On Sat, 6 Jul 2019 at 12:42, Mateusz Konieczny <matkoni...@tutanota.com>
wrote:

It is not feasible to do for a typical mapper to record "light level in
> lux".
>

Sadly, however, it is the only objective way of specifying the light
level.  And even then, it's
easy to do it wrong if you don't account for the angle of incidence.  Aim
the sensor at the
light and you'll get a higher reading than if you lay it flat on the ground.

but it is not helping with problem what would be a good threshold between
> lit=yes and lit=no
> on footways
>

In the UK, BS 5266 requires a minimum illumination of 1 lux along the
centre line of
escape routes.  Up until 2011, BS 5266 required a minimum of 0.2 lux along
the centre
line of escape routes.

IIRC, the 0.2 lux figure was approximately equivalent to the light of the
full moon, which
was deemed adequate for some work in shipyards during WW II.  Certainly
it's more
than necessary to take an unhurried stroll along a footpath after your eyes
have
acclimatized to the darkness.  It may be the minimum that was once
considered
adequate to allow evacuation from a dark building after a power cut (no
time for eyes
to acclimatize) but it's not the minimum needed to follow a reasonable
footpath if you're
not in a hurry.

So it is preferable that everyone has their own definition of what is
> lit=no/yes and
> recommend that "in case of doubt is it lit=yes or lit=no feel free to
> choose either"?
>

It's hard for most mappers to accurately measure.  It's hard to agree on a
suitable
figure, because the amount of light necessary depends very much on the
nature of the
path (an asphalted footpath can be safely followed with far less light than
is required for
an unmade path over rocky terrain).  OTOH, if there are lights along the
path, it is clear that it is lit.
If it's the sidewalk of a lit road, it's lit.  In any other cases, it's
probably safer to say it's unlit.

-- 
Paul
_______________________________________________
Tagging mailing list
Tagging@openstreetmap.org
https://lists.openstreetmap.org/listinfo/tagging

Reply via email to