OpenStreetMap logo OpenStreetMap

Changeset When Comment
35418764

Hello and welcome to OpenStreetMap!

It looks like some tags got deleted from node/1241221826/history (including the name and "information=ofice", which says that this is a tourism information office and not, say, a signpost). Was that deliberate? No problem if it was, but I can add them back if it wasn't...

Cheers,
Andy

35619303

Re "Is there a way that JOSM warns when producing changesets with large bboxes?" I'm not aware of one, but I've not looked either. If there is one I'm sure that people on the help site or on one of the main lists (or in the German forum) will know.

35582900

Perhaps a bit of background about the layout here will help. Historically, there were 4 tracks down towards Nottingham, not 2. at node/2202969410 Someone's added a pair of abandoned railways, and they did used to join the MML at about that location. Handily, wikipedia has a photo from 1983:

https://en.wikipedia.org/wiki/File:Clay_Cross_station_site_geograph-3110788-by-Ben-Brooksbank.jpg

Around 10-15 years ago the 2nd pair of tracks towards Nottingham were removed, and the trackbed on the other pair was changed, which resulted in them being resited slightly.

The reason there's more of a gap between the pairs of tracks at the bridge because that's where the station was (island platform). The station's gone but there's now a bridge pillar there, which is why the two pairs of 4 lines are fairly close at the site of Clay Cross Junction, separate a bit to go under the bridge, and then get closer together again. You can get a bit more of an idea from http://claycross.org.uk/History/railways.htm .

It's also perhaps worth mentioning that at around node/2203012981 there's some new-build work taking place (looks like a new passing loop located where some of the old trackbed was), so that bit in OSM is incomplete or soon to be so.

35457641

@RobJN Aside from whether or not _any_ of the changes in this sequence were a good idea or not, which is debateable (poor Haslewood Road in Newton Aycliffe way/63334615 will be wondering what's happening to it) you clearly have a preference for mechanical edits to data in the UK. That _is_ something worth discussing and I suggest that you write a "position piece" for talk-gb explaining your thoughts. It's about more than whether a particular node is tagged with highway=traffic_calming or not; it's more about a "philosphy of mapping".

35457641

Rob, when was this discussed on talk-gb? 2 days ago you were clear that it hadn't been ( https://lists.openstreetmap.org/pipermail/talk-gb/2015-November/018071.html ). As far as I can see no-one replied to that. You asked the question "Does this mean we can now look for someone to un-revert this changeset?" but no-one answered - not even you saying "well, I've not heard anything so I'm going to do it".

35582900

The "two-lane road" analogy isn't a good one because there's no "passenger_lanes=8" tag/value on e.g. way/30175700 - there's "lanes=4", and also "lanes=4" on the other carriageway way/40895418 . In my opinion the same would work here - just use the implicit tracks=1 on each of the four lines and leave it at that. However, if you must tag "there are some lines quite close together here" at least use a value that corresponds to "the number of lines that are quite close together".

35582900

Re "How do you know that the track gauge is 1435mm" I've seen trains with that track gauge go down it and they haven't fallen off :)

35582900

Sorry, but that's rubbish. At way/212750580#map=19/53.18452/-1.40075 there are no joins between the four parallel tracks. In what way is that "two sets of two parallel tracks topographically"?

Adding "nonsense" tags makes OSM data harder for data consumers to understand and harder for them to use.

OSM data is supposed to be verifiable - a local mapper (like me) is supposed to be able to go to a place and say "yes, I can see that". Here the "passenger_lines" is not just superfluous (i.e. something that ought to be derivable from the data) the value that you have used is actually wrong.

35674754

Reverted. See osm.org/user_blocks/858 and comment on changeset/35653117 .

35674018

Reverted. See osm.org/user_blocks/858 and comment on changeset/35653117 .

35653117

This changeset has been reverted in changeset/35675358 .

LeTopographeFou, please do not make further mass edits like this. As you have found out from the comments that other mappers have made on your most recent changes, it's very easy for you to misunderstand what data means. Often data that is flagged as "invalid" by JOSM is very useful, because it highlights work by an inexperienced mapper or where something has gone very wrong (see way/103231400 below for an example of this). If you change this data so that it no longer appears to be "invalid" it will still be wrong (what's in OSM won't match what's on the ground) but people won't be able to find the problem via QA sites any more. Also (like the Truvelo camera on the other changeset), if you don't understand what some data means, don't change it to something that you do understand. You can ask the person who added the data in the first place, you can ask local mappers by adding an OSM note, or you can ask on a country-specific mailing list or forum, or on an international list.

In response to your comment elsewhere, the Data Working Group does view changes that dilute the accuracy of data within OSM, when carried out as here in spite of the opposition of other mappers, as akin to vandalism and will revert them.

However, you are welcome to continue editing OpenStreetMap provided that you do not engage in mechanical edits that destroy the meaning of data like this.

Best Regards,
Andy Townsend (SomeoneElse)
On behalf of OSM's Data Working Group.

35381971

Currently Edinburgh is set as "capital=2" which I don't believe that even Nicola Sturgeon would claim currently :)

The changes can be seen here:

http://osm.mapki.com/history/node.php?id=17898859

Looks like it was 4 since early this year (which I think matches a discussion somewhere; tagging list perhaps?). Whenever changing data that could potentially break people's extracts I'd definitely mention it somewhere fairly public (and talk-gb is probably the best place for that). Also, a changeset comment explaining what was changed and why would be really helpful to other mappers.

35619303

"no;yes" is usually the result of an unintended way merge, so you can usually figure out which ways were merged to create the invalid tag.

I'd also always explain to the person who did the merge what went wrong, so that they won't do it again.

Here just choosing "no" or "yes" without investigating means that the data is just as wrong as it was before but that that error is no longer highlighted by a QA site, which it was before.

35619303

I'm surprised that you were able to resolve way/39326398/history as oneway=no without splitting into two again?

35667995

This looks very like a mechanical edit. Did you discuss it beforehand as per osm.wiki/Automated_Edits_code_of_conduct ?

34795486

Thanks - in that case the footpath should be joined to the A38 carriageways then. As a driver I don't remember seeing a sign for the crossing, so as a pedestrian I don't think I'd want to cross the road there if I could avoid it!

35379263

Hi, I think that something's gone a bit wrong at way/380788371#map=19/52.86637/-1.56548 . Currently there's no join between that footpath and the A50 slip road (which the certainly was when I last walked it). Also the "Derby Nomad Way" relation currently goes down it, which I'm not sure about (see relation/1069633#map=17/52.86596/-1.56571 ). The Derby Nomad Way is signposted on the western side of the A38, and down the steps north from the A38 slip road. Logically to get between the two you'd walk around the northern part of the A38/A50 island, though I didn't see any signs - are you sure the route is definitely signed around the southern and western sides of the island? For info, you'll notice that the northern bit of the island has "sidewalk=both" or "sidewalk=none" on the road to indicate that a roadside footpath is present or not - it's a better way of indicating to everyone a "footpath that is part of a road" rather than one that is separate infrastructure. The path to the west of the A38 is of course separate infrastructure, and mapped as such.
Anyway, hope you don't mind me mentioning this, just trying to help. Any questions please don't hesitate to ask.

34795486

Hi,
Just wondering how way/376352067 gets across the A38? Is there an underpass or a bridge or something, or to pedestrians have to run across the A38?

35649194

(for anyone stumbling across this edit unaware of the history) Coffeedex is/was a "single-tag proof of concept editor" - see @tmcw/diary/28138 for a discussion about it. It's more about "how can we get new people to update OSM with different data in different ways" than "how much does coffee cost here".

16601166

Great, thanks.