Today's editorial in the Citizen

[swf...@gmail.com]

There was an editorial in the Citizen today on the GPS data for OCtranspo.

It talks about the service being suspended due to technical details and mentions "an application was being developed".

I'm wondering if it would be useful, or counter-productive to point out that the suspension was more likely due to 'politics' + give a little more info on the multiple applications that were created?

John

[Alex Lougheed]

The article: http://www.ottawacitizen.com/Transpo+waiting+game/4190821/story.html

The author, Glen, is a friend of the group and attends our hackathons. He might even be on this list.

I'd say this coverage is good, and we should definitely contribute an editorial. This kind of media can scare agencies away projects of change, especially if there's "political" concerns at the top.

I missed the meeting yesterday. Can someone elaborate on why the city shut down the feed? What was the "politics"? Who made the decision? Why?

If we're going to respond, it needs to be no later than tomorrow (2 days). The mayor expressed interest in the group in the past, so we can get him engaged enough to grab his ear.

A

Alex Lougheed
m. 613 700 2358

alougheed.ca  twitter.com/alox

[swf...@gmail.com]

The suggestion I heard that it the withdrawal was really a dispute between OC Transpo and the City of Ottawa regarding who actually owns the data.

I was thinking of writing a letter on the following lines:

Thank you very much for raising the issue of live OC Transpo GPS data.    The feed was available briefly in December and early January and allowed several enthusiasts, including myself, to develop apps that enable bus users to find out where the buses actually are.

Whilst there were some technical issues with the data, there was an excellent contact at OC Transpo that was helping deal with teething problems.   Even when the position data is a little unreliable, it at least confirms that a bus is actually running.   Since OC Transpo is currently short of drivers, many buses do not run at all, just knowing which buses are running is a great step forward.

Keeping the feed running would have made it easier to resolve issues with the data.  Apps could easily have provided a caveat for users that the service is still experimental and data may not always be reliable.

In fact, my understanding is that the real reason for the withdrawal of the service was a political dispute regarding ownership of the data between the City of Ottawa and OC Transpo. It is a great shame that petty disputes are standing in the way of better information for bus users.

John

[Jonathan Rudenberg]

I can confirm that I've heard this as well.

Jonathan

--
Jonathan Rudenberg | @titanous

On Tue, 01 Feb 2011 16:06 -0500, "swf...@gmail.com" <swf...@gmail.com>
wrote:

[Alex Lougheed]

So in other words, the city wants to open it up in accordance with the OD policy, and OC Transpo doesn't. The only reason OC Transpo be fighting for control is if they don't like what the city is doing.

Your piece is good John, but it gets a little to technical. The Citizen isn't interest in publishing the details of the system--they're looking for a story that interests everyone from  little Johnny next door, to Grandpa Johnny in Kanata.

Here's my shot at an editorial:

OC Transpo Needs to Take a Lesson from the City

re: Glen McGregor's January 31st Piece.

Glen McGregor's piece on how buses chronically miss their schedule times, and how internal accountability measurements are skewed to paint their world rosy is a situation we encounter far too often from public agencies afraid of the public.

The City of Ottawa recently jumped over that hurdle. It passed a policy embracing the principles of Open Data: the proactive disclosure of everything from voting districts to lawn bowling sites--all in formats that citizen developers can make useful tools for Ottawans out of.

The city recently opened up a trial project to the Open Data community: the live GPS bus data. The community was in the process of developing applications for real time schedules, so transit riders can get to their bus stop as the bus arrives--instead of waiting in the winter chill for, as Glen points out, way longer than they should be.

Unfortunately, due to some internal disputes, OC Transpo put an end to this. They turned off the data taps over a needless dispute about who owns the data: the City of Ottawa, who wants to set it free, or OC Transpo, who apparently does not. This dispute misses the point. If we want transit in this city to be the best it can be, the data ought to be in the hands of those who pay for and use it: the everyday citizen.

If we want Ottawa to be more livable and innovative in the 21st century, Mayor O'Brien ought to step in and stop these petty squabbles. We can help OC Transpo make the system we all rely on so much better. And we're going to put in the labour for free. Use us.

Thoughts?

Alex Lougheed
m. 613 700 2358

alougheed.ca  twitter.com/alox

[dave.sc...@gmail.com]

Good job so far. I have two thoughts.

1. Jim Watson is the current mayor.

2. I don't think the issue is that OCTranspo doesn't want the data to be free...I think they do, but with a complicating factor. In August, they budgeted for a private company to produce a web app, mobile apps, and an app for down-stream display screens (like the ones at most hubs) to consume the gps data. Opening the data this early allows us to undercut that effort for $0. In the meantime, they've committed to spending several million on apps from the private company (the one that provided the Next Stop Announcement system). Also note that this plan involves incorporating gps data from the Next Stop Announcement system (currently, it appears that the gps data comes from the gps devices mounted to the fare boxes), and adding some predictive heuristics to guess at bus locations based on historical data (when gps locks are unattainable). Source: the budget doc that Jonathan mentioned (http://ottawa.ca/city_hall/budget/budget_2011/transit_budget_2011_en.pdf).

Dave

[Alex Lougheed]

1. I'm stuck in the past.

2. Thanks for all that info. Very informative. I still think it would be foolish to not push ahead though. I think that council would embrace us provided some fire to the contracted folks feet--to see if they can make a better app than us. If they can't, the city learns a valuable lesson in how it gives out contracts. If they can, the city still learns a valuable lesson in how it gives out contracts. We can also spin it so it becomes a win for Watson: fixing the old plan which was put in place under O'Brien.

With no substantial objections, I'm going to submit that letter at 9pm tonight.

A

[swf...@gmail.com]

Agree some sort of letter should be sent -- I like Alex's version, but  the info about the contract with the company got the next stop announcement system is very interesting.    

It may be the company itself got some exclusive deal on providing apps, so they would not have been happy about the data being freed.   Or, of course it might be that OCTranspo does not want the embarrassment of enthusiasts providing apps for free that do just what they have paid millions for.

So long as there is no contractual stuff that prevents free-ing of the data then I agree it is a good opportunity for OC Transpo/the city to learn a lot about the merits of open data.

John

[Sean Kibbee]

I totally agree. To add to the letter, the question should be asked - how much are we spending to have a company create app(s) when a team of Ottawa resident developers are willing to do it for free and are already halfway done?

I also recommend we all contact our councillors. I doubt they know the real story on this one.

Sean

"Alex Lougheed" <alou...@gmail.com> wrote:

>1. I'm stuck in the past.
>
>2. Thanks for all that info. Very informative. I still think it would
>be
>foolish to not push ahead though. I think that council would embrace us
>provided some fire to the contracted folks feet--to see if they can
>make a
>better app than us. If they can't, the city learns a valuable lesson in
>how
>it gives out contracts. If they can, the city still learns a valuable
>lesson
>in how it gives out contracts. We can also spin it so it becomes a win
>for
>Watson: fixing the old plan which was put in place under O'Brien.
>
>With no substantial objections, I'm going to submit that letter at 9pm
>tonight.
>
>A

>*
>*

[Dave Schindler]

Oh, I forgot to mention that the budget draft also mentions another reason why OCTranspo might want to own the apps (or at least the app that runs on the video display units...it's unclear). Look at the 2nd line of this quote from the draft: "Customer interfaces for web, IVR, mobile devices and platform arrival displays will be designed and integrated with existing interfaces. Video advertising stream will be developed with commercial partner to offset maintenance and generate additional revenue."

Dave

[Don Kelly]

On Tue, Feb 1, 2011 at 18:06, Sean Kibbee <se...@kibbtech.ca> wrote:
> I totally agree. To add to the letter, the question should be asked - how much are we spending to have a company create app(s) when a team of Ottawa resident developers are willing to do it for free and are already halfway done?
>

I tend to agree with the sentiment of this statement, but I'd be
careful making it. The key to employing someone to build software for
you is that money (and a contract) buys time, focus and expertise.
There's no guarantee that we'll continue to develop our different
projects once yet-another interesting project comes along. There's
also no guarantee that we'll add the features that octranspo think
they want - we'll add those that are interesting to us.

That said, I work for a company that develops applications for hire.
If it were my employer working on the project for octranspo, my first
response to ~free~ applications would be that we bring that "extra
something special" that might not come out of people working for free
(yes, I realise that this could be perceived as FUD). Furthermore, if
the "free" applications were actually free software, I'd probably
point them out and attempt to build on top of them (giving credit
where due, of course).

Just something to consider,

Don/

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

[Sean]

Good points Don. I would agree that you are far more likely to get a better
product if you pay cash to get it done. However, in a time of rising taxes
and dependent on the costs involved in procuring and paying a company to
perform the service would it serve in the city's best interest to have it
done this way? Add to that the feel good stories around open data and app
contests that engage the community for cost effective solutions and I think
it's a no brainer. If the city doesn't control the product they also incur
much less blame when something does go wrong (and it will, to one degree or
another, with either scenario). After all if it truly was free to taxpayers
than there's much more latitude when it comes to imperfections.

Do we as citizens of this city want a product that's 100% complete for 100%
of the cost or 90% complete for 5% of the cost? As the purse strings tighten
to cut back on taxes (or prevent increases) it really boils down to
effective fiscal management and I'm confident in the spirit of the new
Council to operate in a manner that reflects a commitment to learn from past
mistakes. After all, there's already a system in place which provides
adequate bus schedule views. This open data is all a bonus and not a
critical service so why are we investing more money than we have to into it?

It's likely we've already paid for the development services which, in my
opinion, is unfortunate. Open data is about empowering the people and self
correction at little cost. I'm always amazed at how effective open data can
be as an enabler when you entrust the people to serve the people where data
is concerned. Touting open data than trying to control the medium is just
bad business.

Regardless though, I think we all agree that the feed should be opened back
up. I can't imagine a contract that would restrict the city from doing just
that in parallel with any previous agreed upon obligations. At the very
least it might save some maintenance costs and provide an alternative down
the road. If it's a matter of embarrassment; well, two wrongs don't make a
right and it's time to make it right.

There's likely much I don't know about the situation but I do like the
proposed letter to respond to the article and I hope someone will send it
off. I Also encourage everyone to talk to your Councilor and ask the right
questions. How much are we paying for this when it could be done for next to
free? This city continues to get duped by bad advice and even worse contract
restrictions and it is truly disheartening to watch it happen time and time
again.

Sean Kibbee

[Michael Richardson]

>>>>> "Don" == Don Kelly <kar...@gmail.com> writes:
Don> I tend to agree with the sentiment of this statement, but I'd be
Don> careful making it. The key to employing someone to build software for
Don> you is that money (and a contract) buys time, focus and
Don> expertise.

True.

But, owning the app also lets you control the API and control how much
information is available, and to whom.

Someone there realized that an automated system could simply pull the
schedule, the time points, and the GPS data, and then simply file a
complaint everytime a bus leaves a time point early.

Don, as you know, that a lot of buses are not on time, and as the
Citizen determined from the raw data, staff has been lying (with
statistics) to council for some years about the on-time performance of
OC-Transpo.

--
] He who is tired of Weird Al is tired of life! | firewalls [
] Michael Richardson, Sandelman Software Works, Ottawa, ON |net architect[
] m...@sandelman.ottawa.on.ca http://www.sandelman.ottawa.on.ca/ |device driver[
Kyoto Plus: watch the video <http://www.youtube.com/watch?v=kzx1ycLXQSE>
then sign the petition.

[Don Kelly]

On Tue, Feb 1, 2011 at 22:28, Michael Richardson <m...@sandelman.ca> wrote:
>
> But, owning the app also lets you control the API and control how much
> information is available, and to whom.
>

Definitely true.

> Someone there realized that an automated system could simply pull the
> schedule, the time points, and the GPS data, and then simply file a
> complaint everytime a bus leaves a time point early.
>

Agreed. I'm sure that the political implications of the data's
availability played a key role in the disappearance of the feed.
Honestly, it's these issues that always had me thinking that we'd
never see such a thing. Obviously, the City wants it available but
octranspo does not.

> Don, as you know, that a lot of buses are not on time, and as the
> Citizen determined from the raw data, staff has been lying (with
> statistics) to council for some years about the on-time performance of
> OC-Transpo.
>

Hm... Lying, strong words. Might even be true.

I agree 100% with Michael and Sean. My reason for pointing out the
mentality behind employing an application developer was to add more
nuance to the conversation. Basically: how the PHBs might see us
versus them when in fact there's no difference.

My own opinion is that this data should be free: available and
unencumbered. Whether there is an external contractor should be
irrelevant to whether the data is available.

[Alex Lougheed]

Howdy all,

Given the considerable internal debate as to the merits of hiring private development companies, and due to the consensus that we should still have our danged feed, I left out the contract procurement piece from the letter to the editor. It was submitted basically as drafted above, excepting a couple of language changes, and updating the mayor's name. Below is what I sent out.

We should also draft a letter to the mayor, and cc the city councillor which shows up to our hackathons. I can't recall her name. Anyone?

A

---------- Forwarded message ----------
From: Alex Lougheed <alou...@gmail.com>
Date: Wed, Feb 2, 2011 at 10:42 AM
Subject: Letter to the Editor re: Glen McGregor's January 31st Piece
To: let...@thecitizen.canwest.com
Cc: Edward Ocampo-Gooding <edwa...@gmail.com>


To the editor:

Please consider the following letter, submitted on behalf of Ottawa's Open Data community. The Open Data community is a grass roots, loosely organized collective of software developers, designers, librarians, data wonks and concerned everyday citizens. Together we assist the city implement its Open Data policy, we organize development hackathons and we push for more accessable government/agency information.

Should you have further questions, please don't hesitate to contact me at the information in my signiture.

Yours,
Alex

--

OC Transpo Needs to Take a Lesson from the City

re: Glen McGregor's January 31st Piece.

Glen McGregor's piece on how buses chronically miss their schedule times, and how internal accountability measurements are skewed to paint the city's world rosy is a situation we encounter far too often from public agencies afraid of the public.

The City of Ottawa recently jumped over that hurdle. It passed a policy embracing the principles of Open Data: the proactive disclosure of everything from voting districts to lawn bowling sites--all in formats that citizen developers can make useful tools for Ottawans out of.

The city recently opened up a trial project to the Open Data community: live GPS bus data. The community was in the process of developing applications for real time schedules, so transit riders can get to their bus stop as the bus arrives--instead of waiting in the winter chill for, as Glen points out, way longer than they should be.

Unfortunately, due to some internal bickering, OC Transpo put an end to this. They turned off the data taps over a needless dispute about who owns the data: the City of Ottawa, who wants to set it set free, or OC Transpo, who apparently does not. This dispute misses the point. If we want transit in this city to be the best it can be, the data needs to be in the hands of those who pay for and use it: the everyday citizen.

If we want Ottawa to be more livable and innovative in the 21st century, Mayor Watson ought to step in and stop these petty squabbles. We can help OC Transpo make the system we all rely on so much better. And we're going to put in the labour for free.

- 30 -

Alex Lougheed
m. 613 700 2358

http://opendataottawa.ca/  twitter.com/opendataottawa

My personal address:

139 Patterson Ave.
Ottawa, ON  K1S 1Y4

[Dave Gallant]

A few general points to add:

1.  I was at the (recent) Sunday meeting, and yes, we heard the reason the feed was shutdown was due to internal politics, and we are not privy to any further details.  Current technical issues with the feed are well within the capabilities of the city's IT staff/contractors to solve.

2.  A great deal of what is going on boils down to accountability and transparency.  I believe it likely that some managers/individuals inside OC Transpo would much prefer to pay whatever sum of money is required to a private software contracting company to produce the mobile/web live GPS apps and keep the gps API hidden.  By doing this, they will be far less accountable and transparent to the public, as no one outside will be able to watch/analyze the data feed.  If, for example, buses still continue to be consistently early on some routes and this information is not displayed on the "official" live gps app, OC Transpo will be able to claim "internal technical issues" and the public will be left with little recourse.

3.  It is the engagement of the wider community with the open data set (the live GPS data) that will help move OC Transpo and the city forward into the 21st century.  I have no qualms with a private company being contracted to build out the official live gps application set, but the citizens of Ottawa should have oversight over the quality of the final result, which would happen if the gps feed data was open to all.

4.  I think it is important for the open data development community to demand a basic level of communication standards from the City of Ottawa, in exchange for what is essentially volunteer time on these projects.  I agree with the concept "we're here, so make use of us", but at the same time I don't intend to give away much more of my time if the city can't find a way to communicate effectively with the open data community.  i.e.  having the live gps data feed shutdown without any kind of notice to the community.

5.  It is probably going to take a large PR campaign at this point to get OC Transpo to release the live feed again.  The phones need to be ringing off the hooks at the offices of city councilors.



Dave G.

On Wed, Feb 2, 2011 at 9:52 AM, Don Kelly <kar...@gmail.com> wrote:

--
Move and the Way will open.

[swf...@gmail.com]

Great points.

Re: point 2.  The recent Citizen piece probably was not helpful in this respect -- using the historic GPS data to show the problems. 

In particular, I was curious about the >40% of buses being early.   The GPS data often showed buses early, but the real issue is passing stops early (say more than 1 minute early), the GPS data may not reveal that well.

It might be worth someone with access to the historic data to do a better analysis of this -- if OC Transpo saw open data making them look better then that could help.

John

[Darren Kipp]

On of the issues with the on time analysis of data that I noticed while testing the live feed is that sometimes buses get re-ordered or whatever reason.

I recall seeing an instance of buses on a 15 minute frequency route being almost exactly

15 minutes early
15 minutes early
30 minutes late

Consequently on a record by record basis these buses are all way off schedule.  If you look across records (which is difficult with 93m rows) the buses are in fact almost exactly on schedule. 

To the person standing at the curb there is near perfect schedule compliance, yet in terms of the data, things are in terrible shape.

[Michael Richardson]

>>>>> "swfiua" == swfiua <swf...@gmail.com> writes:
swfiua> Re: point 2. The recent Citizen piece probably was not helpful in this
swfiua> respect -- using the historic GPS data to show the problems.

swfiua> In particular, I was curious about the >40% of buses being early. The GPS
swfiua> data often showed buses early, but the real issue is passing stops early
swfiua> (say more than 1 minute early), the GPS data may not reveal
swfiua> that well.

If the stop is a timepoint, then the bus has to stop there and wait.
The fact that few buses ever do this is the problem.

If a bus stops at a time point and waits there:
a) the GPS positioning data is going to be more accurate, because
you can accumulate more readings when you are not moving.

b) there is a much higher likelyhood that the data will get reported
on time. Because the bus is not moving, it is not moving from cell
to cell, and it should get a chance to report.

So there should be no technical issues with knowing if a bus leaves a
time point early.

--

[swf...@gmail.com]

I agree, but there are lots of bad ways to analyse the data.

Eg if you just look at all positions reported and count how often the bus is reported early you will get misleading figures, if the buses ever wait at stops in order to stick to the timetable.

Since the Citizen don't tell us exactly what they did, then there is at least this possiblity.

John