[talk-au] NSW Transport Data Exchange (TDX)
inas66+osm at gmail.com
Tue Jan 22 22:48:42 GMT 2013
On 23 January 2013 09:31, Michael Gratton <mike at vee.net> wrote:
> There's no reason why using feeds such as the TDX is incompatible with
> individual user editing. From first-hand experience, even the PT
> agencies get their data wrong at times, and so the
> legion-with-smartphones out there is still actually quite important.
If the data is going to be user-corrected that's great, and OSM is the
right tool. But the TDX data is updated very regularly, so if I fix a bus
stop today, how do you know whether you should overwrite that with the
next revision of TDX data tomorrow?
A simple question, but I can't think of a simple solution.
> The usefulness of these feeds is two-fold, the first is obviously for
> populating a large amount mostly-correct data otherwise missing from
> OSM, but equally important is getting updates down the track. Bad data
> is after all just as bad as missing data (worse?), as my local
> government GIS friend constantly complains.
Yes, but your government GIS friend probably is quite used to incorporating
data from different sources, into different layers.
OSM isn't yet built for that. Large scale data imports have issues.
Issues with licencing, issues with updating, issues with verifying, issues
with understanding exactly what they mean. The easy part is the importing.
> The challenge then is how to merge feed updates along with
> individual-generated changes. It's not impossible, it just requires the
> right data, some conventions, some code and a bot.
That certainly defines the nature of the challenge. I'd be happy to see
the spec of exactly how you think we can stir this into our data soup. If
you intend retaining user created data while discarding the imported data,
you're likely creating the exact issue that the data issuers wish to avoid.
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Talk-au