[Talk-us] Mappy Hour / Many Mappy Minutes

2017-09-27 Thread Martijn van Exel
Hi all,

We had our first virtual mappy hour tonight and it was a lot of fun to chat
with fellow mappers. I wrote a short diary post about it, also announcing
the next edition on November 1:
https://www.openstreetmap.org/user/mvexel/diary/42393

Martijn
___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us


Re: [Talk-us] Missing imagery in JOSM

2017-09-27 Thread Mark Wagner
On Wed, 27 Sep 2017 12:14:03 -0400
"Mark Bradley"  wrote:

> What happened to the USGS topographic map imagery that was available
> in JOSM until a few days ago?

Still there for me, though something's changed: when you zoom out, it
no longer switches away from the 1:24000 maps, but instead displays
them scaled down.

-- 
Mark

___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us


[Talk-us] Missing imagery in JOSM

2017-09-27 Thread Mark Bradley
What happened to the USGS topographic map imagery that was available in JOSM
until a few days ago?

 

Mark

 

___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us


Re: [Talk-us] I 85 Express Lane (Atlanta, Georgia)

2017-09-27 Thread Jack Burke
> On Sep 24, 2017, at 5:22 PM, Minh Nguyen 
wrote:

>
> I might be inclined to map the lane as a separate way, but only because I
don't know of any routers that currently
> recognize the change:lanes tag. [1] Either way, it makes sense to treat
HOV and toll lanes similarly.
>
> [1] https://wiki.openstreetmap.org/wiki/Key:change
>


As to treating the two types the same, I concur.  The specific type of lane
is immaterial; that they have different access restrictions than normal
lanes is what is important.  As to mapping them separately, I'm...still
strugging with that (mentally, not in terms of making the edits).  We
already map some types of access restrictions without drawing separate
ways; e.g., hgv:lanes=no|yes|yes for sections of highway where big rigs are
not supposed to use the far-left lane.  On the other hand, we do it that
way because there isn't any legal restriction on vehicles moving between
lanes, like there is with the double-solid-stripe separated HOV/toll
lanes.  Using hov:lanes=blah|blah|blah is actually documented, whereas
toll:lanes= is not, yet as far as I know, not a single router supports HOV
tags in any form, which is a problem in Atlanta, because we have some
HOV-only freeway entrance and exit ramps.  I guess that's why some wiki
pages suggest doing access=no and hov=yes for such link roads.


On Sun, Sep 24, 2017 at 8:47 PM, Tod Fitch  wrote:

> I strongly recommend mapping them as one way with the hov:lanes=* and
change:lanes=* showing the restrictions and
> where you can transition between the lanes. That is the best tagging I
know of to accurately reflect the actual configuration.

> The reason I feel strongly about it is that many miles of HOV lanes in my
are were mapped as separate lanes (done
> before I moved to the area). And now CalTrans is repainting those to
allow entry and exit at anyplace (change from
> a variety of old paint styles to broad dashed white striping). If the HOV
lanes had been tagged as a single way with
> change:lanes showing the restrictions on moving to and from the non-HOV
lanes it would have been trivial to change
> the mapping to conform to the current paint. Mapped as separate ways you
have to go back and remove the separate ways
> then the paint changes which is a pain.

I understand your point; I'm just not sure if "difficulty in making changes
to match what the DOT does later" is reason enough to draw it more simply.


> When there is a solid line (I’ve seen white, yellow and a variety of
parallel white/yellow versions of the “don’t change
> lanes here striping), there are generally designated areas for
transitioning between HOV and non-HOV lanes. If you map
> the HOV as a separate way then you need to add any number of virtual
cross over ways.

OR merge the HOV lane into the main road for the length of the transition
area, and start a new separated road where the transition section ends,
which is what I was considering doing if I try to draw them separately,
because as you point out, the crossover ways become a problem.


> I wondered why OsmAnd and Maps.me
> were always telling me to do slight rights or slight lefts in those areas
until I looked at the mapping and saw the
> separate HOV lanes with virtual link ways connecting it to the non-HOV
lanes. In essence putting in separate HOV ways
> where they don’t actually exist doesn’t always help the router.

I agree that it can be confusing when a router gives seemingly spurrious
and unnecessary directions like that; it seems to be the proximity of the
separate way to the main road that is the problem, from what I can tell.
Although OsmAnd seems to use the smoothness (or lack thereof) in road bends
to make its slight left/right direction statements; I know of some roads
where there are no turns, side roads, exits, or anything else to leave the
main road where it still says slight left/right, simply because the angle
of the way as it crosses a node appears to be over some configured number
of degrees.


Let me throw an additional scenario into the discussion.  On the Veterans
Expressway toll road in Tampa, they are adding some new separated express
lanes to the road.  The "barrier" is both a double-solid-stripe *and* some
orange plastic vertical bollards:

https://www.mapillary.com/map/im/i98UEIXDDgAfhhqzMxD1IA

Because there is an actual physical barrier, albeit a "soft" one that cars
can easily drive through, I think this one _has to_ be mapped separately.
But I'm open to discussion on it.


> FWIW, the JOSM plug-in that deals with change:lanes shows things nicely.
And I suspect that routers like OsmAnd
> will be supporting it soon (if not already) as they support turn:lanes
pretty well.

change:lanes is one of the most confusing tagging schemes I've seen yet.
 *sigh*  I guess this means I actually have to start trying to figure it
out.


> Extending this somewhat: I’ve traditionally mapped on and off ramps with
the link leaving/joining the