We already have factual “this is just what was added” comments from things such as StreetComplete. That’s not so bad, in the context of a StreetComplete changeset, where it’s obvious that someone is answering questions on their phone (because that’s what StreetComplete is).

That would be less useful to generalise to other changesets, because it’s missing the “why”. A completely random sample shows that people do put a fair but of description into changeset comments that simply couldn’t be determined “by AI”, like here.

… and more:

I’ll try and test some QR codes at my local bus stops

There are a few (actually, only 5) examples of departures_board=realtime on phone. I’m not convinced that this is a great tag value but “there is a QR code that you can scan” does seem like useful information to me. I did start adding note tags locally but that’s very incomplete - I soon found out I needed to worry about which stops actually existed and which not first. It’d need integration with one of the mobile editors such as Vespucci or StreetComplete etc. ideally.

I know of some bus stops where the bus_display_name and bus_speech_output_name differs between the two bus companies that serve them.

That’s another good point. I deliberately didn’t use a “:” in bus_display_name or bus_speech_output_name to avoid them looking like different languages but different names used by different bus companies is a different case again.

To address a couple of these questions:

Wouldn’t it be better for the bus route operator to fix such inconsistencies on their end?

It’d be great, but in some cases the work that needs doing is physical changes to infrastucture (removing or marking bus stop poles no longer in use), and replacing missing ones). None of the organisations involved have a bottomless pit of cash or staff to make these sorts of changes with, so it’s understandable that there are discrepancies between (a) the five(!) different sources of bus stop names that I know of and (b) the three different sources of “is this stop actually used or not”. I was a bit surprised about exactly how wrong some of the data is; it may be worth doing another diary entry about that at some stage.

QR code URL in the name is a pretty bad user experience. Font size is small, and to make use of that URL one would have to type it in switching between browser tabs since you can’t select the text from the map. I think such info is better left for details view if they tap on this stop where it can be made into a working hyperlink or at least a selectable text to copy.

“tap on the map to query it” is very much a pull requests welcome sort of thing. If someone wants it badly enough to suggest a leaflet plugin to integrate to do it, I’d definitely be interested. Until then, other sites (not least osm.org itself) can do this already. The small font size thing is only a problem if you can’t zoom in; I find that the overzoomed zoom 25 of zoom 24 tiles at the maps.atownsend.org site is plenty big enough to read, and (unlike osm.org) it’s running a version of Leaflet that tries to zoom rather than fetch on a “two fingered zoom” which makes it pretty easy to read on first load.

… tiny letters are too obtuse to understand without a legend

There is a legend that you can zoom in and out on. It’s actually merged into the data in the middle of Australia.

“letters vs symbols” is an interesting one. Pubs and fast food use symbols for most things, including coloured flashes on the underline for wheelchair access and outside seating/beer gardens. The maps use names into one of four tags (English, Welsh, Scots Gaelic and the default “name” tag), so it might have been nice to stay away from letters for that reason (and also not hardcode a currency symbol), but here I used stylised letters because I couldn’t think of a way of encoding e.g. “there is a realtime display here” in a symbol in so few pixels. If anyone can think of a better way of encoding the information without using letters then I’m all ears!

visually impaired people, who probably wouldn’t see the icons anyway

It’s often worth mentioning that people suffer from many sorts of visual impairment - “not being above to read a display or timetable at a bus stop” includes many more “forgot my glasses” than “completely blind”. Also, that’s why I mentioned Garmin Satnavs at the end - if you have control of what names appear in the map you have control of what’s spoken by the device when it tells you something it is navigating to.

@slice0 - can I ask a genuine question?

What makes you think that forum thread is about you? It was started in December. The first reference to anything related to you was in your own post linking to your own diary 17 hours ago. This post was edited to reference this forum thread only yesterday, and that was after the thread had been closed for the reasons described there.

Morocco, Western Sahara, SADR.

If you believe that what is in OSM does not match what is on the ground, I’d suggest discussing it in the forum. Also it’s worth reading this.

Ahem. At the risk of pouring cold water on one of the dumpster fires, it should be noted that the “world-class trolling” comment is here.

I don’t believe that the poster was being entirely serious, given that the next suggestion was “Perhaps you should check the Geneva Conventions to see if :popcorn: is allowed to be used against lawful enemy combatants? I am now wondering if we should send the Red Cross into this thread…”

There’s a long tradition of trying to make serious points through humour - it helps the message get across. See any episode of Last Week Tonight, or going a bit further back, Clarke and Dawe or Beyond the Fringe.

No-one’s harrassing you. People (including me) have been trying to help you understand why the reaction to some of what you’ve been saying and doing has been negative, and trying to help you contribute to OSM in a way that won’t cause such friction.

The last time I looked at the shop key in Britain and Ireland to see what people were using esoteric didn’t trouble the scorers at all. I added rendering support** for new_age because that was used at the time.

With OSM you seem to have to revisit keys at least every couple of years to see what changes there have been…

** on the map style that I look after at map.atownsend.org.uk

Hahahaha.

Where I live, “city itself” has records of lots of things, but they’re not always accurate. For example I know (and now OSM knows) where the bus stops are, but they don’t 100% match where the city thinks they are. Their list is maybe 90% accurate, but that’s no consolation if you’re looking for a bus stop that does not exist.

… so I was going through the diaries as usual, in order to hide the spam, and came across this.

It has all the hallmarks of a spam diary entry - almost entirely content free, apart from the lack of a link to some other site and the fact that it’s written by someone who’s actually edited OpenStreetMap.

May I respectfully suggest that you might want to consider adjusting your communication style so that it’s a bit less likely to confuse in this way? I mean no offence - I’m just trying to be helpful (and you did write “I urge you to reevaluate your relationship with making requests”).

– Andy

(who checks diary entries for spam with a DWG hat on but is writing this in an entirely personal capacity)

With regard to “ease of setup”, the docker guide is probably the closest to a “create a server in one command” that we have. The “sudo apt install” at the top of “manual” instructions actually does almost all of the work for setting things up, but the “extras” such as replication (which there are at least 3 different ways to do, and at least three different sets of servers to do it with).

More information about postgres tuning for larger databases is definitely something that would be useful - earlier versions of the switch2osm guide and the precursors to it did have that but it’s not been something I’ve had a chance to test recently.

So no matter what they write on switch2osm, your own tile server is mostly magic.

:)

Pull requests welcome - or even an issue describing what did or did not work (including versions etc ).

I tend to test the “vanilla replication” version about once per new OS, but the pyosmium version more often since I use it myself.

We get that you care, but OSM is a community project - we have to work together or not at all.

What this means is that if you think one thing, and everyone thinks something else, you’re wrong. I think that the rest of the Australian OSM community have been incredibly tolerant so far, but if after 27 days you continue as you did before you’ll just get prevented from editing again.

One thing that might prevent this being an issue would be if you were to promise (perhaps in an OSM diary entry) that you won’t go back to your old ways, and that you do intend to work with the OSM community, not against it, going forward.

so are we going to follow the guidelines for the gawler section

I don’t know about “we”, but “you” aren’t going to be doing anything until you have changed your attitude towards the rest of the OSM community.

that same user is just randomly deleting secondary roads now, https://www.openstreetmap.org/changeset/146839145#map=15/-34.8170/138.6552

No, they are not, as a quick look at an example in that changeset https://osm.mapki.com/history/way/173474800 will reveal.

Saying things like this that simply are not true is at best unhelpful to your cause.

When a large number of mostly incorrect changes have been made it’s normal practice to revert back to the status quo ante and then discuss which, if any, of the problematic changes were actually valid. If you believe that an intersection has been upgraded on the ground and the aerial imagery available to OSM is out of date, then providing photographs would help everyone consider what is the correct classification for OSM now.

As https://wiki.openstreetmap.org/wiki/Australian_Tagging_Guidelines/Roads#Road_Hierarchy says, “The standard practice in Australia is generally consistent with the global definition” - which means that mappers need to consider the best tagging on a case by case basis.

Setting your profile picture to an image containing CENSORSHIP and posting here that “the DWG rules placed on my edits have already ruined the map” is a bit of an odd step for someone who wants to “find a middle ground and move on” :)

a new DWG policy

There is no “new DWG policy”.

as directly lead to old outdated and factually incorrect information being put into the map.

Evidence please (for whatever it is that you are claiming)

,just the fact that it no longer has a route assignment (B19) means it is no longer a state road and does not meet OSM requirements for a primary classification.

In that snippet you seem to be talking about one particular road. It is entirely reasonable that if the classification of one road in the real world has changed, then that might affect the classification that one road in OSM.

What is not reasonable is where (on your profile) you said “I have made every single Major Traffic Road, Primary Road and Secondary Road a 1 to 1 with the Legal and Official Government Source”. This destroys the work of other mappers and will not be tolerated.

As far as I can tell from https://resultmaps.neis-one.org/osm-discussion-comments?uid=435139 , people have been trying to help you with this since 2022. It would appear to me that you have around 28 days to reconsider your attitude to other mappers in particular and the project in general.

Hello,

Your profile currently reads “In South Australia I have made every single Major Traffic Road, Primary Road and Secondary Road a 1 to 1 with the Legal and Official Government Source, The OpenStreetMap Project has full permission to use”. It has been made clear to you many times that just because a source is legal to use, it doesn’t mean the value judgements that that source made are appropriate for OSM. You decided the you were right and literally everyone else in OSM was wrong, continued, and earned yourself a one-month block from the project.

Let me ask you - if you continue like this, what exactly do you think will happen next? Do you think that literally everyone else in OSM will somehow come around to your point of view, or do you think that they will be asking the DWG to extend your block? If you think it’s the former, then I have a bridge I’d like to sell you.

Perhaps, instead of carrying on the the same vein, you should do a bit of listening. Re-read what people said when they explained how OSM’s highway classification works (both in Australia, and around the world). Reread some of the other things they said when they tried to to change your approach to the project. If you don’t reconsider, then in a month’s time you’re going to be in exactly the same place that you are now (i.e. not able to edit OSM).

Best Regards,

Andy, from OSM’s Data Working Group.

Sheepfolds are functionally different - often present in upland areas for when a farmer wants to gather sheep together (e.g. for shearing). A pinfold is for rounding up “stray” sheep in a village (and, according to some of the signs, billing the owner for not looking after them properly).

https://taginfo.openstreetmap.org/tags/historic=animal_enclosure#overview hasn’t troubled the scorers yet, but if it did we’d have to be careful not to group together items which had different function. At least “pinfold” and “pound” were functionally the same thing!

To be honest, I suspect that the question “should we map official data into OSM 1:1” is a bit of a red herring.

The bigger issue is that the writer of this diary entry is simply unable to discuss with other people about things; it’s as if they can’t imagine a world in which other people have evidence and experience that they don’t have.

I suspect that if there was no disagreement on road category tagging we’d have similar dramatics about something else.

** Andy (writing in a personal capacity)

“pinfold” would definitely be preferable to me to “pound”, as “pinfold” doesn’t have several other meanings (some similar, some very different).

We wouldn’t want people to get confused between (say) a police dog pound or a car pound and one of these historical animal enclosures.

At the risk of adding to the confusion, “penfold” is also used.

Also, it’s a bit odd that there are none of these in Scotland and few in Wales (but there are in Ireland and England). Maybe there are other names too? You mentioned perhaps trying to talk to CADW elsewhere; maybe the Scottish National Trust might be another avenue?