Disk space is cheap-ish (not at cheap as it once was, but blame Seagate or WD for that). Don’t really see why anyone would complain. (After all, if an edit is problematic, it can be reverted wholesale.)

For future reference, the correct procedure when you find bad data is to Just Fucking Fix It. If anyone gets in your way, remind them that while you admire their principles, they have a massive stick up their arse and they’re getting in the way of you Just Fucking Fixing It. Corollary: writing long diary entries, creating lots of diagrams, and Rule 7 violations in general also detract from Just Fucking Fixing It. (As does writing lengthy responses in the comments.)

If you’re looking to create an application using tiles offline, then consider redering your own or using another provider, rather than directly using the tiles that appear on our front page. They’re mainly intended as a demonstration, and as an aid to our mappers. Take a look at switch2osm.org for more information on this.

Entirely unofficial answer: I’d assume that they’ll be produced on the same basis as before, after the technical work for the licence change is complete.

As far as I can see, they are there. Where are you looking?

What nonsense. Keep reading. Scroll down a bit. Oh, look, there’s the proposed timetable.

(With apologies to marscot for clogging up the diary entry further.)

Er, it was in the announcement of the downtime.

Between Sunday 1st of April 2012 (no joke) 8:00am (GMT / UTC) and the morning of the 4th of April 2012 the primary database server will be switched to read-only mode so that we can bring the new server (ramoth) onstream, bought thanks to your donations. […] When service is resumed, we will begin the license change as a background process, redacting any data whose authors have not consented to the new terms.

Depends on what their source is. If they’ve imported it from a public-domain source, then it’s not theirs, plain and simple - that would be an example where it would be safe to odbl=clean it. If the decliner has actually moved things, or has traced from something else, then odbl=clean would not be suitable. Something that would have been nice was the opposite odbl=dirty to explicitly tag objects known to come from incompatible sources, though understandably this would have been open to abuse by TWSNBN.

FWIW, this isn’t licence-changeover downtime, this is install-a-new-server downtime. The licence-changeover work will start after that (though not necessarily immediately).

I don’t know if they’ve changed, but half the time my problem with MapDust was that the report didn’t tell you where the person was going - it’s all well and good saying the app gave you unexpected directions, but we can’t even begin to figure out what’s wrong if we don’t know where you expected to go.

“His point is the map key really needs to be updated. “

No, he’s referring to this:

Someone had drawn in a bunch of non-existent features in northern Greenland, presumably to see what they all look like. As for testing, we have api06.dev.openstreetmap.org, but it looks like it isn’t rendered. I can see why - it’s intended for people to test software against, rather than for looking at.

I’m sure that things like this are exactly why we have the test environments.

I believe in certain other places, the phrase would be "Oh boy, here we go ..."

Did you take the names from on-the-ground survey? As in, visiting the village and seeing what's on the signs? If so, make sure you mention this should you have to report it to DWG.

Unfortunately, I don't believe we can do a straight upgrade from BY-SA 2.0 to 4.0, since we'd still have to go through this process, and we'd still be getting the inherent flaws (non-applicable in the US amongst others, uncertainty for reuse, etc.). If anything, the problem with the process has been that it's taken too long. We should have cut it off after a year back in 2008 before idiots started importing BY-SA data.

A POI need not be a literal point, and there's no reason to insist on such. If other people's software doesn't cope with a polygon POI, then their software is defective. The only reason to not do this is in the case of a building containing multiple POIs, such as a hotel with a public bar, or a block of flats with multiple entrances leading to different address ranges. Otherwise we get two objects representing one thing on the ground, and an opportunity for the data on the two to get out of sync.

Something else to remember is that the Bing imagery may be out of date, so without some local knowledge, you may be tracing buildings that just aren't there or missing some that are.

Nice idea, but you'll need a new name, since we already have two osmxapis. The second one is sometimes known as jxapi, since it's a Java reimplementation of the original xapi, but if you're aiming for Overpass then you're implementing something different to the other xapis.

The picture at http://pathetic.org.uk/former/a40m_westway/ is the sign on the Gloucester Terrace entrance. Is it still there?

chillly: "Cyclists and pedestrians" is misleading. Pedestrians enjoy special privileges, cyclists merely think they do. The law you would invoke used to be the Road Traffic Act 1984, which allows local authorities to place such restrictions through traffic roders. The equivalent power for trunk roads in London is held by Transport for London, who will undoubtedly have a copy of the relevant orders for you.

There were definitely signs in the eastbound direction last time I was there, both on the mainline and on the slips from Wood Lane. Perhaps somebody forgot to replace the old NO signs at the other end?

Reverted. Cyclists and pedestrians are still prohibited on the Westway, motorway status or not.