Stream: general

Topic: TomTom - Fix inconsistent highway intersection


view this post on Zulip Tomek_Koz (Jun 16 2023 at 11:05):

Hi Latvia community!
I bet some of you remember me from State of the Map Baltics last week, where I had a talk about TomTom’s cooperation with and contribution to OpenStreetMap.
As said during the presentation (which I think will be available shortly as a recording for those who could not attend) I want to share a couple of MapRoulette challenges that can help fix map issues in your country.
You’re welcome to try these out, and we appreciate any feedback you may have.

Fix inconsistent highway intersection: https://maproulette.org/browse/challenges/39501

You can find more information about each challenge in the description in MapRoulette.
Our editing team may also work on these challenges in a couple of week.

If you have another topic you’d like to discuss, or other map issues you’d like us to look into, feel free to send me a personal message or contact osm@tomtom.com.

Thank you!

Regards,
Tomasz

view this post on Zulip Rihards Olups (Jun 17 2023 at 16:30):

A detailed description of checks would be great for all challenges - although here it seems very simple, it would still work much better regarding involvement and trust.

Also, harr!
a competition

view this post on Zulip Colby (Jun 19 2023 at 10:36):

that LVM road import is such a mess
quality

view this post on Zulip HellMap (Jun 19 2023 at 10:42):

Is this a place where later unclassified got converted to track, but not all the segments did?

view this post on Zulip HellMap (Jun 19 2023 at 10:43):

The messy thing I think is that tiny offshoot, which is likely a turning circle?

view this post on Zulip Rihards Olups (Jun 19 2023 at 11:45):

...I'd still make all LVM roads unclassifieds. Making them tracks might be satisfying from some angle of data purism, but makes for useless map data.

view this post on Zulip UrSuS (Jun 19 2023 at 11:46):

I found few LVM that indeed are more like tracks, I set them to track and put a note.

view this post on Zulip HellMap (Jun 19 2023 at 11:55):

Yeah, most LVM roads seem to be wide gravel/compacted two-lane roads but some are barely tracks. I am considering a compromise of putting grade1 on solid LVM roads if we are going to mark them as tracks. For example, https://www.openstreetmap.org/way/124255091 . At least then the map isn't useless showing a track next to it the same way it shows this road.

In the example, this is LVM road:
also-track-lol.png

and a track it later "turns into"
track.png

Both would be marked as grade2 if we didn't consider LVM different. That's just so useless trying to plan routes or something. (I wish there was a third classification, but at this point it would be impossible to add another highway type to OSM.)

OpenStreetMap is a map of the world, created by people like you and free to use under an open license.

view this post on Zulip HellMap (Jun 19 2023 at 11:56):

At the very least, I am opposed to converting them from unclassified to track without actually surveying their grading and nearby tracks - otherwise it just "hides" the road.

view this post on Zulip Tomasz Kozlowski (Jun 19 2023 at 13:14):

Rihards Olups said:

A detailed description of checks would be great for all challenges - although here it seems very simple, it would still work much better regarding involvement and trust.

Also, harr!
a competition

I will forward this query to our team, so we can try to make the description more detailed if we are talking about the processes.

view this post on Zulip Rihards Olups (Jul 24 2023 at 18:28):

HellMap said:

...

Both would be marked as grade2 if we didn't consider LVM different. That's just so useless trying to plan routes or something. (I wish there was a third classification, but at this point it would be impossible to add another highway type to OSM.)

Why not make both unclassified?
While data purism is close to my heart as well, making map data useful ranks higher.
Marking LVM roads as tracks makes our map partially useless for data users - one example below.

Also if we have an example of a road that is marked in LVM data but is actually barely a track, please share that - would be interesting to see Mapillary images and doublecheck LVM data.

users don't care about our grade* polishing

view this post on Zulip HellMap (Jul 24 2023 at 18:49):

Rihards Olups said:

Why not make both unclassified?

Probably wrong topic and as we're talking about LVM roads again. But to answer - because, unless you want to go against OSM highway classification strictly speaking, they are tracks. Someone will just revert it later and you cannot prove that you're "right". They are land access roads that don't lead to any properties other then forests and meadows. Unless we establish local consensus to map one way or another, there is no discussion we can point to when changing one way or another. Last discussion with more replies ended up with a split of opinions (was it on Signal?).

view this post on Zulip Rihards Olups (Jul 24 2023 at 19:05):

Yeah, somebody should move these to a new topic :)
If data purism scares away users, we won't even have enough users to be pure about. Even the track page on the wiki talks about "mostly", "often", "generally" - and UK might not have such roads like the LVM ones that could be highway=primary in some regions...
It is also important to remember that wiki is not always right or complete - it is a documentation of rules community makes, not the other way.
Long-timers might recall voting with a trout, and https://wiki.openstreetmap.org/wiki/Category:Users_against_tag_voting (which includes Template:No_Wiki_Fiddlers) has many prominent users ;)

view this post on Zulip Notification Bot (Jul 24 2023 at 19:14):

4 messages were moved from this topic to #general > LVM ceļi by HellMap.


Last updated: Dec 22 2024 at 03:46 UTC