> doc, then improve the wiki doc.
>
> Pieren
>
>
> _______________________________________________
> Tagging mailing list
> Tagging at openstreetmap.org
> http://lists.openstreetmap.org/listinfo/tagging
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/tagging/attachments/20110616/f73322d3/attachment-0001.html>