[OSM-talk] [OSM-dev] Thoughts on an enhanced GPX api

René Affourtit raffourtit at gmail.com
Tue Jul 28 12:43:09 BST 2009


On Tue, Jul 28, 2009 at 12:33 PM, Ævar Arnfjörð
Bjarmason<avarab at gmail.com> wrote:
> On Tue, Jul 28, 2009 at 9:04 AM, René Affourtit<raffourtit at gmail.com> wrote:
>> So when a junction is reconstructed a local user can place a bounding
>> box over that junction and all GPS points in that box are marked as
>> outdated (or deleted, or whatever). Maybe some extra safety needs to
>> be made by only allowing users active in the specific area to do this,
>> or only users who upload traces.
>
> The problem with this is that it's a broken solution to an already
> limited system. We shouldn't have to /remove/ GPS tracks depending on
> age, but rather have the ability to mark segments or points of them as
> trusted (amongst other things).

Maybe I wasn't very clear, I'm not talking about removing whole
traces, but about marking points inside an area  as invalid.

e.g. Just south-west of Breda in the Netherlands the highway A16 has
been moved a few hundred meters to the west over a distance of a few
kilometers. Whatever traces that were made for that highway are still
valid, except for these few kilometers and the connecting junctions.
We want to keep the traces of the highway, and only 'remove'* the part
that has been moved.

(*) replace with delete/ mark invalid/ mark untrusted as you like.

> ...
> One problem with this is - as has been observed - that the data gets
> less useful for everyone as more traces are uploaded. We can devise
> hacky solutions to this such as not serving old traces via the API.
> But that's just a lame workaround which'll remove a lot of valid
> tracesi E.g. I've surveyed footways that have been there for
> centuries, and probably aren't going anywhere soon.
>
> What if the GPX API worked like this instead:
> ...
> * The data is versioned, and anyone can edit it
>
> I have a lot of GPX tracks that could be improved, e.g. by deleting
> point clouds. I'd like to edit them using normal OSM tools, have those
> edits versioned (so they can be rolled back), and have other users do
> those fixes for me. Just like with the OSM data I upload.

In my opinion traces should be cleaned up before being uploaded,
however I confess that I often don't do that :-)




More information about the talk mailing list