I support a policy of deleting “dead, never used” accounts, if only for the purpose of data cleanup. I’t becoming embarassing to see the “1M accounts” number being anounced, when it is really meaningless. The “nearby users” feature is a completelydifferent thing, something based on whodidit would indeed be better.

But deleting accounts needs to be very conservative. How about :

• The account should not have contributed any map edit, help question, diary entry, or anything else.
• The account should not have logged in in the past 12 months.
• After 12 months, send an automated email once a month warning about impending deletion (at this stage, simply logging in would reset the counter).
• After 6 warning emails, go ahead and delete the account.

We can argue and tweak those numbers, but I think that with those kind of guidelines, account deletion becomes reasonable. Now we just need to agree that it is usefull.

Well done :) Did you map all the routes by surveying on-board, or some other method ?

I’ll take a wild guess that by “background map” you mean “satellite imagery” ? If so, bare any strange config mistake on your part, you’re probably already using the latest Bing imagery available.

Microsoft kindly allows us to use its imagery, but OSM itself doesn’t include imagery. Aquiring imagery is expensive and time-consuming, so it is common to have to use old or low-quality imagery.

One thing you can do (depending on where you map) is to use other imagery sources. osm.wiki/Aerial_imagery lists a handfull. I suggest you use JOSM, which makes managing and setting up imagery sources more convenient.

You also have the possibility to use your own image files. Be they a photograph or a hand-drawn plan, as long as the licence is alright you can use anything. Again, JOSM has a handy way of loading images and scaling them properly for tracing.

Lastly, for questions like this, try the help center instead of your user diary.

I completely agree that there has been a lot of unjustified drama (mainly from the French, but from the DWG too) in this subject. A lot of heated discussion, with some arguments that are just opportunistic “debate ammunition”, but also some real core issues.

The single most important one in my view is that the reasons behind the “use a dedicated account for imports” rule are dubious. All other quibbles, the fact that it complicates contributing, the fact the rule started to be enforced only recently after years of happy cadastre workflow, the emotions about being contacted in a foreign language and be acted uppon by “authority”, the anger from endless threads where everybody seem to stand ground… They would all fade away if the rule had a clear benefit that matched the cadastre use-case.

However, I have yet to see a clear argument that shows that dedicated accounts must be used.

• “It’s an established rule” is self-referencing and irrelevant.
• “contributors should map highways first / the building that is a huge unwanted bloat / imports are full of errors” is off-topic.
• “It is needed for quality assurance and licencing issues” is very tempting but doesn’t seem to hold technically, because a changeset tag (for example) seem to provide the same functionality (to my uninformed self it even sounds better because simpler to use and easyer to automate). The DWG does a very important job and we should facilitate it as much as we can, but this doesn’t seem to be a good tool.

I have always seen the separate account thing as a semantic tool rather than a technical one. If this changeset isn’t really my work (because it is fully automated or because I am using unmodified data from someplace), then it is cleaner to use a different account. Not a requirement, just a nice thing to do. And the distinction between my work and work that needs a different account is a subjective thing. Maybe I want to distinguish the mapping I do for my day job from the mapping I do as a hobby ? I’d use a separate account for that. But you can’t create a hard rule out of that.

Harry, check that wiki page again, I added a link to the page that explains more toroughly how to do the building import, and it makes a real effort to be clear on the fact that this work is for experienced contributors and requires a lot of work. Maybe that page isn’t as easy to find as it could be (it better now, anyway), but on the other hand the french community has been monitoring for the tell-tale cadastre errors long before the DWG stepped in, and educates users when necessary.

It’s just a joke. Probably in response to the “ban potlach !” calls often seen on IRC (also jokes). C’m’on, it’s funny :) (and I’m a josm user).

I’d say the “does the road have a painted centerline ?” test is not a good way to distinguish residential/tertiary/unclassified, as there are plenty of counter-examples.

Residential is for roads that have a lot of houses lining it (surprise ! :p) and that are not otherwise significant for routing. Unclassified is the same without the houses. Once it becomes “significant”, tag as tertiary. Yes it gets subjective, but there’s no good way to avoid that. See osm.wiki/United_States_roads_tagging and osm.wiki/United_States_Road_Classification for the US.

That said, thanks for the presets tutorial.

Well spoted. I’m sure the incorrect attribution wasn’t intentional. Please go ahead and contact the popcorn devs.

And the best way to remember those details is to take a photo :) See also osm.wiki/Beginners’_guide

I do not think that the damage to OSM’s reputation is that great. I have seen very few comments pointing to OSM as the cause of Plan’s crappynes. Most complains point directly to Apple instead of the data providers, or point to issues which are not data-related.

The OSM attribution is last in a long list which you have to actively look for. Most people do not know any of the data providers listed here (or even know what a “map data provider” might be). Anybody saying that the Plans debacle is rooted in its use of OSM probably had a grudge against OSM to begin with. In that context, I don’t think that the clearer attibution you suggest would make any difference.

As a mapper, I’d love to know precisely where my work is used. But I’m affraid that making that a requirement would be too burdensome for most users of data mashups. I’d say Apple (and others) would not have used OSM at all if that was the case. So the current attribution guidelines are a good compromise, I think.

Another possibility is to tag the building with “Palazzo Pamphilji”, and to add a node inside it for the embassy amenity.

Thanks for the work, I’ve been meaning to move beyond plain mapping for a while and this should be helpfull.

I’m wondering would you consider making your overlay a bit more visible/official by declaring it in the standard layman overlays and/or merging it with the http://svn.openstreetmap.org/applications/utils/gentoo/ overlay ? Unless your plan is to get everything merged in the portage tree ?

Cheers.

The border-to-center line makes no sense to map, since the “arm” is always moving.

I’d also rather tag the farmland or the field itself, rather than the irrigation mechanism… But it’s true there’s not a lot of detail to add beyond what Sundance pointed out.

Indispensable quand on veut obtenir une coordonée précise :) Ça prend plus de temps mais ça vaut le coup. On peut aussi revenir (au moins 2h) plus tard pour rajouter une mesure et améliorer encore la fiabilité.

C’est toujours moins précis dans les bois, mais avec du glonass :/

Ton post ne dis pas si tu as utilisé la fonctionalité “average waypoint” de l’etrex pour créer un waypoint plus précis ?

If there’s any developers to blame, it’s the gatekeeper ones, for the misleading “error” message and the frightfull workaround.

The bot did the job it had to do, correctly. The problem is that many people did not fully grasp how that worked, thought “I accepted the CT, so my edits are safe”, and didn’t lookup cleanmap or OSMI to check the facts.

Sadly, Andew is probably not the last person to get surprised at the result, even though there was a lot of communication and a lot of extra time before the redaction. Not sure we could have done much better, the returns had diminished to nothing much at redaction time.

For what it’s worth, most people have cut their losses by now, and resumed mapping as usual.

The main osm website does have such a highlighting feature (here’s an example link for Paris), it’s just not linked to the nominatim searches. Sorry for the annoying FOSS answer but… patches welcome.

By your description, this road is indeed a ford, if only for a few hours each day. Even fords on rivers depend on the amount of water currently present (some riverbeds go from completely dry to flood-like during the season or because of a dam). The timescale is different but it’s the same issue. What counts as “passable” also depends a lot on your vehicle.

All in all there’s no hard rule, but if you can cross it under some circumstances, I’d say it’s a ford.

Definitely ford=yes and tidal=yes. The former is usually shown on maps and probably avoided by most router configs.

I’m tempted by some access tag, but usual time-based ones won’t do. There’s one use of access:tide=yes in taginfo but it doesn’t make much sense to me. access:high_tide=no or access:tide= sound better. You could always add an access=limited or access=tidal; most routing engines will not know what to do with the value, and simply avoid the road.

As for your actual question of the highway= value, just tag as if it weren’t tidal (the other tags already defined that). From the look of it and the tagging guidelines for France, I’d go with tertiary.

Oh… and this is not a waterway (as currently tagged). Waterways are for boats and such.