SomeoneElse's Comments
| Changeset | When | Comment |
|---|---|---|
| 39249542 | Hi,
Personally I've been suggesting for some time that places like node/448554221/history should have "name:sq" set to the Albanian name, "name:sr" set to the Serbian name and "name" set to the name used by the majority of the locals (which in the case of Fushë Kosovë I presume would be Albanian) - so thank's for adding "name:sr" there.
|
|
| 39206684 | Actually, a photo from the other side would make it much more clear - the pipes would be clearly visible :) Seriously though, please don't make "armchair" edits on the basis of "Mapbox User Feedback" without doing at least a smidgeon of due diligence. There are at least two very active local mappers in this area and the person who originally mapped the bit of the "canal" under the road (me) is still active. Where there are active local mappers, it's always preferable to contact them (either by a changeset discussion comment or a map note) rather than just guessing. |
|
| 39182096 | Er, "move to locator" should be "moved to description" of course, and there should be a "source=survey" on the changeset. |
|
| 39175944 | Thanks for fixing. This has happened a few times from this account; I've commented on changeset/38517811 asking why. |
|
| 38517811 | Hello,
Bonjour,
|
|
| 39121288 | Also, this is a large changeset covering most of Europe. It'd be much easier for mappers to check changes local to them if it was split up regionally. |
|
| 39121288 | Did you contact the previous mappers of way/146503489/history to explain why bridge=swing was "wrong"? If not, then you can be sure that they'll continue to add more bridge=swing elsewhere. |
|
| 39099539 | Hi,
Also, this is a huge changeset - it be much easy for local mappers to see what affects them if it was more geographically localised. The Welsh end, relation/3492463/history , does indeed look like a change that needed making but seems to be nothing to do with self intersections. |
|
| 39097756 | As I said on changeset/38959528 the day before you added this, the changeset description you've used doesn't explain what you did or why, and the changeset is huge. Please do use localised changesets, and explain in the description what you changed and why. |
|
| 39069408 | Please don't just say "I'll be more careful" and then just carry on as normal the very next day, as you did in changeset/39105735 . If you're connecting two roads, such as way/302319853 and way/226723933 in this example, restrict your changeset to just that area, and add an informative changeset comment, not some #gibberish. Here "Joining an access road to houses set back from Markfield Road to Markfield Road itself" would have been a meaningful comment. In my experience based on these sort of edits locally, every edit made by Mapbox remote mappers needs to be checked by local surveying mappers; it helps us to do this if your edits are confined to one local area and have meaningful changeset comments. |
|
| 39117825 | River and stream data is an interesting case where all sources need to be taken into account. In your example way/45991154 is from NPE and is obviously ripe for improvement, but rivers move or are moved, and in many areas it's clear to see from the Bing imagery and GPS surveys that the "official" data is wrong - the stream is no longer where it was. However, OS OpenData is really useful in helping to align Bing imagery and GPS traces (and in the absence of GPS traces in upland Wales I'd generally trust its alignment over Bing). From experience the folks at the OS can also be somewhat "optimistic" when thinking that something is actually a stream. |
|
| 39117825 | I don't believe that this boundary, as imported, already has been "generalised to some extent". See node/4170092738 , node/4170092739 and node/4170092740 . The existence of those three so close together suggests a case of "garbage in, garbage out" rather than any genuine "level of detail". |
|
| 39117825 | @shaunlewis you need to have the discussion with the wider community via talk-gb, not just two people who happen to have noticed this changeset discussion. Did you read osm.wiki/Import/Guidelines ? |
|
| 39117825 | @shaunlewis The boundaries that you have uploaded seem to be grossly overnoded. Please don't do any more data imports without discussing with the wider community first. For imports in GB, the best place to do that is the talk-gb list (https://lists.openstreetmap.org/listinfo/talk-gb). Re "too many nodes", just zoom in anywhere on the new boundary relation such as at osm.org/#map=19/51.82530/-3.28319 and look at the number of nodes in the new relation. No attempt seems to have been made to consolidate immediately adjacent ones. |
|
| 39117825 | @The Maarssen Mapper Not quite - We still have relation/357283 (old boundary, all the tags) and relation/6194617 (new boundary, many fewer tags, possibly too many nodes). |
|
| 39117825 | Oops - looks like you've already created the new relation relation/6194617 in changeset/39148379 . I'd still suggest to go back and discuss and correct the things that I mentioned above.
|
|
| 39117825 | OK, done. It took 3 goes (to 39152803) but eventually got there OK. When you're ready to think about trying again, there are a few things you'll want to think about:
|
|
| 38466231 | Previous tagging was not a "mapping error" - see changeset/28049400 for details. |
|
| 39117825 | I'll have a go at reverting them... |
|
| 39117825 | The normal approach for large areas two big to handle via one large way is to have lots of ways as type "outer" (see relation/2176657 for example). Re the "fill colour" that's a decision for individual renderers, of which the 5 different map styles on the osm.org site are only examples.
|