David Earl wrote: >Sent: 13 May 2008 6:48 PM >To: Sebastian Spaeth >Cc: OSM >Subject: Re: [OSM-talk] [OSM-dev] Developers requested to help provide >"completeness" tools > >On 13/05/2008 15:35, Sebastian Spaeth wrote: >> Frederik Ramm wrote: >>> Once we have a few applications in place that get viewed by *many* >>> people, we could just have a button somewhere along the margin of the >>> page that says: "I know the area and what I see here looks correct". >> >> Given that this will be the default very soon ( :-) ), I'd rather have >> the notes API where people can click and say: "there are streets missing >> here, I know that". No warm fuzzy feeling, but more helpful in >> identifying weak spots. > >My main motivation in wanting this kind of facility is not so much to >help _us_ identify what areas need attention, rather to help our >_consumers_ know whether they can have any confidence in what they are >looking at. That's why I think there needs to be a very straightforward, >not overly onerous, but useful, metric, even if this has more levels >accessible to those in the know. It also means that wiki solutions just >don't cut it (I've been updating completeness pages for the areas I;ve >been doing since I started, but it doesn't help someone looking at the >map). > >Often anyone of reasonable intelligence can tell somewhere is not >complete because only the main streets are there, but I have come across >quite a number of places where a reasonable number of apparently random >residential streets have been done, and whose density would suggest to >someone who doesn't know the area that it is os plausible, when in fact >it may only be 30% complete (for roads and names) or less.
I get this same view. All too often I look at a place and think wow, that looks complete, but when I drum down into the data a bit it its clear that there are general gaps and the density of streets is not what you would expect. That's why I was testing out a completeness metrics method. But I agree with you, what we are really after hear is a simple way to convey a level of map usefulness and relevance to the user. Cheers Andy > >David > > >_______________________________________________ >talk mailing list >talk@openstreetmap.org >http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk > >No virus found in this incoming message. >Checked by AVG. >Version: 8.0.100 / Virus Database: 269.23.16/1430 - Release Date: >13/05/2008 7:31 AM _______________________________________________ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk
