Re: [Talk-GB] Historical Mapping in OSM

[Ian Spencer]

Dave F. wrote on 27/09/2010 12:02:  On 26/09/2010 23:47, Richard Palmer wrote: Dear all,     I'm involved in a project looking at the history of a street in London     (The Strand). One of things we'd like to do is produce a 'animated' map     showing the changes along it from the C18th onwards. I've had a go at     adding some attributes to OSM for building dates and then generated     different tiles for layers each century, (quick demo up at: Is this what we should be using OSM for? For me OSM is for *current* data (I've been deleting any demolished buildings). This project sounds great, but should be set up as a separate entity. The database would become far to cluttered otherwise. I've an abandoned railway in my area where sections of it are now invisible having been completely leveled & used as agricultural land. Someone's tagged these as railway=abandoned. I don't think they should. Dave F. I think there are a couple of issues: 1) The OSM database is not designed with dates in mind for any purpose, even current issues such as recording a temporary diversion. 2) It does not seem to provide a way for extensions to be readily made available through the main project. So, as it stands it seems to be something of a bastardisation to add dates as no renderer will be able to cope with a non-existent building that you have added, or different versions of the same version. My initial reaction is that it would be inappropriate to add dating into the core database as it has the potential to add a significant overhead to the general usage for something that might well be viewed as a specialist application. On the other hand, it seems that OSM should embrace anything that gathers mapping data into an open database, as it might generate the killer mapping application - something that a traditional mapping view might miss. A Google search quickly brought up another historical mapping project, http://forum.openstreetmap.org/viewtopic.php?pid=95872 so there are interests in mapping timelines. More generally, it would seem to be a good thing to have a central repository of additional data while not burdening the core application - thinking of, for example, walking directions to add into a routing application - rather than allowing this extra information to be on external applications which leads to fragmentation of effort. I'd also suggest that this should be discussed on the main talk group as it is not GB specific.

_______________________________________________
Talk-GB mailing list
Talk-GB@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk-gb