Answers in council to questions

[Don Kelly]

(http://www.ottawa.ca/calendar/ottawa/citycouncil/tc/2011/02-16/TTC%20Inq%2003-11%20-%20Open%20Data%20City%20(Wilkinson).htm)

In case anyone missed it, that appears to be the "official" answer to
council on the GPS api.

Don/

--
karfai [AT] gmail.com
http://www.strangeware.ca
http://blog.strangeware.ca

[John Bain]

Reading that annoys me!

To add insult to injury I was at the Queensway station waiting for my bus and i noticed that the OC transpo mobi site and the schedule board at the stop showed 2 different times for the same bus that ended up being late anyways ... 

[Don Kelly]

One part which caught my eye, on re-reading, was:

"As well, there are several issues related to the open data model,
such as enabling third parties to deliver OC Transpo real-time bus
arrival information, which if it were to continue, could create false
expectations and negative results for our customers. These third
parties are not delivering real-time data but a forecast based on the
data they have, which may or may not be consistent with what is
happening in actual operations. Control of real-time data by the
transit authority is consistent with what other transit authorities
throughout North America are doing, such as the Chicago Transit
Authority."

What's meant by "forecast"? Was the feed that the node.js site was
consuming only a forecast? Or, is he trying to spin the conversation
to suggest that the "real time" apps are only working predictively
from the static schedule? AFAICT, the feed we were consuming was not a
forecast but the actual data out of the stop announcement system (or a
sibling of that system).

Anyone know the answer?

[Jonathan Rudenberg]

The feed was from the old "Bell system". They are in the middle of
deploying a new system from Clever Devices, and replacing a bunch of
custom internal systems that they currently use. Once the new Clever
Devices system is fully deployed, they'll release realtime data to
customers via a COTS app.

I talked to Mr. Mercier about this a while back. This is what he said
about the data we had access to:

"The current live system is used by operations to help lift the fog off
the streets, but there is still fog."

It does not reflect detours, mid-trip cancellations, and a variety of
other things. He was adamant that if the data is realtime it must be
directly from OC Transpo (not modified by developers), and that it
always be completely accurate.

Jonathan

--
Jonathan Rudenberg | @titanous

[Don Kelly]

On Mon, Mar 21, 2011 at 10:53, Jonathan Rudenberg <jona...@titanous.com> wrote:
>
> It does not reflect detours, mid-trip cancellations, and a variety of
> other things. He was adamant that if the data is realtime it must be
> directly from OC Transpo (not modified by developers), and that it
> always be completely accurate.
>

This is a /much better/ explanation than the one in the link.
"Forecast" carries a lot more meaning than "does not deal with
events".

[swf...@gmail.com]

This para from OC Transpo was a little mis-leading.   It suggests that the third party apps were taking the data provided by OC Transpo and then making a forecast.   The reality was that the OC Transpo data itself was a forecast.   It had the position of the bus (but not the time that the bus was at that position) and a forecast for when it would get to the stop you are interested in [aside: given this data and timetable data we could have deduced the time of the GPS bus location -- would have been a fun thing to try -- of course better to just have that provided ;) ]

Anyway, as I understand it, things have moved on since the answers to the council and once OC Transpo have their new, improved data available I thought they had agreed to create a feed that third party apps can use.  

Reading Jonathon's response I'm not so sure now -- that sounds like they are just going to create a COTS app.

John