Re: [talk-au] answers to the difficult questions
On Wed, 17 Feb 2010 19:32:42 +1100 Liz ed...@billiau.net wrote: I believe its actually dumped into the town sewerage system which in the Murrumbidgee means it is all going to be recycled into drinking water for Adelaide after we've used it 10 more times. And will probably improve the quality of the water here anyway ;) -- =b ___ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-au
Re: [talk-au] Marking non-existent roads...
On Wed, 12 Aug 2009 19:41:10 +1000 Liz ed...@billiau.net wrote: On Wed, 12 Aug 2009, John Smith wrote: Does anyone have a suggestion on marking non-existent roads, so people don't waste time trying to map them? when you get to his area, non-existent roads are often roads into properties, and belong on the old Telstra database from which they were robbed, but not on a normal map For all that it saves mappers time putting this data on there, surely it's bordering on infringement issues putting reference to them in there. Especially things like honey-pot streets. Darrin -- =b ___ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-au
Re: [talk-au] Australian Rendering
John Smith wrote: You can do this I guess it's a matter of rendering. I think this discussion has highlighted an underlying issue that is beyond the rendering issue however. The underlying issue is that you don't consider administratively equal a suburb and a rural named area despite the fact that the various state governments and australia post do. Australia post doesn't necessarily follow that, Gympie has a number of areas inside the town limits but all mail is addressed as Gympie, eg South Side, Corella, Araluen, Monkland, are the ones I can think of off the top of my head. Even Australia post has to bend to the perceptions of the locals regardless of what's gazetted, either that or Auspost has a later version of gazetted data than the ABS released. All but 'South Side' are listed on the Australia post site as valid postal areas in postcode 4570 as an equal category to Gympie. Yeah this is a bit of a pain when you get duplicates in particular. Really a renderer specific problem however. I liked the idea that surfaced a while back of optionally being able to specify a 'centre' to a boundary relation which should be where the name is rendered and only rendering it centrally if that is absent. Don't think it's made it into the renders though. If one of us can figure it out, we can render it. :) Yeah well, that I'd be all in favour of seeing done :D Darrin ___ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-au
Re: [talk-au] Australian Rendering
On Mon, 3 Aug 2009 19:14:46 +1000
Liz ed...@billiau.net wrote:
On Mon, 3 Aug 2009, Darrin Smith wrote:
What about the 90% of Australia that isn't metro areas and these
boundaries are all over the place and don't line up with towns,
They still provide general enough information to be useful.
That same 90% also has a large number of source=landsat (or similar)
roads on them, which aren't very accurate either, perhaps we could
remove them also since they're just clogging things up? ;)
they're usually smaller or larger depending if the town grew
So they need 'fixing' this doesn't make them invalid.
they are not lined up with reality in our area
and I haven't actually got time to move Lake Wyangan ('locality', not
the water) back where it belongs
but the ABS has let it and Bilbul take over Beelbangera
http://www.openstreetmap.org/?lat=-34.2533lon=146.0973zoom=14layers=B000FTF
Not sure what your point here is Liz apart from interesting information?
Although it does back up what I was saying about the ABS begin
incomplete not having every designated place - I assume Beelbangera
doesn't even exist in the ABS data?
Still it seems to have the 2 places you mention in vaguely the right
place (Not in Queensland for example ;) so as always it's better than a
void.
--
=b
___
Talk-au mailing list
Talk-au@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk-au
Re: [talk-au] Australian Rendering
On Mon, 3 Aug 2009 19:59:48 +1000 Liz ed...@billiau.net wrote: On Mon, 3 Aug 2009, Darrin Smith wrote: Not sure what your point here is Liz apart from interesting information? Although it does back up what I was saying about the ABS begin incomplete not having every designated place - I assume Beelbangera doesn't even exist in the ABS data? Still it seems to have the 2 places you mention in vaguely the right place (Not in Queensland for example ;) so as always it's better than a void. the information is useless here for suburb / village boundaries. But until someone has the time to fix it, it will provide someone who doesn't know the area well enough information to be in approximately the right place (same as most landsat traced roads are for example). I can't actually find anywhere else to compare that boundary to so for all I know it's accurate anyway :) -- =b ___ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-au
Re: [talk-au] Australian Rendering
On Mon, 3 Aug 2009 20:50:22 +1000 Liz ed...@billiau.net wrote: On Mon, 3 Aug 2009, Darrin Smith wrote: But until someone has the time to fix it, it will provide someone who doesn't know the area well enough information to be in approximately the right place (same as most landsat traced roads are for example). I can't actually find anywhere else to compare that boundary to so for all I know it's accurate anyway :) no, the place name information would be more accurate. that would at least get you to the post office for the area Lucky someone put the time into that then isn't it. Lucky that's the case for every other locality in Australia... oh wait, it isn't! -- =b ___ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-au
Re: [talk-au] Australian Rendering
John Smith wrote: --- On Sun, 2/8/09, Darrin Smith bel...@beldin.org wrote: Why remove suburb boundaries? Because they aren't suburb boundaries, they're ABS boundaries. In 95% of cases they're close enough, are you going to throw out the baby with the bath water and dispose of the cases where people have adjusted them to be correct? Even in the places where they aren't they're the closest we're going to get for a long time. Darrin ___ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-au
Re: [talk-au] Australian Rendering
John Smith wrote: I'm not saying they're inaccurate, I'm saying they create noise in some areas rather than showing useful information. Completely a matter of opinion, and again the same thing could be said about innumerable bits of data that don't fit into your perception of what needs to be on the map. In the middle of western regions in NSW there isn't much of anything, let alone something of significants as a suburb boundary, but this is why I suggested using alternate Are you sure they are not suburb boundaries? All over SA these very cases you are talking about ARE suburb boundaries, nearly the whole state (at least south of Goyder's Line) has 'suburbs' defined which as I said before the ABS almost lives up to matching. It's not just place=one_size_fits_all, and in this case suburb boundaries would be useful, boundaries in rural Australia is confusing, no one navigates like that. They work exactly the same as suburb boundaries, they divide the level above (postcodes) into areas that the Geographical name boards of our various states have defined under a particular name. What about for every day use, that would be a once in a while use, which is more important, as I find them completely unhelpful, distracting even as I'm not used to seeing maps drawn that way so it makes quickly referencing an area a much bigger task. Maybe not every day, but probably once a week or so for me they are useful. So let me see if I get this correct, because you aren't used to maps looking this way we should change the way the data is represented? Navit makes it even worst by making them very obvious and it confuses things trying to navigate that way. Ah, so there's a certain amount of 'tag for the renderer' happening here also. That's just it, they're ABS boundaries and they are where they are for a purpose, statistics mostly, should we be really messing with them en masse, but I think we need to selectively show such boundaries or we should be drawing new boundaries that are more meaningful. These are PLACE names not ABS names. They are quite often the aggregation of several ABS areas into 1 for that coincide with a place. They are valid locations, they are valid suburb boundaries, in fact in some areas ABS doesn't actually have all of place names that exist and have boundaries. We should have no hesitation correcting ABS data because they don't necessarily reflect individual ABS survey areas anyway. Franc put a lot of work into bringing them in, the list talked about it at the time and I believe it was accepted that they would be wrong in places but it was the best we could get given the whole Australia is so big issue. Darrin ___ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-au
Re: [talk-au] Australian Rendering
John Smith wrote: --- On Sun, 2/8/09, Darrin Smith bel...@beldin.org wrote: Completely a matter of opinion, and again the same thing could be said about innumerable bits of data that don't fit into your perception of what needs to be on the map. The same argument could be made for UK v AU opinions of what a map looks like, it's obvious you are used to one way a map looks like and I'm used to another, additionally it seems we use maps differently which may make different things relevant. That's a given for any 2 people. However how the same thing is store in OSM in 2 different areas is the same. And that is what we are talking about here. Are you sure they are not suburb boundaries? All over SA these very cases you are talking about ARE suburb boundaries, nearly the whole state (at least south of Goyder's Line) has 'suburbs' defined which as I said before the ABS almost lives up to matching. Now we're going into very subjective territory, a suburb to me is a section of a metro area, not a patch of dirt in the middle of no where with only a handful of people. What we call the ares is not the issue. Even though just about any information I've found which refers to the divison of post code areas into named areas calls them suburbs I will accept perhaps it's not the best choice here because it's obviously confusing you. I apologise for that. So let me see if I get this correct, because you aren't used to maps looking this way we should change the way the data is represented? That's the entire point of the exercise, we all see maps differently, otherwise we'd be happy using the UK rendered maps. By represented I meant in the OSM data. The underlying data should be consistent. Ah, so there's a certain amount of 'tag for the renderer' happening here also. That's true for a lot of tags, otherwise roads would all look the same and wouldn't show differently to indicate the most used roads. This is all I'm really looking for, to differentiate areas into importance and labeling everything the same isn't descriptive enough. These areas are of the same importance, they are the geographical name divisions that lie within post codes. These are PLACE names not ABS names. They are quite often So add a place tag, what would be less than hamlet? They are not a point, they are a named area, documented in the states naming registry and everyone who lives in those areas lives in that named location. the aggregation of several ABS areas into 1 for that coincide with a place. Actually it's not that simple, some areas are bigger than the boundaries some areas cover less than the boundaries. Yes, that's been talked about before, they're the best we've got now, we fix them up when we know more about them. They are valid locations, they are valid suburb boundaries, in fact in some areas ABS doesn't actually have all of place names that exist and have boundaries. They seem to be all named to me except non-incorporated areas which they list as unclassified. Please read what I am saying. I did not say Not all the ABS data has names I said the ABS data doesn't have all the names. There are even more of these names in existance than the ABS data would lead you to believe. We should have no hesitation correcting ABS data because they don't necessarily reflect individual ABS survey areas anyway. What's wrong with it exactly? Have you been reading my emails or just jumping on the points that you like? Many of these areas are multiple ABS survey areas joined together, so if you were to say make this layer the abs data layer it would be wrong anyway because it doesn't have *all* the ABS survey areas. Which I'm grateful for, it helps marking out things with low res sat imagery and you don't have to survey like rivers. I you really want them accurate you still need to survey, I've found plenty of cases where ABS data doesn't match exactly what's on the ground. But it's a darn sight better than what has existed before. Darrin ___ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-au
Re: [talk-au] Junctions (to name or not to name)
On Sun, 28 Jun 2009 09:41:38 +1000 Rick Peterson ausr...@iinet.net.au wrote: Hi Folks, Quick question about roundabouts. (junction=roundabout) Originally, I didn't name roundabout junctions, but when I validate my work in JOSM, it identifies them as 'Unnamed Ways' in the warnings section. I've tried naming a few using the name of the primary road that connects with the junction. The validation warnings disappear, however, the rendered work looks messy as the streets AND roundabouts get named at close zooms. http://lists.openstreetmap.org/listinfo/talk-au Since SOME roundabouts have individual names (e.g. Britannia Roundabout here in Adelaide) so the renders are doing the right thing printing names differently for the roundabout. The issue is that a normal roundabout which nearly always involves more than one road really can't be appropriately named and that JOSM's validator actually is buggy in the sense that it shouldn't be asking you to name every junction=roundabout. For a long time I used to put in name:link=road to shut it up but then after a private discussion with jackb way back he helped me realise that that was not appropriate and that it's really the JOSM validators fault. I believe there's a bug report in about it somewhere but it's obviously not on the current coders radars at the moment. -- =b ___ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-au
Re: [talk-au] Hi all ...
On Wed, 17 Jun 2009 05:30:43 -0700 (PDT) Delta Foxtrot delta_foxt...@yahoo.com wrote: I don't have a problem with uploading gps traces to osm but I can see no benifit if I'm just going to edit them in josm anyway and given that they are all one second data that's a lot of data to put on the osm servers when not really necessary. The benefit I see is from law of averages, to a point, if you get 50 people tracing out the same ways, you average the points and neglect any spurious data and you should have a very accurate plotting. Have to agree about 200% with Delta on this one, I've seen way too many cases of people who have obviously used their 1 gpx track to move and existing surveyed way to follow it when if they'd used theirs and the 20+ traces that were available on osm (and uploaded theirs so the next person would have 21+ to use) they'd realise their track was a off that day. I've also seen too many cases myself where I do a first trace down a road to get a path then put together 3 or 4 over a period of time and realise the first trace was off. The moral of the story is never ever ever use a GPS trace in isolation if there's any others available in the area, and make you traces available so the next person working in the area can benifit from your input also. -- =b ___ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-au
Re: [talk-au] How to map out streets the most efficently
On Tue, 16 Jun 2009 07:22:42 +1000 Liz ed...@billiau.net wrote: On Tue, 16 Jun 2009, Delta Foxtrot wrote: Thoughts, hints and tips will be appreciated. Personally I'd concentrate on doing one or two on the journey up and one or two on the journey back. One large and one small? I'd agree. It's not a race so focus on doing the job fully to the level you are happy with on a smaller scale and come back later to finish it. If you try to rush and shortcut you'll only have to go back later to finish the things you find out you've missed the first time. The number of POI's and small side alleys I've found off obscure back streets of rural towns you really need to check every street if you want to consider everything mapped. -- =b ___ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-au
Re: [talk-au] Maritime boundaries
On Sun, 14 Jun 2009 04:59:36 -0700 (PDT) Delta Foxtrot delta_foxt...@yahoo.com wrote: --- On Sun, 14/6/09, Darrin Smith bel...@beldin.org wrote: The reason I chose to put in the (roughly estimated) 12nm boundary was that from the research I could find it's the *legal* definition of the extent of full australian territory, i.e. when you are inside 12nm you are in Australia and all laws apply - as is the case with most countries from what I could determine. The 200nm is to do with resource exploitation and full teritorial rights do not exist in this area. The other countries around where world where people have added maratime borders and been at the 12nm limit also from what I can see. I should have read up on it first I guess, but that's only partially true. Since UN based conferences in the 1960s various nations have signed up to various international treaties, that originally went from 3nm, to 6nm, to 12nm and now various countries have put in requests for recognition to 200nm, Australia being one of them. Exceptions in Australia's case exist where it meets territorial waters with PNG and they have agreements in places as to who owns what, and around several islands which only extend 3nm. http://en.wikipedia.org/wiki/Territorial_waters What I can't figure out is if these treaties have been ratified and Australia's submission for 200nm was/is accepted/valid, or what's going on. http://en.wikipedia.org/wiki/Territorial_waters#Submissions_with_recommendations The URL provided as a reference for all the maratime boundaries I entered makes it pretty clear that 12nm is our territorial water boundary, and references the Torrens Strait issue: http://www.ag.gov.au/www/agd/agd.nsf/Page/InternationalLaw_AustraliasMaritimeBoundariesandZones I never bothered going into heavy details in Torres Straight because getting that remotely accurate would require using maps provided by the gov. which of course raises the ugly issue of copyright. Too hard basket to translate all the written versions into meaninful maps, and too much chance of offending someone who lives up that way if I got it wrong ;) Interestingly that page I linked lists the Exclusive Economic Zone (the 200nm case) as extending from the outside of the 12nm limit, which would actually make it 212nm from the baseline case. Also if someone is interested in getting the 12nm truely accurate they'd need to look at the 'Internal Waters' cases because from some maps I've seen around the place there are quite a few bays classified as 'internal' which I didn't count as internal when I drew up the lines, so the border could be further out in some areas. This page: http://www.comlaw.gov.au/comlaw/legislation/LegislativeInstrument1.nsf/0/3075C0CCC553EF84CA25711400120045?OpenDocument links to a definition of the baseline (including lon/lat points) if anyone's really interested. -- =b ___ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-au
Re: [talk-au] Maritime boundaries
On Sun, 14 Jun 2009 06:02:18 -0700 (PDT) Delta Foxtrot delta_foxt...@yahoo.com wrote: --- On Sun, 14/6/09, Darrin Smith bel...@beldin.org wrote: Interestingly that page I linked lists the Exclusive Economic Zone (the 200nm case) as extending from the outside of the 12nm limit, which would actually make it 212nm from the baseline case. An exclusive economic zone extends for 200 nautical miles (370 km) beyond the baselines of the territorial sea, thus it includes the territorial sea and its contiguous zone.[3] A coastal nation has control of all economic resources within its exclusive economic zone, including fishing, mining, oil exploration, and any pollution of those resources. However, it cannot regulate or prohibit passage or loitering above, on, or under the surface of the sea, whether innocent or belligerent, within that portion of its exclusive economic zone beyond its territorial sea. Before 1982, coastal nations arbitrarily extended their territorial waters in an effort to control activities which are now regulated by the exclusive economic zone, such as offshore oil exploration or fishing rights (see Cod War). Indeed, the exclusive economic zone is still popularly, though erroneously, called a coastal nation's territorial waters. http://en.wikipedia.org/wiki/Territorial_waters#Exclusive_economic_zone Yes, I'm fully aware of the 200nm official UN definition, I just found it interesting (hence the use of the term 'interestingly' at the start of the sentence there) that that page appeared to be in error. -- =b ___ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-au
Re: [talk-au] Yorke Peninsula and South Australia
On Mon, 25 May 2009 18:50:15 +1030 Graeme Wilson wandere...@live.com.au wrote: If you need something checked anywhere, ie street names etc, as long as it will only take a few minutes as I am passing through, then make a list and I will see what I can do. On good thing to get since you mentioned the YP is the path of the B89, I've added it down to north of wallaroo but I can't remember where it goes from there so if you're in that area.. :) -- =b ___ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-au
Re: [talk-au] Causeways
On Tue, 26 May 2009 07:31:01 +1000 Liz ed...@billiau.net wrote: On Tue, 26 May 2009, Delta Foxtrot wrote: My original question was in relation to concreate slab crossings which technically aren't fords because they dry far more often than wet, and they aren't raised at all so they're not bridges. I can't find an example of what I mean, I'll have to take a photo of one and post it online in the next few days. the definition of it being wet is a pommie problem you need rain before they get wet we don't even have a marker for rivers or lakes which are seasonal (ie, usually dry) i'm not at all bothered if you label a ford ford when the creeks dry - if the creek had water it would be wet?? I agree Liz, I was just thinking pretty much exactly the same thing a few minutes ago when Delta's post arrived. It's totally a factor of the state of the watercourse in question. Luckily for us a couple of Fords I went through on the weekend did actually have a spot of water in them, it's been nice this last month to finally see some of that long forgotten rain :D -- =b ___ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-au
Re: [talk-au] Manipulating ABS suburb data, and related data..
On Mon, 27 Apr 2009 22:40:04 +1000 Liz ed...@billiau.net wrote: On Sun, 26 Apr 2009, Liz wrote: 1433 nodes split it into 600+ and 700+, still couldn't upload the change. I wait on the bug report - someone has another problem with relations and they may be related problems Merkaartor let me bypass the problem and split the way however, now one length has two relations member of multipolygon Oxley member of mulitpolygon KeriKeri other sector has only one relation member of multipolygon KeriKeri attempting to add this sector to multipolygon Oxley fails transfer aborted due to error condition 'Interesting' problem back to bug reporting and somewhere out here (Balranald region) is a way with over 3000 nodes so it will choke too at some stage :-( I'm working on 2 relations like this on KI that sound like they are having the same kind of issue. ANY edit of any form to the relation would cause an error - EXCEPT deleting I eventually tried on one of them. So I've just kept a copy of all the details about the relation, deleted the old one and created an identical one that matched the original, THEN I could edit the new one without any trouble. I suspect the same thing might work with the 'Oxley' relation (I think your successful split was because the way that kept the realtion was the one that kept the original way Id # and therefore there was no edit actually required on the relation). -- =b ___ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-au
Re: [talk-au] Manipulating ABS suburb data, and related data..
On Sun, 19 Apr 2009 19:47:04 +1000 Ben Kelley ben.kel...@gmail.com wrote: Hi. I think I'm with Darren on all counts. I think the only thing I'd add is where local knowledge tells you that the ABS data aligns to some previously unmapped feature (e.g. a river) that cannot be made out on Landsat (no Yahoo coverage). There I'm tempted to add the natural feature data to the existing ABS way, and leave the ABS attribution in. I have seen a few places where the ABS suburb boundary seems to exactly match a river or creek outside the Yahoo coverage, and there is no easy way to get a GPS trace for the creek. Oh yeah, I've done that for a couple of things too now that you mention it :) And yes I've left the attribution tag in for sure. I'm doing some surveying around Tamworth NSW this weekend, and I was looking forward to uploading my traces tonight, but of course the OSM database is being upgraded! Know that feeling, I have several days of Easter holiday traces to go in also. *twiddles*thumbs* -- =b ___ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-au
Re: [talk-au] suburb boundaries
On Sat, 21 Mar 2009 12:29:52 +1030 Darrin Smith bel...@beldin.org wrote: On Sat, 21 Mar 2009 11:46:44 +1100 Franc Carter franc.car...@gmail.com wrote: Hi all, The upload has completed (much faster running from dev). There were a couple of problems:-] * Gruyere and 'Wandin North - Bar' in Victoria, which I *believe* I have fixed * Beatrice and Ellinjaa in Queensland which are too complex for me to fix as I don't have local knowledge cheers That made a serious difference the the speed of things, wow. Now to resolve the differences between my own boundary work and the ABS stuff in northern adelaide, and at a first glance I must say I'm glad I told you to upload the data anyway, because there's a couple of places where I think the ABS is more correct than my results (and a few the other way also of course ;) Going to be fun correlating the two. :) Futher poking around I've found the 'Unclassified SA' 'suburb', containing over 100 segments scattered all over the state, I assume most other states will have a similar object, what's the thoughts of everyone on this case? Is it really needed? (I assume it's just a category in the ABS data that's come across wholesale). Seems to me anything not in side a suburb boundary would be considered unclassified anyway? -- =b ___ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-au
Re: [talk-au] suburb boundaries
On Sat, 21 Mar 2009 13:38:40 +1100 Franc Carter franc.car...@gmail.com wrote: [snip] Futher poking around I've found the 'Unclassified SA' 'suburb', containing over 100 segments scattered all over the state, I assume most other states will have a similar object, what's the thoughts of everyone on this case? Is it really needed? (I assume it's just a category in the ABS data that's come across wholesale). Seems to me anything not in side a suburb boundary would be considered unclassified anyway? I noticed a small number of those in NSW and decided to ignore them and just put them, that might have been a bad idea ;-( LOL, Well I guess we just need to decide if a 'unclassified' suburb is appropriate or not. If we decide it's not we blow away the relation and problem solved :) Or if keep it should we somehow flag it slightly differently so that we know it's not an actual suburb called 'Unclassified', although there are weirder names around ;) -- =b ___ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-au
Re: [talk-au] suburb boundaries
On Sat, 07 Mar 2009 23:51:35 +1100 b.schulz...@scu.edu.au wrote: Hi all, It's really nice to see suburb boundaries popping up around the place, it just makes the map look that little bit more professional. Yeah it is isn't it, Franc has done some nice work. There seems to be some naming redundancy in the NSW data though. The previous nswgnb import (before my time, don't know the source or history of this) placed suburb names all around the place, and now the ABS suburb import is repeating the data. An example of this is here: http://www.openstreetmap.org/?lat=-32.96227lon=151.65078zoom=16layers=B000FTF What are people's thoughts about this? Should the data be sanity checked for naming consistency then the nswgnb node deleted? Or should it be kept incase some renderer doesn't understand the suburb relation? Like perhaps it would be easier for the mkgmap devs to keep the nodes rather than have them write code to make a node from the relation. You've hit an interesting point here, one I've thought about a few times without coming to any reasonable answer myself. On the one hand we have the case where we leave around we're making allowances for any program that for whatever reason doesn't support 100% of the OSM data structures. On the other we're having multiple version of the same data which have to some how be kept consistent creating more maintenance of the data. After many years working with a number of databases I've found every un-necessary duplication of data leads to headaches, but there's inevitably going to be software that gets out of date and people are going to expect the data to change for it rather than update the software :/ I think what the boundaries need (and this is where a spot where it being a relation comes in handy) is a way to make a centre node, if the centre node is there then it can be assume it will display the name of the boundary, otherwise the renders should display their own boundary. This proposed option seems to be close to that kind of thing: http://wiki.openstreetmap.org/wiki/Proposed_features/add_center_in_Relation:boundary Of course then we'd have to get the renderers to recognize that fact on not render a centre text for a multipolygon with a 'centre' (unless they already do that?) (This has the benefit of moving the name display to a place more appropriate for the suburb rather than the exact centre of the area, I'm sure we can all thing of suburbs where the demographic centre isn't the physical centre) -- =b ___ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-au
Re: [talk-au] suburb boundaries
On Wed, 25 Feb 2009 22:50:20 +1100 Franc Carter franc.car...@gmail.com wrote: I know very little about the rendering, but I would suspect not. each boundary is going to divide two suburbs and may of them are quite short - so I would expect that representing them on a generic map is quite difficult to do. Observe this: http://www.openstreetmap.org/?lat=-34.7875lon=138.6422zoom=14layers=B000FTF You can see 2 Parafields, 1 grey (the place= one) and 1 black (the relation based one). You can also see a black '5107' from a postcode boundary I've put in in that area too. So the answer is yes it will render. -- =b ___ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-au
Re: [talk-au] suburb boundaries
On Wed, 25 Feb 2009 23:33:09 +1100 Franc Carter franc.car...@gmail.com wrote: Nice. Do you want me to try to exclude some suburbs so as to not overlay the areas you have already done ? Nah, there's only about a dozen that I've got fully completed and they're all close enough together I'll just review what's there and merge the 2 sets of data as part of the manual review process. There's probably another 2 dozen that I've only got partial boundaries of so we'll need the abs data to finish them off anyway. -- =b ___ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-au
Re: [talk-au] Suburb boundaries - getting close
On Mon, 16 Feb 2009 22:09:15 +1100 Franc Carter franc.car...@gmail.com wrote: Ok, it seems my conversion script is now producing sane results so it's time to work out what the final output should look like. The first question that I think we need to answer is, how do we represent the data in OSM, there appears to be 3 options:- 1. Closed ways 2. Relations 3. Borders with a left/right tag My vote is for #2, and I'd be strongly against the use of #3 since it's essentially the system #2 set out to replace and is so dependant on way direction and making adjoining suburbs all match directions vs left/right will be painful. #1 is a fine choice in city regions but I think it will cause ways to be too large in country regions, it also prevents someone telling which suburbs a boundary way lies in. Then we need to decide on what tags to apply to the data. The raw data has three fields * STATE_2006 A numerical identifier for the state the suburb is in * SSC_2006An identifier provided by the ABS * NAME_2006 The name of the suburb, which may have the old name in '()' after it. So, my initial proposal for tags is:- * name=? (with any old name removed) * source=Based_on_Australian_Bureau_of_Statistics _data (ABS ask for this) * ABS:reviewed=no * ABS:STATE_2006=? * ABS:NAME_2006=? * ABS:SSC_2006=? The 'ABS' part is just a suggestion - It's a bit short for my liking My thought: Make it au:ABS:... that way it flags it as an Australian thing, and within Australia I don't think there's too many multiple uses of 'ABS' in this context :) We also need to decide where these tags go - nodes, ways, relations. And if we go for the left/right approach a decision on how to I think how far 'down' the tagging goes depends on how we want to handle the update every 4 years. - If we plan to do a point by point check each time then we probably need to tag each node with a unique ID number to detect changes. - If we plan to do more of a diffing of the 2 data sets and updating changes only then we can probably get away with just tagging the data to the ways. I think the 2nd option is going to work better for us in the long run (given how much adjusting the boundaries are looking to need anyway). Of course if we choose option #2 above then I think both ways and relations will need to be tagged, although the ways will only need the source= tag and the unique ID #. -- =b ___ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-au
Re: [talk-au] Suburb boundaries
On Fri, 13 Feb 2009 23:21:11 +1100 Franc Carter franc.car...@gmail.com wrote: I'll have a think about whether can I work out something clever to see how well postcode boundaries match suburb boundaries. I suspect I am not going to be able to process both the suburb and post code data together to get one nice import as the size of the data blows the memory on my machine away (4GB). I wasn't meaning import the data together, just that a few people taking the suburb boundary data and applying which postcode each suburb is in could derive the postcode boundaries from that relatively easily. Give some people who might not be able to get out and collect data something constructive to do :) -- =b ___ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-au
Re: [talk-au] Suburb boundaries
On Fri, 13 Feb 2009 14:44:50 +1100 Franc Carter franc.car...@gmail.com wrote: After some nashing of teeth and swearing I have script that converts the ABS data in to a set of non-overlapping ways with some minimal info on the ways. I'd like some volunteers who I can give some subset of the data to (name your subrubs/areas) to have a look over and see if it 'looks ok' (i.e correct enough and no pathological cases I have missed). Then, we can start making some more solid decisions about exactly what form we want the data uploaded in. I'll have a look at the Northern Suburbs of Adelaide if you like, compare them to the existing data I've put together :D (Elizabeth *, Salisbury * and Munno Para * if you need a list of names to match :) -- =b ___ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-au
Re: [talk-au] Suburb boundaries
On Thu, 05 Feb 2009 20:23:07 +1030 Jack Burton j...@saosce.com.au wrote: Consider two suburbs, A B, whose boundary is currently defined by a river. Now let's say that by the time the next ABS update occurs, that boundary has changed, and a small part of what used to be suburb A has become part of suburb B (it can happen). Since the ABS data contains only suburb boundaries (and no separate way for the river itself), and we're using multiple segments per boundary, and someone has helpfully merged that boundary segment with the way that forms the river (as I think you suggested earlier, to avoid stacking up ways on top of each other), there'd be no method for the update mechanism to know whether the course of the river itself has changed (and therefore so has the boundary segment, so it should move the way that defines both) or whether the river has stayed where it was but the boundary no longer uses that part of it (so it should split ways, create a new one, then add it to the boundary relation). This is an automated process, if it can be explain logically the computer can be made to do it. As I said before, as soon as any points are moved things become complicated anyway. If I were implementing this part of it (note Franc is only talking about a one-time import at this stage anyway, so we are talking somewhat theoretically): I'd uniquely identify each common boundary between 2 suburbs that we make a way. Use a diff mechanism to detect a change on said boundary, and look at the data, updating and adjusting a way that hasn't been modified at all and removing and replacing the way if it's been changed beyond the ability to adjust. With a single closed way around each suburb, the problem does not arise, since the update process does not need to care about the river itself (and should be clever enough to detect that another way uses some of the existing nodes, so duplicate those nodes instead of moving them). You fob it off so simply but there's a lot of work in your solution also. Following your example any time a minor change happens to a suburb it's likely to re-align every node on the boundary back to the original place, in fact it will most likely have to remove re-add the entire way since it won't be sure which nodes are which any more, someone could have added more, removed some, etc. You could tag every node I guess, but seems a lot of bloat for small gain, and similar gains would be made to the relation model with individual tags anyway. So we have the boundary solution which when a boundary changes only has to modify 1 shorter way along the common boundary between the suburbs that change or the way solution which most likely requires the whole way to be replaced on an update, possibly removing other adjustments made on other parts of the way. From this point of view the boundary solution requires less far reaching changes than the area solution. Of course any unique ID is risky anyway because it can be accidentally removed, but that's the risk I guess :D -- =b ___ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-au
Re: [talk-au] Suburb boundaries
On Thu, 5 Feb 2009 20:53:01 +1030 Cameron osm-mailing-li...@justcameron.com wrote: How much do suburbs change anyway? Perhaps any changes could simply be introduced manually. ~Cameron Yeah I did think that might be an easier solution, I was addressing automatic updates because jackb brought them up :) And as soon as we modify them in any way (align with road for example) I think it probably comes out the more appealing however we put it in initially. I know for example the SA Gazette tends to provide information about suburb boundary changes, probably before they make into the ABS structures. I imagine the other states have similar channels. -- =b ___ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-au
Re: [talk-au] Suburb boundaries
On Thu, 5 Feb 2009 14:26:13 +1100 Franc Carter franc.car...@gmail.com wrote: There are two issues that I have come across with converting to osm:- 1. What way do we want to represent the data, e.g closed ways or relations consisting of borders - something else ? I'd personally prefer border relations. But given Franc and I seem to be the only significant creators of relations in .au anyway (A search of the australia.osm reveals we're the only two with 100 relations) I don't think the majority of regular osm mappers have got relations yet. However I think relations are the way data like this is going in OSM. 2. The more technical problem that the boundaries are defined fairly precisely (or more accurately there are lots of points defining the boundaries). So the .osm file is very large - so eyeballing it in josm is not going to work. So I'm interested in people's suggestions of how we want to represent the data and on methods we can use to sanity check the data before we upload it. Lots of the cases are along roads/rivers/railways I imagine to make them align with what we actually have on the map, lots of review is going to have to happen once it's actually in the map anyway. Given nearly all suburb boundaries are multiples (one suburb on each side). I'd think 1 way for a common boundary between 2 suburbs and joining up all those ways for each suburb in a relation would be the way to go. Then people can review them in areas where there's existing data and re-align them down the middle of roads they run along or remove the chunks than overlap single ways and add those ways to the boundary. ___ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-au
Re: [talk-au] Suburb boundaries
On Thu, 5 Feb 2009 15:52:43 +1100 Franc Carter franc.car...@gmail.com wrote: From a 'philosophical point of view', I tend to agree that suburbs are made of a set of boundaries between adjacent areas. This was not how I did it in my first (very quick) attempt ;-( An advantage of having to sort out the legal issue means you get a bit of time to fiddle around trying out options before you get the full a-ok and import it ;) The data is in shapefiles that define each suburb boundary individually, so I'll have a think about how to extract out the individual borders (suggestions welcome) Hmm, so there's no real surety that 2 adjacent suburbs even share the same boundary? Perhaps then the single area option might have some merit from a 'getting the data in there' point of view or we write a convoluted script to correlate things... One question about aligning them that springs to mind is 'what should we align' - I wonder if the accuracy of the data is better than the average accuracy of a gps or yahoo imagery. That's a tricky question because it might be more 'accurate' because it might measure to an exact positional definition but is that useful or relevant to the OSM structure whereby a boundary down the middle of the road is more conceptually accurate Guess we have to get a small sample of the data into a city somewhere where we have plenty of GPS as a trial run (once we have the full ok). and see how it correlates to reality. GPS + Yahoo never correlate enough (at least in SA) to make it possible for both to be relevant :) ___ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-au
Re: [talk-au] Suburb boundaries
On Thu, 05 Feb 2009 16:29:39 +1030 Jack Burton j...@saosce.com.au wrote: 1. What way do we want to represent the data, e.g closed ways or relations consisting of borders - something else ? Closed ways (areas) - as that's how ABS define them, so it will make merging updated ABS data into the OSM Australia dataset (each time ABS update their dataset, which is presumably quite regularly) significantly easier. This isn't really relevant. Given the amount of data involved an automated process will have to be developed to bring it all in, so this process can just be re-utilised on any update. ___ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-au
Re: [talk-au] Suburb boundaries
On Thu, 05 Feb 2009 17:18:53 +1030 Jack Burton j...@saosce.com.au wrote: But I'm still not a fan of relations for suburb boundaries - even more so, now that we know that the authoritative set for Australia (the ABS data) is organised as a set of polygons (one for each suburb), since we'll presumably want to continue using this dataset for updates (e.g. when new suburbs appear, or old ones are split up/consolidated/renamed/etc.). This could be accomplished really easily if each item of ABS data was tagged with a source_ref:ABS (or whatever) set to corresponding object id from the ABS dataset (I'm assuming it uses such things - most large datasets do). And as soon as we edit that data in any way, such as you yourself suggest doing lower down in your reply, then updates from ABS are only ever going to be able to be imported as diffs - a straight 'update to these values' will break any adjusted edges and move other ways around. This will require extensive processing and either option can be handled at that time. You also assume the ABS is actually going to be (a) recently up to date (b) accurate, I'm not holding my breath on either, so blindly syncing with any changes they make is not necessarily wise. It still seems to me that the simplest possible set of data to define a suburb is the location of its town centre (a single node) and the outline of its boundary (a single way). [And we already have most of the nodes for NSW towns/suburbs in the OSM dataset, from another bulk import of government data - adding suburb areas from the ABS data would give us complete definitions for NSW, and the hard part done already for other states] I isolation this makes sense, when included with other features such as roads etc it makes more sense that if a suburb boundary runs down a road that road is some how flagged as part of the boundary. Stand alone areas make extra work correlating that kind of data. Also, consider the case of a user downloading a rectangular section from OSM (since I'd imagine most of us do that, rather than deal with enormous planet or country files), where a suburb boundary intersects the boundary of the rectangle downloaded: If we use the single way method, the OSM API will give the user the entire suburb boundary, even the bits that are outside the rectangle - so every suburb that has any part of itself within the rectangle will have its boundary fully defined within the user's osm file. If we use the relation method, only those segments of the boundary which have nodes within the rectangle will be supplied - leaving some suburbs with incomplete boundaries in the user's osm file. If the user doing the download is not prepared to handle the relation issue with respect to boundaries they will probably encounter far greater problems that just suburb boundaries. Multi-polygon relations for example will suffer from exactly the same problem. The issue is that the down-loader needs to be aware of the data structure and not make the data structure adjust to handle his in-competencies. For example in JOSM it's a matter of a 3 clicks to request all the ways of boundary. There are already issues of ways with too many nodes causing downloading problems for the OSM servers, a single area for a whole rural suburb (or one of the bigger boundaries like a council) is easily going to exceed reasonable limits of way length, and unlike a way where you have to download the entire way every time it's viewed, with relations you can choose to download only the relevant parts, and the whole lot if you need it. Should you happen to not have your download's bounding box cross any of the suburb boundaries with either method you may just end up with no suburb data at all anyway. Assuming you can rely on suburb data from a small are download is a little naive. The only situation I can think of where a relation would be necessary for a suburb boundary would be when one suburb exists wholly within the boundaries of another suburb - but we already have the multipolygon relation for that (and I can't think of a single Aussie example of this off the top of my head - in fact, the only one I can think of globally is Vatican City being wholly within Rome - although we do have a similar situation for state borders with NSW/ACT). It's also able to be handled by the boundary relation with enclaves and exclaves which are designed for exactly this reason. ACT/NSW being a prime example of exactly that. Interestingly there is now support in multi-polygon relations for outer and inner ways to be broken into multiple ways (mapnik and josm already handle them properly) rather than being single areas, further indicating that stacked ways is not considered the ideal solution to these problems. This can be done either way - since with the one closed way per suburb method, the nodes along shared portions of boundaries should be common anyway: a mapper fixing an incorrect boundary would still only
Re: [talk-au] Adelaide out of copyright street directory
On Sat, 17 Jan 2009 22:32:12 +1030 Cameron osm-mailing-li...@justcameron.com wrote: All, The directory is now available at http://www.osmaustralia.org/gregorysmaps.php Thanks to Matt White (Gaffa) for the hosting! It's a great resource for naming streets traced from Yahoo. When doing so, please tag your work as source = Gregory's 1949 (or more commonly, source:name = Gregory's 1949). This is of great help if someone else does a survey and comes up with a different name to the street directory. I like the way the index map says Gregory's Street map of Victoria. Encouraging on a quality contol level ;) -- =b ___ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-au
Re: [talk-au] Adelaide out of copyright street directory
On Sun, 18 Jan 2009 09:36:43 +1000 (EST) i...@4x4falcon.com wrote: On Sun, 18 Jan 2009 09:21:02 +1000 (EST) i...@4x4falcon.com wrote: Nick, My worry with the Old map + yahoo combo is that some armchair mappers may over-ride existing surveyed data (I've already seen surveyed stuff over-ridden by someone looking at yahoo a few times) maybe even as far as changing the name to the old map name. :/ But I think the overall gains that will be made in getting the map to a basic usable state so the 'masses' can operate with it, and provide the detail feed back is more important. Until we have that we're just making a toy for ourselves and we can strive to make it perfect but it'll be useless to everyone else. This is really simple. If the source=survey don't change it unless you have resurveyed it. Under no circumstances change source=survey data with traced data. Yeah, what he said, although perhaps the later line could be: Under no circumstances change source=survey data with traced data unless you have obviously newer and more accurate data. Just a thought :) -- =b ___ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-au
Re: [talk-au] Mt Barker - roundabout
On Tue, 6 Jan 2009 22:59:36 +1100 Nick Hocking nick.hock...@gmail.com wrote: There was a large roundabout missing in Mt Barker ( Intersection of Mann Street and Adelaide Road). Previously I have been unable to fix this since there was a relation on Adelaide road which made it impossible to split the road and add the roundabout without havinvg to delete the relation and re-edit it afterwards. However, I noticed, tonight that the relation has disappeared, so I took the opportunity to slide in the roundabout. (I had surveyed the roundabout a few days ago.). Does anyone know how to edit a road that has a relation on it. Which Editor are you using? Any remotely recent JOSM will happily split a road with a relation and leave both parts of the road in the relation which you can then remove 1 part of the road (if it's no longer appropriate of course) from the relation by removing it from the properties section of the way. -- =b ___ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-au
Re: [talk-au] Major road cleanup
On Sun, 21 Dec 2008 23:24:47 +1100 Nick Hocking nick.hock...@gmail.com wrote: Splitting the way should not effect the park at all Stephen, Every time I try to select the way, to split it, one of the parks get selected instead. I don't seem to be able to get to the road to add tthe two bridge end nodes, unless I move the parks out of the way. This entails ungluing them and then it is about impossible to get them back again. The street is Masson Street, running off Northbourne Avenue in central Canberra. Feel free to have a go at it and tell me how it's done. Also, (Don;t tell Liz or Darrin) but there's a mini roundabout there that I think should probably be a normal one, but I can't get at it either since it is also part of both the parks. Safe from me in Canberra, I limit my Editing to S.A. ;) Actually the the very problem you're facing here is the very reason why I ended up opting for separate parallel ways when doing things like parks, it just makes managing everything a lot easier despite the fact it's more work to setup. Like I said before in these cases there seems to be no real guidelines just preferences of people. I find (at least in OSM) the trick to selecting a way like that is to do a drag select across and area covering that particular way (and of course you'll get a few points and maybe a few other ways also) and then selecting the way you actually want to edit from the 'current selection' window. -- =b ___ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-au
Re: [talk-au] What gives with roundabouts?
On Mon, 15 Dec 2008 00:22:05 -0800 (PST) bluemm1975-...@yahoo.com wrote: So after realising this I can't actually stand in support of junction=roundabout on a point (or some other similar proposal) as a permanent fixture, but would fully support it as a 'temporary' tag to indicate at some point someone with my kind of island obsession comes along and puts in the details. I thought the same when I first started mapping, as I wanted to show centre pedestrian islands like in the Melways. But the wiki is very specific http://wiki.openstreetmap.org/index.php/Tag:junction%3Droundabout It says that normal pedestrian islands aren't meant to be drawn as two separate ways (flares). I guess you need to add a comment to the discussion page or on Talk mailing list to propose something different. Therefore given the wiki definition, there isn't anything gained by having 4+ nodes when compared to a point and some kind of diameter value. I think perhaps you miss-understood me, or perhaps judge my island obsession a little too extreme. I wasn't actually talking about those island which I personally don't draw until they're at least a car length or so, big enough to get in the way of a u-turn or obstruct you turning into a driveway off the side of the road. (obviously not right on a roundabout, but that's what I use as a rule-of-thumb idea about which islands to include). You say given the wiki definition there isn't anything to be gained, yet the wiki definition says: A standard size roundabout with up to four exits can be drawn simply using four nodes in a diamond shape. To me that sounds like the definition says a roundabout is four nodes. -- =b ___ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-au
Re: [talk-au] What gives with roundabouts?
On Mon, 15 Dec 2008 12:08:54 +1100 Ian Sergeant iserg...@hih.com.au wrote: + When you cross this kind of roundabout when cycling, or with a learner driver, you don't have to worry about the characteristics of the road you are crossing (since you never turn into the traffic of the cross road, you just cross the roundabout). This isn't just about cyclists and learners. Its about the nature of the intersection. I'm having quite a bit of trouble understanding your point here, what difference does whether it's a node or a loop-way have on the characteristics of the side road? What difference does the size of the roundabout have to do with this? Surely you *do* turn into the traffic of the cross road, twice in fact, once for each direction, once yielding, once with right of way? Again this happens for all sizes of the roundabout. (In fact I thought that was the point of roundabouts, to reduce the points of contact to a minimum and make the laws of yielding right-of-way very clear - and to slow you down whilst doing it of course) + It represents what is on the ground accurately. Often there is less of an actual diversion than many other traffic calming devices, which are not mapped. To draw it as a deviation in the road, just isn't what is there. The only other traffic calming device I can think of that this applies to is a 'chicane', perhaps I missed a few options? Humps and their variants cause no change in the traffic flow, neither do Chokers (all listed in map features), what others are there? Given nearly all small roundabouts occupy close if not the entire road width, using the intersecting roads carriageway as part of the loop, this means the average deviation is about 1/2 a road width out and 1/2 a road width back or nearly 1 whole road width. I'm pretty sure I haven't seen a chicane deviate more than a road width (that would put you on the footpath at some point), so it's really a close call here as to whether there is '*less* of an actual diversion than *many* other traffic calming devices'. + These have a very standardised appearance, and should be represented in a standardised way, like a template. The benefit isn't just in time-saving, but in identifying that all these roundabouts are very much the same. What standard things about a suburban roundabout are there that don't equally apply to a large roundabout? The only standard things I can think of about all suburban roundabouts are: (1) They go clockwise around a central raised island (2) All approaches to the road are divided by at least a smaller splitter island (3) They have one lane (4) They have a roundabout sign displayed on all approaches Beyond this they are as varied as everything else on the planet. The only one of those that doesn't also apply to much larger roundabouts is (3) but it applies to some much larger roundabouts, so it's a bit of a non-starter. If we look at the things that vary between suburban roundabouts we find the following list: (1) The structure of the centre varies wildly beyond being raised: - some have a garden bed, some are paved, some are dirt - some are a single tier, some a 2-tiered, some are 2-tiered with the lower tier being traffic-able. (2) Some have extensive street furniture in the centre, some don't. (3) The splitter islands may or may not have signage. (4) The outer edges of the approach roads may or may not has extra curbing added to tighten the road approaches (5) The roundabouts may occupy just the existing road surfaces or may extend out beyond them, and given the variety in road widths different approaches to the same road can yield radically different judgements by people. (the extended area is often quite large on the opposite side of a 'side' road in a T-junction) (6) Not all of these are even circular, some are elliptical in shape, and of course there's the classic egg-timer shapes. (7) Quite a number of circular ones have funny offsets on opposite roads which mean your have to adjust the centre lines of the adjoining ways to make it line up (suddenly map doesn't match ground any more). 6 and 7 at least can be handled by actually drawing the roundabout as a way I guess, but then it defeats the purpose of the standardisation idea. I'd suggest in fact that suburban roundabouts don't fit any kind of template at all. snip the rest for now because you raise a good point about the decision order, in fact I'd even go so far as to suggest (given jackb's disagreement with me on what it means) we need to make an explicit list of what exactly defines a roundabout, I think Liz's list at the top of her PDF was a good start, plus a no-parking addition). -- =b ___ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-au
Re: [talk-au] Wild guess surveying
On Mon, 15 Dec 2008 10:35:01 +0900 Andrew Laughton laughton.and...@gmail.com wrote: Yes, very easy to fix, and I have fixed other roads that were also wrong, the worry is, how many others need fixing and where are they. Maybe a polite message could solve the problem, or maybe a rough position is better than no position, and there is no problem. I think there's actually 2 issues you've hit on in this. One you outline here and the other is the issue of what the original author used as a source for the estimation. On the issue you have listed here I'd suggest at some level it would be a good thing to have rough estimations drawn in, at least for major features (which landsat can provide if nothing else), an empty block of map just doesn't help anyone at all really. Data can always get more accurate as time goes by as someone with more specific information refines the paths, much like you are doing in this case. When new data obviously over-rides older data in the map people should not have hesitation correcting things. The other issue is a potentially nastier one, especially given that landsat supports something approximating the traces you made in this case, I worry that your suspicions may be correct, or perhaps that the person who drew it in based it upon personal experience from a long time ago (dodgy source at best ;). I think a polite message suggesting that you are concerned about the source of his data might not go astray in this case. -- Darrin Smith s...@salseast.org ___ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-au
Re: [talk-au] What gives with roundabouts?
On Fri, 12 Dec 2008 09:10:32 +1100 Ian Sergeant iserg...@hih.com.au wrote: As is everyone - but we can't forget that a linear road is always going to be a representation of a 2 dimension road surface, and currently that is what we have to work with in OSM. If you were drawing the full road width in OSM, the road wouldn't actually deviate at all for a mini-roundabout, it would just be drawn within the width of the road. Mapping a 4-node deviation in the road for a mini-roundabout isn't actually what is on the ground, either. The question remains, how to best represent what is on the ground. But this interpretation is entirely dependant upon the width of the road, an older 5-6-car-wide inner suburban road can eat up a 4 car wide roundabout and make it 'mini' whilst a new suburb where they build roads only slightly wider than 2 cars it's have a smaller 2-3 car wide roundabout where the full structure of the road differentiates radically. Following your reasoning we would have the latter case represented at something closely approaching reality and the former case rendered as if the road junction is a big slab of asphalt. In addition, if we follow this reasoning we might as well throw out the bridge as a way concept for around 50% of all bridges, most of them a small foot track across a creek and similar, but also including largish bridges such as the South Road/Port River Expressway bridges in Adelaide. The bridges are only as wide as the lanes going over them and there are walls built up on either side right next to the lower roadway, so there's no bridge that extends outside of both roadway more than centimetres. So to map what's on the ground by this stance it should be a point. I plan on submitting a proposal for the roundabout tag, where you can add it to a node like a mini_roundabout, for use in simple suburban type roundabouts. Something like junction: inner_width=3mcould specify the island size, making it possible for pretty rendering. Weird intersecting ways or large roundabouts would have to continue as is. Oddly enough, these seems almost completely contrary to what Darrin is arguing, and aligns well with that I would like to see happen. I really don't care whether the tag is called mini_roundabout or something else, I think the junction is best represented by a single node. Darrin believes that it is better represented by have a loop. If I was to accept that your above point was valid (obviously I don't), then I would agree with BlueMM, because I believe there is a fundamental difference between the mini and a real roundabout (the presence of the centre Island, even if it's not marked on the map). I thank you Ian that you seem to have summed up my position better than I have, and that this discussion has clarified for me that the issue is very much tied to the presence of the island. -- =b ___ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-au
Re: [talk-au] *Round*abouts
On Fri, 12 Dec 2008 19:31:34 +1100 Liz ed...@billiau.net wrote: http://en.wikipedia.org/wiki/Roundabout#Mini_roundabouts Mini-roundabouts can be a painted circle, a low dome, or often are small garden beds. Painted roundabouts and low domes can easily be driven over by most vehicles, which many motorists will do when there is no other traffic, but the practice is dangerous if other cars are present. Right, 3/4 of that backs up what I'm saying. Yes it says small garden bed but further down in the same section it proceeds to eliminate it as a mini roundabout. Whilst it may be physically possible, it is illegal for vehicles like cars, which can turn around the mini-roundabout, to go over the painted island, or around the wrong way- vehicles should treat it like a solid island and proceed around it. (In practice, few motorists obey these rules). This backs up what I'm saying exactly. Notice how they talk about it being physically possible for a normal car to go *over* the island. I challenge you to find an example in any of the pics of the roundabouts we've discussed where a conventional car can cross them in any manner at speed and not be significantly damaged. When I read this I see that motorists may be found driving over these things, but it's illegal; it's dangerous Yes, so, it's illegal and dangerous to go through a red light, but we map traffic signals. It is very useful however to know one is approaching a roundabout which DOESN'T have a centre island for exactly this reason however since there are drivers who will break the rules it helps to be fore-warned to be a little more careful. A slightly larger version of a mini-roundabout, sometimes called a small roundabout, is designed with a raised centre surrounded by a sloped overrun area of a different colour from the roadway and up to a meter in thickness called a truck apron or a mountable apron. The wording in this sentence is inconsistent with the rest of the article, as further down in that same document it says: The centre island also MUST BE ABLE TO BY OVER-RUN BY LARGE VEHICLES. If this is not possible, perhaps due to plants, or street furniture it is considered a small roundabout NOT A MINI ROUNDABOUT and as such must adhere to the stricter roundabout guidelines. (My emphasis added) It doesn't say parts of the centre island, it says THE center island. A little flange is not the whole center island. And I fully suspect the roundabouts in the area where you put a mini in the other day that there was no flange there anyway, so it's a little academic in those cases. And did you know how the sentence I quoted indicates a small roundabout is NOT a mini roundabout. So the text of the document is doubtful at bet. I challenge you to find a picture posted by someone NOT furthering this lists mini_roundabout agenda that show a mini with a raised centre island. English wikipedia won't help you, neither will the german one, nor will the OSM Wiki. -- =b ___ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-au
Re: [talk-au] Talk-au Digest, Vol 18, Issue 5
On Fri, 12 Dec 2008 20:45:13 +1000 Thomas Schroeder thomas.schroe...@red-sofa.de wrote: Hi, I am running JOSM on a Pentium D machine with 2 cores [but only with 32-bit linux (Fedora 9)] and no problem so far. I would suggest to try the latest JVM / JDK. My Java is this: java -version java version 1.6.0_10 Java(TM) SE Runtime Environment (build 1.6.0_10-b33) Java HotSpot(TM) Client VM (build 11.0-b15, mixed mode, sharing) It was a triple core AMD Phenom, running FreeBSD and I was using the linux-emulated java at the time, which I think was the problem, one I switched to the native version it's been running tickety boo every since. Thanks for asking though :) -- =b ___ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-au
Re: [talk-au] What gives with roundabouts?
On Thu, 11 Dec 2008 22:56:55 +1100 Matt White mattwh...@iinet.net.au wrote: There are emergency boom gates across the EastLink and CityLink tunnel entrances (well, back aout 300m), used to stop traffic entering the tunnels in an emergency, or when there is an accident in the tunnel. I believe they wer put in after the Burnley tunnel fire a couple of years ago. Oh of course! That makes sense :) Should have thought of that myself :) In fact now that you say that I realise there are a few of them on the southern expressway here in SA. What we need now is a sub-committee to determine the terms of reference for the committee to determine the way roundabouts are mapped (with it's attendant oversight and audit committees). If we are lucky, I might even be able to organise some cake to help the ruminations ;-) Oh that gave me quite a laugh. :) -- =b ___ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-au
Re: [talk-au] *Round*abouts
On Fri, 12 Dec 2008 06:53:44 +1100 Liz ed...@billiau.net wrote: Starting again about *round*abouts The renderers accept two methods of drawing a *round*about. There is a quick and easy method ideal for where two simple ways cross. Somebody has decided that this is called a mini-roundabout, and honestly I don't think that this item actually exists in Australian road design. There is another method which needs someone to draw in four points *square*about if we are lucky; *quad*about otherwise; and can extend to *dodeca*about if someone insists on creation of a beautiful form. With a way indicating one way drawn around the nodes, we can have a roundabout junction for the renderer. The renderers note a collection of straight lines and we don't get anything *round* at all. JOSM makes awful circles if you request it to make a circle. Merkaator is a bit better but still makes deformed shapes. So the maligned mini-roundabout is the only one where current renderers make a circle on the map. This doesn't hold much water I'm afraid. Following this line of reasoning all the curves on ways are invalid since they don't follow the exact shape of the road. Besides as I said in another post this is completely a rendering issue, they could be adjusted to turn the 4+ points into a circle/ellipse anyway. You are still equating mini_roundabouts and roundabouts as exactly the same thing. Minis were defined as a very specific subset of roundabout which all the pages YOU provided say are small roundabouts that can be completely driven over. I notice you have failed to address this very specific argument completely. Darrin, where does this *core* idea come from? I had not seen it until you mentioned it. Ok, what's the very first line on the osm wiki about junction=roundabout: A roundabout is a road junction where the traffic goes around an island in the middle. *Around* an *island* in the middle. From the English wiki page about roundabouts also: Pedestrians are (usually) prohibited from the central island All vehicles circulate around the central island in the same direction The island is a no-go no-mans land. -- =b ___ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-au
Re: [talk-au] Traces of Naracoorte, SA
On Sun, 7 Dec 2008 22:25:25 +1100 Liz [EMAIL PROTECTED] wrote: On Sun, 7 Dec 2008, Matt White wrote: Cameron wrote: I wonder if it would be feasible to map the whole of Naracoorte for $400? To do it on the cheap you could camp at the showgrounds. Would need power for a laptop and battery/gps/dictaphone charging and I don't know if they have that. Bus (one way from adelaide) is $57.10 for an adult or $28.55 concession or you could drive there. Then bike or drive around for a couple of days. Could do the data entry there and then go out and recheck some areas, or could come home and possibly have another visit later. Does anyone know about tracing from arial photography? The town might have some which could be of use and save time surveying. I'd do it off my own bat, just for the fun of it, but if I go anywhere near Naracoorte, my clients end up finding out and bail me up for tech support or software feature requests or something... Naracoorte isn't that big - probably mappable in a weekend easily enough (perhaps 70 streets, plus POIs). Definitely a small place - two days of cycling would do a good job. Our Adelaide based student has gone to Orange for summer to work, so he can't be persuaded to go there until after Uni starts next year. You would still have time to trip out to the caves and map a bit out there as well. After Matt's original post I was pondering much the same idea, figured $400 would more than pay for petrol and a hotel going on recent trips I've done around SA this year where I've combined orienteering with a little OSM. Auburb, Saddleworth, Hawker and Meningie all only took me about an hour each so I can easily see someone wrapping up Naracoorte in a whole weekend given Naracoorte can't be much bigger than twice those 4 put together. Have a talk to your clients Matt I might even be interested in doing it myself if I could find the weekend to fit it in. I know from talking to people in Riverton and Saddleworth (mums place) that a lot of small towns really wish they had a usable map avaliable, so handing them a relevant pdf dump from osm made them very happy. Futher thinking about this had me wondering just how much latent interest for these kinds of maps there might be out there and whether one could setup a 'map a town' register where people pledge amounts to get the town mapped on OSM and then when it reaches a threshold to fund someone you call in the donations and do it. More of an issue here than overseas I think due to our very sparse population, if there's 1 serious mapper every 100,000 people then that's only about 200 to cover all of .au and 1,500 for the .uk. And it's so far for many of us to get to areas relatively unmapped. -- =b ___ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-au
Re: [talk-au] Edits in and around Mt Barker, SA
On Mon, 20 Oct 2008 21:22:20 -0700 (PDT) [EMAIL PROTECTED] wrote: Looking at your B37 Alexandrina Flaxley Rd roundabout, you don't need oneway=yes(it's implied), clockwise(just draw it in a clockwise direction), ref(roundabout's don't inherit route numbers, it's for when roundabout's have specific ref numbers [in Europe I think]). Can you explain why that roundabout wouldn't have a B37, given it's actually part of the B37 route? If you leave out the B37 then you're leaving a gap in the B37 ref's, surely that is inconsistent? They are relations to tie different ways together to form logically-connected ways, eg. routes. They are relatively new to OSM and I haven't played with them much, but others have added them in my area. The only strange thing I can see is that the motorway_link for the B37 M1 are added to the relations, which I believe is wrong because they are on/offramps, not the actual freeway that the route follows. Are you sure you're talking about the right relation? Since I created the M1 in relation in the last few days, and just checked it I can see I messed up and actually added a point by mistake (now fixed) but the ramps are part of the Princes Highway and South-East Freeway street relations (since they are parts of those 'streets'). * Lots of the area's don't need area=yes, like parking/schools/landuse etc. Is this yet another crazy OSM inconsistency? Surely any of those closed loops are implicitly areas? In fact I notice mappaint in josm tends to render closed versions of these as areas without any redundant area=yes tag, so I'm not alone in my thinking here. * I think the highway=traffic_signals on B37 next to Cornerstone College should include crossing=traffic_signals (or just crossing=pelican), as it appears to be a pedestrian crossing, see http://wiki.openstreetmap.org/index.php/Key:crossing. Perhaps a twig to the map features page with respect to this is need, I find the reading of it very unclear and had been assuming from how it's written that highway=traffic_signals is used for all traffic signals (since they all controll traffic)... -- =b ___ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-au
Re: [talk-au] Road numbers in Garmin
On Wed, 22 Oct 2008 07:03:42 +1100 Liz [EMAIL PROTECTED] wrote: On Tue, 21 Oct 2008, Darrin Smith wrote: But not all sections of the road in one state have the same code (Sturt Highway being exactly the case in point, it is NOT National A20 for the last ~5km between Gepps Cross and Pooraka. I don't know about the case around Mildura so I can't comment there. I can defintely give more examples in Adelaide, Grand Junction Road IS signed A16 at each end and 'National' A16 in the middle. so you have crazy bureaucrats too? Yup, I think they make it everywhere, like cockroaches.. :) -- =b ___ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-au
Re: [talk-au] Edits in and around Mt Barker, SA
On Wed, 22 Oct 2008 09:40:49 +1030 Kim Hawtin [EMAIL PROTECTED] wrote: Darrin Smith wrote: On Mon, 20 Oct 2008 21:22:20 -0700 (PDT) [EMAIL PROTECTED] wrote: Looking at your B37 Alexandrina Flaxley Rd roundabout, you don't need oneway=yes(it's implied), clockwise(just draw it in a clockwise direction), ref(roundabout's don't inherit route numbers, it's for when roundabout's have specific ref numbers [in Europe I think]). Can you explain why that roundabout wouldn't have a B37, given it's actually part of the B37 route? If you leave out the B37 then you're leaving a gap in the B37 ref's, surely that is inconsistent? i didn't put this roundabout in, so when i see stuff like that its hard to know what is the right thing to do unless we put it to the list. It's been there for a while, my question was more directed at [EMAIL PROTECTED] in this case, I personally think the B37 should be there as it is (I'm probably the one who added that tag, can't be bothered confirming it right now ;) since you have to travel through it as part of travelling along the B37. and i'm yet to find a decent resource for what roads are named routes. i suppose i need to make notes from the big green signs huh ? =) That's how I've generally been finding them, having a little interest in highway routing I've been looking around for resources and the BGS's are the most reliable source I can find, even the old copyrighted sources are a bit lacking in accuracy. Unfortunately there are still a number of areas where even the BGS's are lacking although I have to give TransportSA credit, travelling around SA this year I've found a number of places where brand new signs with routes have been ereceted in the last couple of years (since I was last there) so I have some hope that sometime in the future the route system will be pretty consistent. * Lots of the area's don't need area=yes, like parking/schools/landuse etc. Is this yet another crazy OSM inconsistency? Surely any of those closed loops are implicitly areas? In fact I notice mappaint in josm tends to render closed versions of these as areas without any redundant area=yes tag, so I'm not alone in my thinking here. I've seen areas around adelaide where they specifically have. it looks a whole lot better to see areas like commercial and schools marked in, its easier to see and the labels are marked up better. i don't understand how the renderer works, but putting in the area attribute helps for the mapper to identify what the thing is for. well it helps me get a better idea about what im editing anyway. =) ] Oh man! I feel like a wally :) I just re-read what bluemm was saying there and he's totally saying the same thing as me anyway! Can we just pretend I didn't write my original paragraph? :) -- Darrin Smith [EMAIL PROTECTED] ___ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-au
Re: [talk-au] diary entry with interesting visualization of users contributions
On Mon, 13 Oct 2008 22:08:11 +1100 Matt White [EMAIL PROTECTED] wrote: Darrin Smith wrote: On Mon, 13 Oct 2008 21:22:53 +1100 Peter Ross [EMAIL PROTECTED] wrote: http://openstreetmap.org/user/BlueMM/diary/3679 The site he got that from www.itworld.com as referenced in the image has some really nifty features I must say. Mised an 'o'... www.itoworld.com :) Wups indeed I did, can I blame the keyboard do you think? ;) -- =b ___ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-au
Re: [talk-au] boxes around cities
On Tue, 2 Sep 2008 18:21:54 +1000 Liz [EMAIL PROTECTED] wrote: could the person who drew rectilinear boxes around most australian cities explain the reasoning behind these boxes and their positions? the box for Canberra goes outside the State / Territory boundary which is an unlikely place for the city boundary i can't see the reason for these boxes could the author of them explain for me? Are these the Yahoo coverage boxes you are talking about? I noticed the one for Adelaide appeared a few months ago and confused me until I realised that's what it was for. -- =b ___ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk-au
[talk-au] landuse=* and highway=* intersections
Heya everyone, don't die of shock that there's a post to the list... Recently chatting with another OSM'r the issue of whether it was appropriate that landuse=* tags and highway=* tags should be completely seperate or if they can share common nodes in some places. Kinda came down thinking it really didn't matter too much but for neatness at the very least. We ended up wondering what other's opinions were on this? -- =b ___ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk-au
Re: [talk-au] Mapping Tee Place or Y Court
On Sat, 14 Jun 2008 13:40:33 +1000 Nick Hocking [EMAIL PROTECTED] wrote: While cleaning a whole lot of duplicate ways in the ACT OSM data, I noticed an unusual merthod of mapping a small section of road that is in the shape of the letter T. There seem to be 3 ways of tagging this road. 1) Tag the upright as a way. Then Tag the horizontal as another way, with the same name. 2) Do a way like an upsidedown L then another way for the left hand half of the horizontal or 3) Go up the upright, then left to the end, then back along the same track to the junction, then over to the right hand end. This results in having one continuous way but does involve having two bits of the way on top of each other and going in opposite directions. This might cause a routing program a bit of a dilemma. The same problem occurs with a Y shaped road where the top bits can be fairly small. What is the best method of tagging these and if I use methods 1 or 2 (as I have been doing) should I also add a street relation to tie up the two ways into one entity (also with the same name). I don't know what's the 'best' but I've been using method 1 for T roads and 2 for Y roads and a relationship to tie them together. But I'm only one person... I definitely think #3 is not the best option however. -- =b ___ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk-au
Re: [talk-au] Tracing items.
On Thu, 5 Jun 2008 14:38:20 +1000 (EST) [EMAIL PROTECTED] wrote: Hi All, Just a general comment and a couple of suggestions. Having been adding items into osm for the last year I've noticed people spending a lot of time tracing roads from landsat, yahoo, etc. While this is admirable work is this really a useful exercise. Someone will eventually drive along this road with a gps upload it and then erase the traced road. Give that some parts of Adelaide for example have got 2 years without anyone doing a GPS track on the, having somebody have at least put in the main roads with an overhead trace gives a framework to place things on. Likewise there are a number of people who contribute ONLY GPS tracks but no names or other details, and often don't ever trace their GPS tracks. If someone has traced from YAHOO etc and put the road name in then someone else can come along and combine the 2 pieces of information to make the more accurate data. Surely getting whatever data we can at whatever level we currently have is a better methodology? Surely even if an area were to be fully GPS traced and annotated people should still be occsionally checking things anyway and will notice things that slip by? Or worse still will not bother to drive down that road as it's already in osm and not tagged as survey and therefore not sure should they replace it. So it never gets surveyed accurately. Anyone with 1/2 oz of a clue can work out the difference between a yahoo/etc traced map and a gps track, for starters if you've uploaded you GPS track and made it public it's obvious. Then of course there's the source: tags. So my suggestion is instead of tracing roads, trace things that can not be easily surveyed. eg railway lines, rivers, powerlines. Railway lines are easy to survey, catch a train ;) Or fix the coastline. There's still lots of areas where the coastline is not accurate and needs to be fixed up. Anyway just my 2c's. And if someone enjoys tracing roads or some other feature exclusively are we to tell them to go away because it's not our idea of the best way to spend their time on OSM? -- =b ___ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk-au
Re: [talk-au] Kangaroo Island SA
On Thu, 5 Jun 2008 18:29:12 +1000 Liz [EMAIL PROTECTED] wrote: On Thu, 5 Jun 2008, Darrin Smith wrote: It's tragic to see that as usual with most of the people who dive in they haven't actually checked out the AU or SA wiki which make it pretty clear what should be primary roads in rural areas and have gone primary crazy... I did the B23, and if I didnt put tags on it I can tell you it should be source:yahoo. -- I've driven a select few SA roads in the SE corner (from my point of view) I waypoint each corner and try to record the name of the road, then when bored trace the side roads off landsat. Despite initial complete confusion in my head about what road is what grade, I think I've got it right at last. I wasn't talking about your roads Liz :) they seem to be pretty reasonable 99% of the time ;) From what I can see is your work you seem to go on the everythings unclassified unless it's special type thinking, whereas a few of our newer mappers in SA seem to follow a everything remotely main is a primary and as that flows down we end up with every 4th road a secondary and every 2nd road a tertiary. Makes a great rainbow, but not so useful as a street map. -- =b ___ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk-au
Re: [talk-au] Tracing items.
On Thu, 5 Jun 2008 23:25:06 +0100 80n [EMAIL PROTECTED] wrote: On Thu, Jun 5, 2008 at 5:38 AM, [EMAIL PROTECTED] wrote: Hi All, Just a general comment and a couple of suggestions. Having been adding items into osm for the last year I've noticed people spending a lot of time tracing roads from landsat, yahoo, etc. While this is admirable work is this really a useful exercise. Someone will eventually drive along this road with a gps upload it and then erase the traced road. Or worse still will not bother to drive down that road as it's already in osm and not tagged as survey and therefore not sure should they replace it. So it never gets surveyed accurately. So my suggestion is instead of tracing roads, trace things that can not be easily surveyed. eg railway lines, rivers, powerlines. IMHO Yahoo tracers should focus on lakes (and other bodies of water), parks and buildings. After that maybe roads with no public access. Tracing public roads should be a last resort, they'll get done much sooner using GPS if they are not already traced from Yahoo. I'm not saying that Yahoo tracers should not do roads, but please do lakes and buildings and stuff first. Give the GPS guys a chance to do the roads. It's a big planet we are not going to run out of unmapped space any time soon. If that's the case, why be picky about what anybody maps? Let those who want to contribute in their own way do it. It may get them traced quicker leaving them blank, but it gets them MAPPED quicker if someone does them via Yahoo/etc. By you argument you'd be happy to sit there with huge empty swaths of map whilst someone wastes time doodling buildings (what relevance those have on a street map is debatable, although they do look very cool) and not adding useful navigational data to the map. It's trivial to find out if an area hasn't been gps tracked by firing up josm and searching for roads without the source:survey tag if you want to GPS map an area. -- =b ___ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk-au
Re: [talk-au] new mapper with bad experience
On Sun, 25 May 2008 15:15:32 +0930 Andy Zivkovic [EMAIL PROTECTED] wrote: Hi all, A few weeks ago I found out about OSM. Actually, it was when someone posted about it on the Linux SA mailing list. Anyway, I checked OSM and noticed that none of the streets in my suburb were mapped (Hallett Cove, SA). So, I thought I'd have a go at doing it. Welcome Andy, I noticed your work down there the other day, good to see someone down south making some contributions :) Looks nice so far. So, this has somewhat dampened my enthusiasm for mapping, but I think I'd still like to finish off my suburb at least. I'll obviously be a lot more careful where I stop and will try to avoid places where I feel obligated to pull over as far as I can. Oh yeah, it's a challenge at times to park quickly and safely :) I've also found it's very useful to make sure you let people go past even if they're quite a safe distance away when pulling out because you inevitably end up pulling up again pretty quickly anyway :) -- =b ___ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk-au
Re: [talk-au] (LONG) Adelaide Highway Classification (was: Highway Classification Issues)
On Mon, 10 Mar 2008 16:53:49 +1030 Jack Burton [EMAIL PROTECTED] wrote: On Mon, 2008-03-10 at 11:47 +1030, Darrin Smith wrote: OK, to take this a step further I'll start the ball rolling in Adelaide: (As we get a consensus I'll write a Adelaide/South Australia Wiki page to reflect the decisions, I'm happy to do that) To date, I've been following the guidelines on the Australian Roads Tagging wiki page, specifically those for Urban Areas (since everything I've done on OSM so far has been urban). *SIGH* I can see we're not going to see eye-to-eye on this which is what I feared. My interpretation of what's vaguely defined on there makes me think you're not. This is not to say you're wrong just that you have a different interpretation than mine... (northern A15 being the classic example) That's what this thread is about I guess, getting the interpretation sorted out. Rest assured I won't change any existing definitions from now on until we sort this out. I propose that all A routes in Adelaide and only A Routes are labelled trunk. I don't think this makes sense. Here's why: Firstly, I'll get the two red herrings out of the way: if I recall correctly, the Gawler Bypass is signposted as A20, but it's actually a motorway. Similarly, the Port River Expressway signposted as A9 (and A13 for the part that used to be called the South Rd-Salisbury Hwy Connector Road), but also is a motorway. But I don't really think you intended the trunk definition to include those two. Ok, good point, that's a lack of my definition which should have included unless the route is superseeded by being near-freeway conditions in which case it should be a motorway or something similar. The motorway tag I think always dupes any other lower tag. Leaving those two aside, trunk routes (at least in urban areas) imply the big, heavily trafficed, wide, long, most significant roads within the greater metropolitan area. This isn't the definition as I see it from the wiki: highway=trunk. Metroads in the cities where they exist, or other similar cross-city trunk routes in cities where they do not. Going by how metroads are used in the relevant cities there are roads of lower quality that some of the roads you are proposing to eliminate labelled as turnk roads (southern end of metroad 3 in sydney comes immediately to mind). All the met-roads in other cities are about the cross-city nature of things rather than the quality of the road. I would suggest the only 2 A routes in Adelaide that don't fit this rule are the A22 and the A14. And yes the southern portion of the A15 past Norlunga Centre is another case that's debatable. However even though I don't think they deserve it, I think it's much easier to define it as all A roads and be able to display that than make a list that everyone keeps debating about. For example: I think everyone will agree that for Adelaide that includes: * The metro portion of Main North Road * The metro portion of Main South Road/South Road * The A21 city bypass route (with all its myriad of names) * Port Road * Anzac Highway as far as the coner of Tapleys Hill/Brighton Roads Actually I'm not sure about the tagging of A5 on this one as a side note. IIRC last time I was down that way the A5 stopped at Brighton Road (no A5 ahead at that point, and nothing at all at the next junction). So perhaps we need to check that's the case (I don't trust my memory enough on this one). and pull back the A5 there :) * Glen Osmond Road from the cnr of Greenhill Road to the Freeway Part of this road isn't even an A Route, so does this suddenly open us to defining other roads as trunk if we think they're busy enough? * The metro portion of Victor Harbour Road (at least as far as Aldinga Road, beyond that it's no longer within the urban problem domain :) * The portion of Fullarton Road that has an A reference This is one easily questioned for example, what makes it special/ sure the A1 goes along it but if Glen Osmond road is a trunk to carry all the traffic into the city, what purpose does Fullarton Road server except to carry a few wandered who didnt make the A17 turn past the city? * Portrush Road/Lower Portrush Road/Hampstead Road * North East Road * Grand Junction Road How far up? It's A to N.E.Rd beyond that it's not, I assume you mean only that far? or maybe futher? by the east end it's only 1 lane each way carrying less traffic than Lower N.E.Rd or Hancock Road (Source for this info later with another way of complicating things :) EDIT : I see you mention this lower :) * Port Wakefield Road Most of us will probably agree that it also includes: * Cross Road (A3) * Salisbury Highway/John Rice Avenue This is another one I though long and hard about trunking, after all it's right next to Main North Road raelly (2km seperating them)... I'm 50/50 on Payneham Road/Lower North East Road -- it's a road of importance (definitely at least primary
Re: [talk-au] Adelaide Highway Classification (was: Highway Classification Issues)
On Mon, 10 Mar 2008 11:52:00 +1100 Ian Sergeant [EMAIL PROTECTED] wrote: Anyway, rest assured this is a live debate. Check out http://wiki.openstreetmap.org/index.php/Proposed_features/Highway_administrative/physical_descriptions That's an interesting page, good to see people are addressing that issue on a global level. Should that proposal go through it automatically eliminates one of the options as 'valid' ... Shall have to keep an eye on it. discussion of reference definition, vs physical definition Ian's comments OK, in a sense it's good to see people as unsure as I am :) However, in rural areas of NSW, the system doesn't work so well. If you use the reference method, you will find that there are a handful of state highways, a couple of auslink roads, and that leaves 99% of all the roads without a reference classification. This would dramatically reduce the usefulness of the resulting map to use a reference classification. Most roads would look the same. Many main routes between towns have no reference classification at all. Perhaps NSW will one day get it's funding to finish the MABC roll-out, that might help a lot, but yes until then I don't envy your position. South Australia is a little better off with a nicely defined set of A B roads to guide things along. It would be nice if Australia had a reference system that would work comprehensively. It doesn't, and that leaves us always requiring a certain element of subjectivity. 3 States have (TAS, VIC, SA), 2 are part way there (NSW QLD), it's a start :) I would say - if there is a workable reference system for a particular area, then it is best to use the reference system, and make a correspondence to the OSM types. Document the area and the reference system on the wiki, and coordinate a discussion to ensure there is a consensus for that area. Right this is where I kind of got to with Adelaide, I guess my first email was a call out to start such a discussion so I'll change the subject to reflect that... If you can come up with a practical, yet unambiguous and objective, system for all of Australia, that would be great. Short of laying seige to the roads departments and councils, I don't think that is going to happen. Yeah, that's a good dream that one :) I'm sure if you have ideas for improvement, or a workable reference system for Adelaide, then you just need to convince people of the benefits, and update the doco. OK, to take this a step further I'll start the ball rolling in Adelaide: (As we get a consensus I'll write a Adelaide/South Australia Wiki page to reflect the decisions, I'm happy to do that) 1) Trunk Roads in City I propose that all A routes in Adelaide and only A Routes are labelled trunk. I can understand some hesitation from people with respect to the A22, parts of the A16 because they are low quality roads, but if we're going to tag to a reference pattern they need to fit. 2) Definition of rural vs city area I propose that the area bounded by lines joining Two Wells, Gawler, Birdwood, Mount Barker, Willunga, Aldinga and the Coast line are defined as City area, and that areas outside these are considered Rural (We can define other city areas around Mount Gambier/Whyalla/Whatever if people have definitions?). I think the current Rural definitions as provided on the Wiki are pretty close to spot on for these areas. 3) Primary Roads in City: There are about 5 B Roads inside the definition of the city area, otherwise there's a whole bunch of roads in the city itself which server the cross-city tasks the road definition suggests these should be. However I think there are currently way too many roads in Adelaide marked as primary which AREN'T serving significant cross-suburb purposes (Prospect Road is one that immediately comes to mind). I would like to suggest we the mappers of Adelaide draw up a list of Primary roads which are the only ones that should be marked primary. 4) Further levels: For later, a few steps at a time :) I'm particularly would like input from those guys mapping lots of Adelaide with me (jackb, justcameron, adhoc?) since you guys and I will tread on each others toes if we're not seeing eye-to-eye. -- =b ___ Talk-au mailing list Talk-au@openstreetmap.org http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk-au
