[talk-au] NSW Transport Data Exchange (TDX)

Michael Gratton mike at vee.net
Thu Feb 21 03:00:27 GMT 2013

On 20/02/13 22:24, Andrew Harvey wrote:
> Hmm... seems like they changed from CC-BY back to a custom license?

Yeah, there's some onerous clauses for attribution and re-distribution
and what-not. Pity. I notice you have a copy under the CC-BY licence, I
wonder if that would that be suitable for use for a initial import to OSM?

> On 22/01/13 23:17, Michael Gratton wrote:
>> From the sounds of it, one thing they are anxious is having old
>> route/timetable data being out in public.
> I'm not sure whats wrong with keeping historical records...
> In my view, the less restrictions they place on the data, the more
> freely it will flow, creating more competition and innovation among
> people using the data and building products incorporating the data...


I figure they don't want random members of the public using a service
that publishes PT timetable data for Sydney, thinking the data is
accurate but that is actually out of date. It would be misleading the
user if the service said (say) there's a bus due in 10 minutes when
there's not, because the timetable has changed. The user would possibly
assume that Sydney Buses is at fault (I know I would) rather than blame
the service with the out-of-date data. This would not only make
Transport NSW/Sydney Buses/City Rail/etc look bad, but it would also
inconvenience the public.

> Probably worth another look though, but I believe,
> * It didn't say which roads were used for the route, just station to
> station (but since the stations are fairly close you could infer)
> * I don't think it used the common route names, but rather each
> timetabled journey was recorded as a separate route.

GTFS feeds (optionally) includes route shapes, which can be used to
render the actual routes. Since GMaps currently only does that for rail,
I assume the TpNSW GTFS feed only contains shapes for that.

There is a notion of both routes and trips in GTFS, where a route is
effectively a collection of trips, and where a trip is a sequence of
stops, ordered by arrival and departure times. Both routes and trips
have human readable names and other descriptive information.

> You can still include this into OSM, its just you need to decide is it
> better to do a mass import, or collect it naturally via the traditional
> OSM way of going out and observing what is out there.
> See http://tianjara.net/data/tdx.nsw.gov.au/. Mind you I haven't looked
> any further into it since my initial investigation.

I did have a look at that, cheers. For a previous paid project I have
some Python code useful for representing and generating GTFS, I was
thinking of extending that if TpNSW fixes their licence.

I think it is worthwhile doing a combination import-and-maintain in
conjunction with the traditional approach, as I outlined before:
- basically a mass import combined with a bot to keep it up to date.
This can be combined with the traditional approach, which definitely has
advantages but I don't think scales to something like keeping PT data
complete and up-to-date.


⊨ Michael Gratton, Percept Wrangler.
⚙ <http://mjog.vee.net/>

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 261 bytes
Desc: OpenPGP digital signature
URL: <http://lists.openstreetmap.org/pipermail/talk-au/attachments/20130221/b8cb726b/attachment.pgp>

More information about the Talk-au mailing list