[Talk-is] footpaths vs bikeways and enthusiastic marking

[Ævar Arnfjörð Bjarmason]

On Wed, Nov 19, 2008 at 1:00 AM, Karl Palsson <tweak at tweak.net.au> wrote:
> How come some parts of the foothpath by sæbraut are marked
>
> highway:cycleway
>
> but others are marked, highway:footpath;bicyle:yes ?
>
> Is anyone making any distinction?  Some footpaths are marked footpath, some
> cycleway.

You brought up some of the points in this E-Mail before at
http://wiki.openstreetmap.org/wiki/Talk:WikiProject_Iceland/Cycleways
and I replied to the best of my ability then. Basically we're using
the schema described at
http://wiki.openstreetmap.org/wiki/Talk:WikiProject_Iceland/Cycleways#Examples

In the case of the Sæbraut cycleway suddenly turning into a footway
that's because at that point it ceases being of the quality one would
expect of most of the cycle network and turns into a pedestrian
sidewalk.

> Why have we marked a footway on one side of snorrabraut, but not the other?

Only one of them has been surveyed, presumably. And if I recall the
majority of the footpath for the south of sæbraut is nonexistent.

> Why are we marking normal footpaths at all?

Because they're features on the ground and as such should end up in
OSM eventually, how they should do that is open to discussion, maybe
the main way they're besides should have tags to indicate that they
have an accompanying footpath, but then we'd have to split up the way
each time the footpath stops and starts, and there would be no decent
way of conveying changes in the distance of the path from the road,
junctions as they relate to footpaths (they often have their own
traffic control) etc.

> If this was an area where cycle access
> was not expected, I'd understand, but it's a residential area, and it's just a
> normal footpath.

Users of the OSM data can't make decent any assumptions like that when
trying to route through the data, and I think gosmore which is the
primary routing software for OSM presumes that you can't cycle on
footways unless it has bicycle=yes.

> Likewise for marking a footpath on the park side of
> Rauðarárstígur.  There's one on both sides, and they're both just regular
> footpaths.  There's also one on every street in that area except
> Skarphéðinsgata.  If you're going to start marking footpaths, you should be
> consistent, otherwise how can anyone know what the map really shows?

No you can't be sure what the map shows you wrt. footpaths from
looking at it now, but I can't see how this problem can be solved by
anybody. The map will always be a work in progress for *all* data,
whether it's normal roads (where we have plenty missing), footpaths or
something else.

> On the subject of enthusiastic marking,  marking the police station car park as
> "amenity:parking" is just outright lying.

No, amenity=parking should be used for any sort of car park, at least
according to the wiki as I understand it. It is a private parking are
however so it should have had access=private to go along with it, I've
fixed that.

> Likewise marking the entire area of a petrol station as amenity:parking.  It's not a parking area!

Do you mean this petrol station?: http://openstreetmap.org/browse/way/25867042

Some of it is a parking area, but not all of it as the map would
suggest. I've marked some of these more diligently though, like the
park area in Öskjuhlíð where I surveyed the parking area and the
greater paved are it was contained in separately:
http://openstreetmap.org/browse/way/26530917

Or do you mean that amenity=parking should only be used for public
parking spaces and not privately owned space or parkin space with
access or time limitations? If so I can't agree with you, it's still a
parking space and any access restrictions should be expressed via
further tagging.

> Do we really need bounding boxes drawn around "suburbs" ?  What's wrong with a
> point marker in the middle of the area?

Yes, suburbs should have areas, since a suburb is an area and not a
point. Just a point marker would tell you nothing about where a suburb
begins or ends.

> There's no official street boundaries to my knowledge, so marking them as if there were is disengenous.

There are no official suburbs at all in Reykjavík, only administrative
regions. Suburbs aren't meant to be official or administrative
however, but a historical segregation of the city into smaller areas
for whatever reason.

> (Miklatún is _not_ in Hlíðar?  says who?  Birkihlíð, reynihlíð, viðihlíð, lerkihlíð,
> beykihlíð are not in Hlíðar either?  Says who?

I was including the streets that ended in "hlíð" when I mapped it,
obviously I missed some.





> While we're on the subject of Hlíðar, someone's connected Bólstaðarhlíð all the
> way round again.  Have they removed the concrete barricades there?  That was not
> an accident, that road was deliberately unjoined.  If the road has been opened
> up again by the city, cool :)  but the concrete barricades looked fairly
> permanent :)
>
>
> Still, I like how much extra stuff is on the map these days.  :)  (I'd
> personally rather see all the roads filled in, properly linked and named, for
> routing purposes, before seeing fire hydrants and shop names in skeifan, but
> each to their own)