OpenStreetMap
I've been making a push recently to seek out postboxes in East Kent. CT1, CT2, and CT5 are already shown as complete on the Dracos website. Now we've only got one box left to find in both CT3 and CT4 before each of those is complete as well.
Finding some of the boxes has been particularly tricky as some the descriptions are pretty useless. For example one of the CT4 boxes is described as "Stone Street". To anybody local to me that means the Roman road of the same name. Pretty much the whole road is within CT4, which covers a sizeable portion of the East Kent countryside. Basically the description narrows it down to a search of about 8 miles in length of road, or about 80% of its total length. Now that's helpful.
I've mainly been concentrating on the CT postal sector, but also straying over into the neighbouring ME and TN sectors from time to time. Today I'm pleased that we've reached a milestone of sorts for CT: 2/3 of all of the boxes have now been found when comparing against the Dracos data. That equates to 708 boxes.
In searching out some of the remaining boxes I've found that a few typos have unfortunately been made in the refs for I'd mapped a long time ago. Thus they weren't showing up as mapped when the Dracos updates from OSM data each Sunday. Some Postgres queries quickly uncovered candidates for these though, followed by a trip to the boxes themselves to confirm their correct refs. A further Postgres query reveals that there are a few boxes elsewhere in the UK with duplicated refs. So, if you live near one of these boxes then it may be worth another check:
CO13 70
DG11 40
E17 35
E4 4
IG2 70
IV1 113
KT10 173
N21 18
N22 6
RG21 210
RH14 81
S40 6
S61 303
SE1 36
SE25 4
SE9 50
SW15 3
WA8 5
Now that we've got reasonable approximations to the outward postcode sectors in East Kent I've been able to make decent use of it to visualise the distribution of the membership for the cycle campaign group that I'm on the committee of. It's really useful having the data available in a compatible licence. Now, if only the Royal Mail would realise this too and open up their data...
Discussion
Comment from Vclaw on 9 November 2009 at 19:11
FWIW, for IV1 113, I mapped those.
Checking my photos, its two postboxes next to another, both for franked mail - one 1st class, one 2nd class (plus another box for stamped mail, with a different ref). Photo here: http://www.flickr.com/photos/vclaw/4089764925/
I'm pretty sure both boxes were IV1 113. Though checking my photos, the ref on one is partially hidden, so it could be different?
IV1 is complete anyway, according to the Dracos list.
Also, I've found that the refs on the Dracos list don't always match what it says on the postbox. A few seem to be swapped about.
(and as for useless descriptions, there's one postbox in IV4 that is only described as "P". I've not managed to find it yet...)
Comment from Gregory Williams on 10 November 2009 at 12:30
I find that sometimes that a new adjacent box has a "1" appended to the ref, i.e. IV1 1131 in your example. Anyway, we've found a perfectly legitimate example of where one of the refs can be duplicated, thus showing why it's never safe to make assumptions about apparent errors.
Hmmm, now that description of "P" is really bad! Add to that the fact that it looks like IV4 is a huge area and looks like you'll have a fun task on your hands tracking it down. A good excuse for just mapping the whole of IV4 I suppose...