Mappo has commented on the following diary entries

Post When Comment
Another new MapRoulette challenge: Suspected Missing / Wrong One Way Streets 4 months ago

That's a really cool use of data to improve the map. But because it might need on-site verification perhaps it should also be fed into one of the QA tools that lets you scan your local area for (potential) problems too.

Intro 5 months ago

You might find the specialised "railway" maps here, useful:

e.g. mapping the different uses of abandoned railways (e.g. for bicycles):

Sochi was not mapped for Olympics 6 months ago

Is there no official data at all? I was impressed with the addressing data when people featured Sochi maps and had assumed it was an import from government held data. If not then that's all the more impressive.

Adding in missing roads from South Australian open data 10 months ago

For inspiration, there was at least a couple of projects in the UK that dealt with crowdsource assimilation of similar data:

Auto roads, part 3 11 months ago

Would it help the algorithm if the starting input where two points on a straight section of road (so all intermediate points would be valid input data too) rather than a single point? Perhaps even a series of straight line sections so the algorithm would only need to connect relatively short sections and would have a reasonable amount of human provided input to learn what the road looks like?

It seems like the algorithms and the editing user interface need to be designed together for maximum benefit. I keep thinking of the Photoshop-like smart select tools where they do an amazing amount of work for you but you still need to learn a few key combinations to get the most out of them by giving it the right starting data and then correcting it when it goes off course.

OS Locator update 201305 about 1 year ago

Is there any evidence that the OS are taking on board the discrepencies found and annotated by OSM by not:name etc?

Kids city map carpet over 1 year ago

Very Cool.

I would probably already have bought one if you could rotate the map within the rectangle to better choose the area contained.

Delete most administrative boundaries? over 1 year ago

I found a lot of the comments that started "OSM is ..." very annoying.

They generally took the form "OSM is amateurs, on-the-ground, collecting manually, non-imported, physical observable data" but even avoiding the import debate that seemed an incredibly narrow view.

Isn't "OSM" also the map renderers, the tile servers, the javascript libraries, the various editors, the communities, the wikis, the concept of being "open", the HOT teams, the various apps that use the maps etc. etc.

And going back to the import debate, if someone did collect all the 3rd party data into it's own layered DB to make it easy to incorporate into map output, wouldn't that also be "OSM"? Why would you not want it to be?

calculate ways instead of drawing or: a look at "average tracks" over 1 year ago

Surely there must be a bug causing this? It shouldn't really be possible for an "average" to wander outside the values like shown in these diagrams.

Or even multiple bugs,

e.g. sometimes the average seems to "cut" the corners, with two widely spaced points, even though there seems to be multiple points between them that describe a smooth curve.

Okay, I don't speak R, but after a glance at the github source I'd guess the reason for there being occasional problems like this is that one of the original tracks is chosen as the master track, based on which one has the most points. But those points are not equally spaced so the accuracy of the results varies with that single GPX track's point spacing.

It doesn't explain the offset seen in the second image, but it seems much more likely for that to be a silly bug too, rather than a fundamental problem with the idea of averaging tracks together.

Untangeling ways about 2 years ago

Oh, and where streams and boundaries are colocated, I've been putting them in seperately, as the invisible boundaries are imported from an external source and the river can be traced and changes over time without (I think) the legal boundary following it. It's a bit annoying trying to draw a trace over the top of another without it being automatically joined. Again, better tools seem like the answer.

Untangeling ways about 2 years ago

I think this is just one example of when mapping at different scales can cause you more work as you return put in more detail than the initial mapping.

Another example is a building being put in as a node, I've not noticed any simple way to convert that to a building outline and carry across the data and history (maybe there is, please point it out!).

My general feel is you don't want to make it more difficult for people to map the areas with low detail just to make it easier for those mapping in extreme detail, you need some kind of balance. And if you can improve tools, then you improve things for both mappers. That also takes work though so you're trying to balance developers, and mappers at different scales and different experience levels. Basically, it's a hard problem, so try not to get too annoyed at someone doing it "wrong".

First post! over 2 years ago


You mention adding surface quality will help CycleStreets routing, is there a wiki page or other documentation or QA tools (I'm thinking a bike focused type thing) for what you consider most useful or best practice? Everything I've found is fairly high-level e.g. adding missing streets. What things other than surface quality are important and exactly what tags will your software recognize?

I'd happily improve my local area, but am not sure how to go beyond the basics.

Potlatch2 over 2 years ago

Also, "traduire" is "translate" in english, not "traduce"

Good luck

Potlatch2 over 2 years ago

You need to extend this file:

So that is as long as these examples for English and German:

I hope that helps.

Anybody know what's happened to Matthew Somerville ? almost 3 years ago

Yeah, if you read that (and any related) thread there's all sorts of problems with the data they release that those receiving it point out. But since we'll be crowdsourcing some QA on it anyway, it's certainly better than nothing.

Anybody know what's happened to Matthew Somerville ? almost 3 years ago

Regarding the link provided by LivingWithDragons. Is there any way to add/edit metadata to the OSM database via a simple HTML form? It seems a bit crazy to require a Flash based map editor to add or correct (for example) a post boxes ref number.

Anybody know what's happened to Matthew Somerville ? almost 3 years ago

I was going to comment that the dracos database was just about to pass the 50,000 mark. I was vaguely aware that there's roughly 100,000 post boxes but I thought I'd google for the info.

It seems they've recently released more up-to-date info on the total number of postboxes (115,654 apparently) including ones taken out of service, more details here:

Project of the Week Pitch Problems almost 3 years ago

leisure=pitch seems reasonable english to me. Could it be a UK/US english thing. Don't they call pitches, fields in the US e.g football pitch vs soccer field?

On the other hand, I've never heard anyone say Tennis Pitch or Tennis Field, only Tennis Court, which may explain why people are less likely to use a tag that seems wrong.

OpenStreetMap Wimbledon. 6 hours to go about 3 years ago

Reminded me of this, a computer vision project that automatically finds tennis courts in satellite images:

Top of the stops over 4 years ago

Does the Naptan info not also have street names (and larger localities)? It would be similarly interesting to see which of those names are not currently in OSM. I suppose the data might need some massaging but if the data exists it shouldn't be impossible.