Talk:UK Mapping Priorities

From OpenStreetMap Wiki
Jump to navigation Jump to search

Discuss the UK Mapping Priorities page here:


What would be really cool is if you could specify the bounding box for a town, and a program found the file size each week, producing a weekly-updated league table of towns. I think this would really spur on mapping by making use of the human competative instinct! Of course we'd need some examples of how bounding boxes should be defined, so that the standard can be comparable. Daveemtb 10:53, 12 May 2009 (UTC)

As a step towards this, I suggest someone should go through and add bounding box data for every town in the table.
Do this by adding links using the new box=yes URL parameter. For example:
This data will help us treat each town consistently (particularly if one person does all the links) and later could be used to automate the process.
-- Harry Wood 16:56, 18 May 2009 (UTC)
Just seen this and have started adding some (see comments below). SK53 20:20, 14 August 2009 (UTC)

Local Knowledge

Some figures could do with updating by someone with better knowledge of what is included in the area, eg Dudley. Daveemtb 12:48, 18 May 2009 (UTC)

August 2009 Status

Various comments, based on a quick look:

  • Bognor Regis looks very incomplete, ratio in the table looks wrong.
  • Milton Keynes centre looks very complete, and although outer areas are missing (e.g., Fullers Slade) I think it deserves to graduate. Missing outer suburbs is pretty healthy for most UK cities.
  • Saint Helens also looks like a graduation candidate.

Scunthorpe is not listed, population 72000, file size 922kB giving a coefficient of 78. So probably should be added. SK53 20:20, 14 August 2009 (UTC)

Landuse Residential

As and when I try & work out the bounds of each place I try and put a naive set of landuse=residential polygons on the map. This has two primary purposes: 1) it makes it easier to work out where the place stops; and 2) it helps visualise how much there is still to do. Of course later on it can be refined and used to produce attractive maps. SK53 18:55, 3 September 2009 (UTC)

September 2009 Status & Changes

Working through adding the priority links stuff, I've been trying to make this more consistent.

I have already weeded out towns under 50000 and moved them onto a second page. In the fullness of time they may become our priorities.

I also think it makes sense to concentrate on towns rather than amorphous areas such as districts. I have already removed Kirkby-in-Ashfield and Sutton-in-Ashfield from Mansfield: this better reflects the wiki page on Mansfield (complete with cake slices), likely mapper interest (local mappers might not take kindly to their home town being lumped together with adjacent ones), and the ability to assess (and, indeed, make) progress. Another reason is to avoid areas which spill across major administrative boundaries. With the exception of suburbs of large cities where the urban fabric is continuous, this also enables the scope of the target to be kept in reasonable bounds.

On these grounds I suggest that Sandwell, Tameside be removed: West Bromwich is part of Sandwell and appears in the table, so this results in duplicate data. For the same reason I suggest that the focus for Rochdale be restricted to the town, not the metropolitan borough, and that Burnley appears on its own. Nelson and Colne may have a combined population to justify their own entry.

Lastly I have added Lincoln and Birkenhead. Lincoln qualifies on any grounds, but Birkenhead has largely been mapped from the outside in, so it is like a ring doughnut: lacking a centre. SK53 20:49, 4 September 2009 (UTC)

Would it be worthwhile moving any towns from these areas to secondary mapping priorities? --Wynndale 10:44, 5 September 2009 (UTC)
My preference is to focus on towns, so I'd rather remove the districts as stated above. As for other candidates for demotion, I think I've done a significant part of the mapping of Birkenhead in the past month and I wouldn't want to use it in earnest, even though I do sort of know the town. I'd be reluctant to think that a large town might be perceived not to be a problem when it actually requires at least one on-the-ground survey. It's why I added the town centre criterion. There's also a more or less complete lack of POIs. Alternatively we raise the cut-off to somewhere around 50, so that we have a few serious candidates to focus on. On the other hand lots of effort went into mapping Sunderland, but I don't think a single way has been edited in the core urban area since the mapping party, so I still think its a priority. SK53 14:28, 6 September 2009 (UTC)


I wondered if we could set some criteria for graduation. I would suggest both:

  • Town/city centre mapped to a good level of detail.
  • A co-efficient below 32.

I've checked the co-efficient of Crawley which is completely mapped and it comes out about 16, so 32 corresponds to 50% mapping. Further lowering the coefficient requires a lot of work, so many places will just linger on the list.

If we can progressively graduate a few off the list, then we might keep generating interest. SK53 22:10, 4 September 2009 (UTC)

The graduation level is going to depend on how much of the dump size is core mapping. --Wynndale 10:44, 5 September 2009 (UTC)
This is a good point, although a definition of core mapping is still required. I presume the starting point is 'road-completion'. Since I wrote the comment I've added a few UK towns to my sites on ITOMapper and its instructive to see how towns which have superficially similar levels of detail, may have a big disparity in the number of ways. Much of this is likely to be footpath, whether in parks, urban areas etc. We really ought to have a better measure than file size by now! SK53 14:28, 6 September 2009 (UTC)