Hallo Frederic, leider zerstört dieser Commit ziemlich viel Küstenlinien. Zu viel, als das ich das kurzfristig schaffen könnte zu reparieren. Vielleicht hast du ja andere Möglichkeiten, es gerade zu biegen. Der Checker von deiner GeoFabrik wird dir daher morgen viele Fehler anzeigen. Ich werde die jetzt erstmal ignorieren und nächste Woche wieder reinschauen... Was war denn an dem Import falsch? Zumindest was die Küstenlinien betrifft, was zumindest heute am Morgen noch alles konsistent. Grüße aus Wolfenbüttel... Andre

The issue was, that node 5862033940 was a double. Fixing is easy... Just load it by JOSM and check it. JOSM will complain about the error and can automatically fix it. Detecting these issues is a bit more complex. I have a set of scripts doing this... Simply to say, just scan the replicates for those double nodes.

These issues are ONLY a problem of the editor program used. So, unfortunately, there is nothing to learn except using a better editor program.

Please keep coastlines consistent! They may not overlap neither have open ends. And the position of the line is position of high tide. A pretty good description of this is at the german wiki page of natural=coastline. Since it seems that you think, that my correction was not good I will not correct it again to avoid edit wars. But other mappers will surely do it...

Please learn that land is always on the left hand side of coastlines.

Please be careful adding natural=coastline to existing lines. Coastlines may not overlap.

Yes, exactly. These dupe nodes are nearly always due to a faulty editor tool. I do a lot of these fixes all day and it seems that all editors do this. To compare the versions you could load the way with JOSM and use the history tool (ctrl-h) to visualize the difference.

These loops or dupe nodes are nearly always due to a faulty editor tool. So no, you didn't do anything wrong. Likely...

Yeah! That's it! Looks great now! Thank you!

Please learn how to draw and tag coastlines before doing so. Most importantly they may not overlap! And the line is high tide. A good description is at the german wiki page of natural=coastline. I've fixed this once and will not do it again. If it remains like this, it will not render correctly.

Hi! As I see these kind of errors very often and in nearly all cases the Id editor is involved, I believe that it is very likely a bug in this editor. As I don't use it I also can't write a bug report. May be you can...?

Well, no problem. I do consistency checks for coastlines. If the coastline here has contact with the coastline of the sea, then go ahead. I do not have a problem with it. But as soon as there is a weir or a watergate then this is disconnected and should be water only.

YEAH! GREAT! It's sooo funny reading your comments that I now have laughed loud. No, really... You must have used REALLY GREAT QA tools which did NOT detect the error you created and lasts for nearly TWO DAYS! That's perfect! Especially for people who live in a perfect world! These STUPID QA tools which detected the error MUST be wrong! GREAT!!! YEAH!! Give me MORE!! Especially that I HAVE TO READ the docs for the QA tools I have written myself! Funny! I want to laugh! I'm really looking forward reading these silly things! It's soo sad, that I'm away for this weekend. But soon it is sunday evening and I can laugh again, I hope...

Well, this is EXACTLY the answer, which I was expecting from you... Sadly... You are just ignoring the facts and it seems that you are trapped in you own thoughts. It is absolutely IMPOSSIBLE that I made a mistake, because I did not edit anything in this area. Just because of a case two weeks before I was so prescient to not destroy facts. The last major edits were YOURS. And as you could have seen yourself, there in FACT were a duplicate way, which you deleted TWO DAYS later after I was complaining about this. Complaining about wrong software and slow hardware is simply NONSENS. And I NEVER have said, that QA tools operate instantly. This is also your brain telling you. The only QA tool which is REALLY operating VERY fast is mine. It takes exactly five minutes to check for coastline errors for the whole planet. You believe that you are without fail, but you are NOT! Just accept it! You cannot blame software which is working correctly and you cannot blame users about errors when they did not edit anything. That is a tactic to deviate from own problems. You should solve them instead!

The discussion here is in the changeset which introduced the error and it was YOURS. But as I said before. I think you are trapped in you own thoughts. The proof is that you are now talking about slow hardware and about generating tiles which was NEVER a problem. We are talking about DATA, nothing more.

Your problem is that you simply do not agree that you made a mistake. The error was yours. JOSM also should have warned you before upload. And you either have deactivated the JOSM checker or ignored the warning. But I'm pretty sure that JOSM is also wrong, right? The tools are working absolutely correct. As you have now deleted one way the error is gone. My tool and OSM Inspector do not complain anymore. But it is soo easy to blame tools instead of start thinking. I will ignore you now. You are just stealing my time. Hope you learn. *EOD*

It's simply WRONG, what you are writing! The way id 686907210 was created at 2019-04-30T14:45:16Z by YOU. And just that you believe (which is the opposite of knowing) that tools like Overpass or OSM Inspector have outdated data does not mean that the data is not there. The data of Overpass is normally not more than two minutes behind. You have now deleted way id 404462767 which is good. But I still wonder about that wilful ignorance. Tools are tools and not more. Using a tool to find problems is GOOD. If you than download current data and the problem is REALLY there than you can fix it. There is NO reason complaining about these tools just because you just want to ignore YOUR faults. I've now learned that even if a user has many edits does not mean that this user is REALLY experienced. That's sad.

OSM Inspector has also detected this error:

http://tools.geofabrik.de/osmi/?view=coastline&lon=70.19559&lat=-49.69507&zoom=13&overlays=coastline,coastline_error_lines,line_not_a_ring,line_overlap,line_invalid,line_direction,questionable,coastline_error_points,unconnected,intersections,not_a_ring,double_node,tagged_node

You can use overpass to download only lines with natural=coastline even for huge areas. JOSM offers a wizard for this.

The tool I use is a own written program which downloads the very recent day and hour replicate files. So I can see coastline errors not later than one hour after they appear.

Line ids 404462767 (which already existed) and 686907210 (which you added) both have natural=coastline and overlap!

Again, there are duplicated coastlines! WHY?? You could have used and refine the old one. Why creating a new one? This time, you should clean up this mess! I will not. This also means, that I completely stop cleaning up coastline around the world for several days. Great job!

No! It's YOU who is creating BAD or INCONSISTENT coastlines!!! STOP DOING so!!! If you want to improve coastlines the ONLY refine existing ones or split them up.