OpenStreetMap logo OpenStreetMap

Changeset Key Mışewre
164278185 4 months ago

PS The CyclOSM layer on OSM website or at https://www.cyclosm.org/#map=14/53.2737/-6.2206/cyclosm are a handy way to see the cycle infrastructure that has been mapped.

164278185 4 months ago

Hi, welcome to OSM!

Many of the cycle lanes you added were already mapped using the cycleway tag on the road.

There are two current approaches to mapping cycle paths: osm.wiki/Key:cycleway and osm.wiki/Tag:highway%3Dcycleway
Please read those pages to understand their intended use, and pros/cons of each approach. Obviously choose whichever approach you feel is best for the situation at hand, but understand the meaning of each, and also that applying both approaches, like has inadvertently happened here results in duplication.

Also, many of the cycle lane/track lines you've drawn (e.g. osm.org/way/1372763000) aren't continuous cycle lane/tracks in reality - there is actually no cycle lane at all for part of Sandyford Road.
The cycleway tags already added to the road reflects this intermittent and varying nature of the cycle lane/track, e.g osm.org/way/108345245

It's unfortunate that the iD editor doesn't give a visual indication of cycleway presence on a road with cycleway tags on it. I made the exact same mistake when I first contributed to OSM.

I hope this all makes sense!

154951060 12 months ago

Re. your fixme. It is effectively one-way due to how narrow it is, and due to the turn restrictions at the western junction which prevent traffic from entering it at that end.
However, there are no road markings or signs to stop someone from driving westwards on this section of road, then parking at the side of the road and later heading back eastbound.
So in theory it is a two-way street, but it's practically one-way.

155135110 12 months ago

This seems to have resulted in stray house tags being added to nodes, in addition to areas (e.g. at the corners of a house).

Also FYI the wiki page for mapping houses recommends "house=semi-detached" or "building=semidetached_house", but I don't see any mention of "house=semi_detached". I could be out of the loop, I'm assuming the wiki is up to date on this.
osm.wiki/Key:house

150814677 about 1 year ago

Ok, in that case I wouldn't tag as highway=construction because navigation software will treat the road as closed and unusable. You could map the affected road as being one-way alternating, and map a separate footpath with highway=construction. However you should consider how long the works are going to last. I read somewhere that road works that will last less than 3 months should not be mapped, and started doing same. I don't think there's a hard rule around this, but the main factor for me is that some navigation apps might not update their OSM data for months, which can have the effect of the app avoiding a road that has since re-opened. In this case since its basically a layout change and not a closure, there will probably be little/no impact on routing choices for navigation apps so it's probably OK to map it here IMO.

150814677 about 1 year ago

Hey, sorry for not contacting you about this. I assumed the road is still open to two-way traffic as I saw some housing estates that are only reachable via that road. Is the road completely blocked?

150726173 about 1 year ago

Thanks!

150695112 over 1 year ago

This looks like the wrong location to add this park. Please re-check.

143911428 over 1 year ago

That's great, thanks for that

148921583 over 1 year ago

Thanks for your understanding!

143911428 over 1 year ago

Thanks for your understanding!

143911428 over 1 year ago

Hey,

In relation to the cycle lane that you added as a separately mapped way:

Please pay attention to the cycle lanes/tracks that have already been added by others using cycleway tags on the way that is the centreline of the roadway. e.g. at Stillorgan Road (osm.org/way/481561458)

For example, there is already
cycleway:left=track
...along a part of this road where you have added a separate cycleway. Your addition conflicts with these tags. This conflict should be resolved somehow.

Accordingly, I want to point out the guidance given on the OSM wiki at this page osm.wiki/Tag:highway%3Dcycleway#When_not_to_use and the various other pages which that page links to, in case you weren't aware.

Thanks

148921583 over 1 year ago

Hey,

In relation to the cycle tracks that you added as separately mapped ways:

Please pay attention to the cycle lanes/tracks that have already been added by others using cycleway tags on the way that is the centreline of the roadway. e.g. at Constitution Hill (osm.org/way/370381577)

There is already
cycleway:both=lane
cycleway:both:lane=exclusive
...along a part where you have added a separate cycleway. Your addition conflicts with these tags. This conflict should be resolved somehow.

Accordingly, I want to point out the guidance given on the OSM wiki at this page osm.wiki/Tag:highway%3Dcycleway#When_not_to_use and the various other pages which that page links to, in case you weren't aware.

Thanks

141519016 almost 2 years ago

The northbound cycle lane here is not really redundant when used in combination with the cycleway:right=separate tag, because "separate" has a special meaning unlike the other values of the cycleway tag.
If the value was something like cycleway:right=lane, then yes,that would be redundant if there was already a separately mapped cycle lane.

When the value of the cycleway tag is "separate" that means the cycle lane/track was drawn as a separate way. It's not a requirement to put that value in when mapping a cycle lane/track as a separate way, but some choose to do so.

About the southbound cycle arrangements on Lombard Street with the parallel car parking, it's a tricky one. In my opinion this should just be a cycleway:left=lane on the main road way since it's so close to the road and you can technically enter/leave the general traffic lanes in the gaps between parked cars without a kerb obstructing your way. However as the car parking on this road does distance the southbound cycle lane from the general traffic lanes, I imagine some would argue that this merits a cycleway:left=track or even a separately drawn way.

I've made a small update on this road to keep the separately drawn way because I don't like to undo effort of drawing a separate way when it is technically valid.

141519016 almost 2 years ago

Hey, welcome to OSM.

This cycle lane wasn't missing from Lombard Street.
It is already mapped using the cycleway tag here: osm.org/way/49974586#map=19/53.34478/-6.24886

See osm.wiki/Key:cycleway for more about this approach to mapping cycle lanes.

Your approach of mapping a cycle lane/track as a separate way is common in some cases, but not typically done for a cycle lane like this which is an inherent part of the road.

I'm not here to convince to you choose one cycle lane/track mapping approach over the other, but to make sure you are aware of these two approaches.

Using both approaches for the same section of cycle lane/track should not be done as it results in duplication of information. So you should either keep the separetely mapped way, or change/remove the cycleway tags on the main road way.

The CycleOSM map layer is a handy way to see what cycle facilities have been mapped. The map layer can be changed in the menu on the right hand side of the screen on osm.org, if you didn't know.

138486685 about 2 years ago

Is there really an ATM here?

137611052 about 2 years ago

I'm not challenging your choice to map the footpath as a separate way. I respect that your approach is one of several acceptable approaches.

1. I don't see why OSM cannot accommodate mapping for both vulnerable and non-vulnerable classes of footpath users.

If you feel that a connection which somebody added between a footpath and road is unsafe for use by a subset of footpath users, then the solution is to tag it as such, rather than delete it.

I imagine that tags such as osm.wiki/Key:wheelchair and osm.wiki/Key:tactile_paving and osm.wiki/Key:kerb and osm.wiki/Key:lit would be useful when it comes to placing importance on the safety of vulnerable users?

For example if you deem an otherwise valid connection that somebody added unsafe to use because it involves a raised kerb, the solution is to map the raised kerb rather than deleting the connection.

2. In the example journeys I gave, there are crossings (with dished kerbs too) at the t-junctions of roads along the route (see See osm.org/node/10754018268 and osm.org/node/10754018234 for example of such locations), but you neglected to connect your footpath with the road at those locations.

It is when connections such as osm.org/way/1180320076 are deleted that the separately mapped footpaths become the only option for the router. Consequently, if the separately mapped footpath isn't connected to the road at the applicable pedestrian crossing points along the way, we end up with these convoluted pedestrian navigation routes.

Have a good night.

137611052 about 2 years ago

Are you saying you want to ensure that the navigation is safe for vulnerable road users?

The walking distance for this trip is 224 metres (your change hasn't taken effect in routing terms yet) osm.org/directions?engine=fossgis_osrm_foot&route=53.27359%2C-6.35571%3B53.27375%2C-6.35324#map=19/53.27371/-6.35417

Once your change takes effect, the walking distance will increase by about 100 metres. This doesn't reflect the reality of what route a pedestrian would actually take to reach that destination from that origin.

Here is another example.
870 metres walking distance.
osm.org/directions?engine=fossgis_osrm_foot&route=53.27902%2C-6.35530%3B53.28299%2C-6.35576#map=17/53.28114/-6.35252

Same origin and destination as above, but in google maps: 450 metres walking distance.
https://www.google.com/maps/dir/53.2790614,-6.3552574/53.2830658,-6.3556906/@53.2812425,-6.3551525,17z/data=!4m2!4m1!3e3?entry=ttu

I would ask that when deleting the footpath connection that existed at the end of a cul-de-sac and creating a separate highway=footway alongside a highway=* road, you ensure that connections are added at junctions with other roads so that pedestrian navigation is not compromised like this.

137611052 about 2 years ago

Hi, please explain why you are removing connections between footpaths and roads?
This is detrimental to navigation use cases.

134115034 about 2 years ago

Hi, it's difficult to tell from your changeset description what you've done, but I've seen several cases of houses and housing estate roads/greens that were once covered by a residential area no longer being covered by a residential area where you are the last person to change those residential areas.

Can you explain why you are reducing the coverage of mapped residential areas to not cover greens and roads, in many cases cul-de-sacs, whose main purpose is for use by residents of the houses covered by said residential area?