I just sent an email to Yarra Trams asking them if they could build a RESTful API for the real time tram tracking.

It would be great if you get that data like this:
http://yarratrams.com.au/tracker/1847/xml/
Returns an xml document with the next trams from stop 1847.
http://yarratrams.com.au/tracker/1847/86/xml/
Returns only tram 86.

I’ll let you know when they get back to me.

Re: caching. The story around the Yahoo Pipes community is that there’s still a 20 minute caching delay for content. They re-jigged it back in March because people were seeing too much difference between what data was available in edit mode and run mode.

Anyway, it’s moot since the RSS service accommodates the merged feeds. Pipes may still offer some features, like email alerts, conversion to iCal (for calendars) or XML, translation into foreign languages and one or two other things (geocoding? widgets?).

I took a look at the Yarra Trams real-time tram tracker thing, and it’s a bastard to get hold of the data: MS ASP.NET AJAX so wget (or Dapper or Yahoo Pipes) aren’t gonna cut it.

But, anyway, well done on rolling out the new features so quickly!

As the site now shows multiple routes for a stop, I’ve gone ahead and stolen your idea added a new brand new feature.

To get to it through the interface:

Click on the stop your after, in your example it’s stop 15 for route #48 or #75.

Then click on ‘Show all routes for this stop’.

After it’s refreshed an ‘RSS’ link will be available down the bottom.

The link will provide an aggregated RSS listing of the stops for that list, very similar to what you did using the Yahoo Pipe.

The link just looks like this:

http://www.tramroute.com/rss/timetable/651,652

So you can customise the routes that are included by removing/adding to that comma separated list.

So hopefully this should help you get to the data your after without being hinder by the caching issue.

But let me know though if there’s anything I can do to maybe stop Yahoo caching so aggressively. I’m pretty sure the server is sending out the correct header, there appears to be a distinct Etag for each request and apparently they are meant to be ‘honouring’’ these (http://pipes.yahoo.com/pipes/docs?doc=troubleshooting) , though possibly in favour of less hits than more.

The idea is that sometimes you have multiple services or stops (like Ben above) and you don’t care which line their on. This Pipe merges the feeds and sorts by time.

http://pipes.yahoo.com/pipes/pipe.info?_id=_POgoLFM3RGWDwfe8TxBKg

(You can put in two stop numbers and labels.)

Unfortunately, Yahoo’s caching rules may render it a little less useful on subsequent pulls. I tried putting a filter in to drop old times, but not sure it’s effective.

-Greg.

But we will wait and see what the iphone provides in a couple of days once released.

EDIT: Looks like the GPS data is accessible on some models of the BlackBerry (via JavaScript) so maybe it will be made available at some point through the iphone as well.

It’s great to see. Any plans for GPS integration with the new iPhone?

I’ll try and incorporate some of the suggestions over the next couple of days.

To answer some of your questions :

- The position of the trams is estimated based of the timetable. Unfortunately it’s not based on live data from the tram tracker system so it’s just a guide at the moment as to where the trams should be. Hopefully one day we will be able to get access to that system to be able to show the accurate position of the trams.

- The direction is represented in both the text describing the direction as well as the route name. You should notice when you click ‘Reverse Direction’ the title will change from say ‘South Melbourne Beach to East Coburg’ to ‘East Coburg to South Melbourne Beach’.

- The timetable information isn’t taken live, but is updated periodically to reflect any changes in the timetable.

- The labels are there to indicate the start and the end of the route for users who aren’t familiar with the geography of Melbourne. I might look at adding a feature to turn them on/off for those that would like to remove them though or when they are obscuring information.