It is obviously incorrect. I do not recall what I might have intended at the time so I have deleted the wetland key for that way. Thanks for identifying the problem and bringing it to my attention.

That was a careless error by me. I have now corrected it to barrier=cattle_grid. I do not recall any gates on that road and, looking at satellite imagery, it appears to be a cattle grid. Further, my abbreviation for cattle_grid was "ca" and I appear to have transposed the letters when I was entering that abbreviation. Thanks for bringing the error to my attention. It brings back a happy recollection of my drive on that road some years ago.

I had visited the general area around Deniliquin and Moulamein in the past and observed farmland, canals, wetland, waterways etc. but this particular location is not open to public access. So, like you, I have to rely on satellite imagery. Most of the farmlands in this area have irrigation canals carrying water from the various nearby rivers. The Bing imagery now available is probably not the same as that on which I relied when I mapped this canal. Today. I found Esri World Imagery (Clarity) Beta most useful. I think that the Esri imagery shows it to be definitely an irrigation canal although the width now appears to be closer to 6 metres, not 11 as currently tagged. I would look again at the various sources of imagery and it is probably more accurate to add an "intermittent=yes" tag. However I do make errors and I am happy if someone, better informed than me, makes changes to improve accuracy of the map. If I were mapping it today, I would map it as an area - I have found that "waterway=canal" tends to render canals as wider than is warranted, irrespective of "width" tags. Those are my thoughts but please go ahead and make changes if you think warranted. Thanks for posting the comment.

I saw recent discussion on talk-au list about this issue. I'm sorry but I cannot assist regarding speed limits. I have mainly concentrated on roads, farms, waterways and, while I have mapped some speed limits, I don't think I have travelled on this road again since my original edits and I cannot recall anything in this location. I share your concern about incorrect information being imported from an unreliable source.

It was my carelessness/inconsistency in spelling that started this problem. Thanks for your work in getting it right.

Hello again Warin
We recently exchanged thoughts about spelling of Toolebuc/Tooleybuc. I think we were both of the view that the NSW and Qld placenames were spelled differently. I know I was a bit confused but I think Tooleybuc with a "y" is the NSW spelling while Toolebuc without "y" is the Qld spelling. NSW Geographic Names Board has the "y" spelling. I do not have a source for the Qld spelling other than what I recorded when I drove on the Toolebuc Road about ten years ago. I think I was correct with the name of the road but recorded one or more farm addresses incorrectly.

Thanks.

No No No. I'm still confused. Queensland spelling is without the "y", isn't it? You are thinking more clearly than I am. Please fix my errors.

Yes, it seems I confused the NSW and QLD spellings. " Tooleybuc" with a "y" seems to be correct for Queensland. Thanks for picking this up. Please go ahead and correct my errors if you are willing to do so. Alternatively I will look but it will be a few days before I can get to it. Thanks again.

Fixed now. Thanks for identifying error and letting me know.

Thanks for acting on the nonsensical "beach" tag. I was feeling a sense of alarm when I saw this area overwhelming other features on the map. I think you are correct in pondering if i using KG classification works for OSM. It is a theory (possibly with some validity - I don't know) and the theorised area could be overlaid on the map for special purposes but it is not an observable natural feature and is not a legally defined administrative boundary so I do not think it is appropriate on the map. Yes we should map natural features but this area is not such a natural feature. I have visited some of the locations within the area and many are neither "desert" nor "sand". Examples are Lake Menindee, the Caryapundy Swamp, areas along the Barrier Highway west of Broken Hill, along the Eyre Highway in Western Australia, areas around Windorah and Boulia in Qld and many more. I saw some wooded areas, some water and wetland, some towns, a lot of scrub and some areas that I could describe only as "dirt". I saw some small areas of sand but perhaps only about one percent of the areas I visited. I agree that beach is incorrect but I also disagree with sand. The KG area can stay in the OSM database and would be found by anyone searching for it but it does not merit a single natural=* tag. I would like to see the individual areas of wood, scrub, water, wetland, sand etc each mapped in OSM - there would be millions of such small areas across the continent. But the KG classification area is not a single natural feature and I suggest removing the natural=* tag. I think you were right to remove the beach tag but I believe this very large area cannot be mapped with any single natural=* tag. I think others have considered how to map identifiable deserts such as Simpson Desert and Great Sandy Desert but have yet to find suitable ways to define these areas. And even if one could define the areas, no desert consists of a single natural feature. Thanks for removing the "beach" tag. I suggest also removing the "sand" tag and not using any natural=* tag for this area.

That's fine. I'm happy to help if I can.

Hello Ian
I have not retained any notes from that journey along Bullagreen Lane. I appear to have noted the road through the state forest to have been a private driveway leading to the farm Widgeree. I added the name of the farm and the address, 2000 Bullagreen Lane. I feel confident that would have been from my observations on the day. I do not seem to have been aware that there was a state forest in that location - probably no signage at that time. So go ahead and change the status of this road, as you see fit.

Ian, you are a careful and thoughtful mapper. Please feel fee to modify any of my contributions wherever you think appropriate.

regards

Michael Cleary

If you look closely at the NSW Base Map, I think it shows that the track is just beyond the boundary of the Conservation Area and is therefore on private land and probably a private road. The road runs through a wooded area but does not enter the conservation area, as far as I can see. I do not recall surveying this area. From what I mapped at the time, it looks as though I drove up to the entrance to 241 Tonninges Road, made some notes, and later added the private road and farms from satellite imagery. Look again and, if you still consider it to be open to the public, then feel free to make any changes you think appropriate. Thanks for the contact.

You are right. The section south of the gate should be unnamed and access=private. Dalmallee Road,the public road, is the not-so-well-maintained track to the west of the road that should be unnamed. I have looked at six maps and the NSW address location service. The farm Dalmallee's official address is 245 Dalmallee Road which would mean accessing the farm from the track. It looks as though the property owner has put in his own private road of better quality and joining Dalmallee Road closer to the Mendooran village area. Thanks for this contact - it rekindled a memory of a happy day driving in the Goonoo & Mendooran area a few years ago. I think that was probably just before we were given access to the Base Map. When I drove down the road, there was no indication that Dalmallee Road veered onto the track or that the well-made road became private. But I can see it now when I look at the map.

Thanks for the feedback. I'll look further at the landform tag - I do recall that reservoirs used to be tagged as landuse but the majority view was to change to natural=water+water=reservoir. While water is a naturally occurring substance, it did seem to me that reservoirs were more appropriately regarded as landuse. However whatever classifications we have, there will always be grey areas or overlap.

Thanks also for the link to the directory of wetland areas- I have browsed but will go back and read more thoroughly.

Thanks again. I appreciate your good mapping in rural areas.

Sorry that was submitted before I finished editing it. The other thing I wanted to add is that I am not familiar with the "landform" tag.

As I said, I am still learning. I have studied a lot of satellite imagery, I have visited many of areas I have mapped (where they are accessible to the public) and I have consulted the NSW DCS Base Map. Much of my mapping relies on judgements and I am comfortable where a more knowledgeable person is able to improve my contributions.

Regards

Hello Warin
Looking back, I think the Bing imagery that I used must have shown the area completely covered in water with no sign that it was intermittent. Now however it is clear from various imagery sources that the area is not always covered with water, so I agree with adding the intermittent tag.

The accuracy of mapping water areas in the DCS Base Map seems variable in different areas and I think satellite imagery can also be helpful. Accurate tagging really requires knowledge of how much water lies in the area and for how long. To me it is a matter of judgement and additional information can inform further judgement. Please feel free to modify the tags if you think it appropriate,

In regard to "floodplain", I think that is unlikely here. I am an amateur and still learning, but floodplains seem to me to be areas where the water is flowing and eventually drains from the surface, albeit leaving the sub-soil quite soaked. I think the area that we are looking at here is sort of basin that fills with water - an endpoint where water fills until it soaks in or evaporates - like the nearby Lake Altiboulka. The difference seems to be that Aktiboulka gets more water and is filled or partially-filled most of the time.

On further investigation, I retract some of my earlier comment. The NSW Base Map is sometimes obsolete and, if there is conflicting information, the other sources are preferred. However the DCS NSW Address Location Service is usually very accurate and, in this instance, it uses the names in the NSW Base Map. I have therefore left most names unchanged on the basis of information from the Address Location Service.

Some of the name changes in this changeset were on roads that I had surveyed some years ago. I see now that the name changes are taken from the NSW LPI Base Map. This map has some obsolete names and features that were once correct but the map is sometimes not updated when changes are made. Where there is another source of information, particularly a more recent source such as survey, the other or more recent source should be preferred. I propose to reverse some, and possibly all name changes where the name has been "updated" to an older obsolete name still shown on the NSW LPI Base Map (now referred to as DCS NSW Base Map). Please advise me if there is any more recent source that would support retention of the names currently shown for these ways in OSM.