Following the “A,B,C” tag rules will keep South Australia consistent with the majority of the other Australian States and larger worldwide communities.
information found on The Australian Guidelines page Australian Tagging Guidelines - OpenStreetMap Wiki will be ignored in favor of the OSM wiki definitions and policies. This is to keep our standards in line with that of the larger OSM community.
The Australian Road classification definitions are a direct copy of the British definitions. Australia adopted the British system when our roads were first being built.
Following this proposal will ensure that data can be maintained and up to date as the network changes.
This information can be cross referenced by anybody even without knowledge of our roads.
If there is no major objection I plan to follow through with edits to reflect this proposal in 28 days from now.
If there are any concerns or questions, even if it is about the history of road classifications in Australia through out the history of the OSM project. I am available to answer.
Tagging roads based purely on routes, with no other considerations, is not a good idea. You’re ignoring that a route that stretches a large distance may go across different classes of roads.
Other states use the Australian Tagging Guidelines where it suggests M=trunk, A=primary etc as a starting point, but also accept that deviations from that can and do occur.
You’re also ignoring the several discussions the community has in favour of discussions upwards of 10 years old, and the various occasions recently where it’s been pointed out that official data is not the best information for road classifications.
So yes, I object. And also, adding that data in anyway is an import and would need to follow the import guidelines separately to this discussion, and as per previous block messages and emails from the DWG you’ve received, making these edits would also go against what you’ve been told.
Edit: I’ll also add, no comments is not the community agreeing with you, you really need an actual positive response, similar to tagging proposals on the wiki.
Thankyou for the feedback based on the above I made made an amendment which I changed “Class B = Primary” to “Arterial = Primary”
reading through other proposals that have gone ahead from this forum there doesnt need to be much of a positive response, I will leave it up to DWG members to decide if need be
Government data isn’t infallible, so in my opinion the knowledge & judgement of local mappers should take priority.
If a particular bit of government data doesn’t make sense, it shouldn’t be in OSM (even if that means a non-local can’t easily cross-reference it).
our road data is very open and researchable via old roadwork notices and proposals if needed, since our roads were first classified on OSM way back in 2008 there have been many changes to road classifications which have been reflected on data sa but not on OSM. If there are circumstances where official data is questionable it can be research on a case by case scenario. When weighing up the official data vs using any other method there are far less chance for error, and if there is a error we can work with local councils to fix both sets of data. Keep in mind I have personally and by hand edited every single road and there hasnt been anything that really stands out as a error in my eyes. Keep in mind government data only refers to Trunk, Primary and secondary roads which by nature has a very slim chance of database error because of how well documented they need to be.
on OSM overall policy
Contributions to OpenStreetMap should be:
Truthful - means that you cannot contribute something you have invented.
Legal - means that you don’t copy copyrighted data without permission.
Verifiable - means that others can go there and see for themselves if your data is correct.
Relevant - means that you have to use tags that make clear to others how to re-use the data.
Can you provide examples of where they already do differ, that way everyone can see where there’s differences between OSM and DataSA’s Roads dataset and be able to provide better feedback.
I would actually want to defer to the consensus from 2008 for roads that haven’t been re-evaluated since then, and haven’t been substantially upgraded (i.e. most of the inner metro area!)
But there was far more to it than just copying any sets of official classifications. All the reasoning and considerations for individual roads are split across several long talk-au threads from that year but it’s summarised at Australia/Archive/WikiProject South Australia - OpenStreetMap Wiki . It’s still inconclusive but I don’t think SA road classification has been discussed anywhere near this exhaustively between 2008 and recent months.
Those would be good to start from, and then look at what’s been upgraded since and what’s become more used from other development since then.
My main objection is just taking discussions and consensus from then and trying to apply it now, when there’s been discussion and comments that do show the opinions have shifted, as well as the pure fact that government classification data just isn’t perfect. It’s a great starting point, but there will be roads where they do actually vary from that data, and that should be reflected.
well the most debated one that comes to mind is this section of main north road in gawler OpenStreetMap , it has always carried the Primary classification until when the gawler bypass was built and this section was officially re-classed as Secondary. I wrote a diary entry on this section slice0’s Diary | the DWG rules placed on my edits have already ruined the map | OpenStreetMap there are many other roads that have been changed from the official classification over the last couple of months.
This proposal is about establishing a future policy as we currently do not have one
If I get some time I might try and pull together an easier to work with list from that archive article, and then get what OSM has primarily had on each road since then, any notable changes I can think of, and that would give a good starting point to consider each road.
I’d be happy to check the DataSA classifications into that table as well, just as another reference point to compare.
And part of the issue I and others have is that the official classifications aren’t always the most appropriate. That Gawler one is a great example, considering freight routes and how people travel to higher class roads from within Gawler, it doesn’t exactly match how the government classifies their roads.
That data was good to bring in 1:1 when there was nothing, because it was a starting point where there was nothing. It’s a good thing to take into consideration too, but it shouldn’t be the only thing considered.
And I’ll just comment that what you originally posted was more along the lines of “I think this is how it should be done, and if nobody makes a big fuss I’ll just do it”. If you want to work on better guidelines that can be used going forward, then talking about timelines and potential edits straight out the gate is not the way to do that.
If you’ve got some examples where DataSA doesn’t match OSM, then give us a list and we can all talk about it. We’ve got the previous discussions that were had and a lot of those were about how the two varied, and we can all talk about what’s the same, what’s changed, and what seems more appropriate.
This proposal removes the need for debate and classifications can be verified. It is not OSM policy to make things up but rather map as close to factual information as possible.
Can you show us where specifically you can see that the classifications map 1:1, and how that should override cases where it’s clearly wrong? Because you haven’t been able to show that despite constantly claiming it.
I’m all for trying to come to an agreement on how these things are mapped, but that means working with others, listening to the issues and criticisms they have, and accepting that using official data for this isn’t a one size fits all situation. There’s plenty of cases where that can be shown, changeset 144928924 is a great example, and is why I consistently try to urge that government data is a great start and is valuable, but should not be the be all and end all for classifying roads.
Officially there are no Tertiary roads in South Australia, our government uses the Collector classification instead. There is no resemblance between the two.
Under this proposal the current classification (Tertiary) is correct because it perfectly matches up with the description of the wiki Tag:highway=tertiary - OpenStreetMap Wiki
this proposal does not classify tertiary or residential roads based on official data, this proposal uses the wiki definitions instead.
The reason is that South Australia does not have tertiary roads, we have Collector roads instead. Which are roads designed for bus routes, that Ruskin Road does not have a bus route so the council call it a residential road because there is officially no alternative.
South Australia does not have “unclassified” roads either, so any other road is classed as residential.
Since I looked into that exact changeset all that time ago I have since realised this and only propose Arterial roads follow official classification.
Okay, in that case can you then show us which roads differ between the two so we can have a better understanding of what you want to change? Pretty standard part of any proposal like this would be to show some of the actual changes you want to make, and it gives everyone a chance to actually see what you’re wanting to change and provide feedback, especially if there are any roads that should be excluded.
This may not be a huge import bringing in brand new data, but it’s still an import and that process should be followed properly.
I’d hazard a guess that tastrax was more focused on showing how government classification data isn’t always the greatest, considering you used it to classify that road as residential in the past.