Toronto Sun: Toronto’s data open but almost useless

[James McKinney]

Replies to David and Michael:

On Thu, Jul 7, 2011 at 4:36 PM, David Eaves <[hidden email]> wrote:
> James, there are always going to be asymmetries in resources and access.
>
> My fear is on this is that you are creating a new condition - data can only
> be released if it eliminates these asymmetries.

That is not my point, nor is it the Sun's, Michael Gurstein's, or
others that I've read. Accessibility is something that comes after
availability in the discussions of open data I refer to. All I'm
saying is "Let's not stop at availability," which I think should be
fairly uncontroversial. I'm not introducing a new condition. I'm
encouraging people to adopt, or at least keep in mind, another,
further goal. The reason people are making noise about this is that
(1) few people are talking about accessibility in the open data crowd,
(2) keeping the ultimate goals in mind, such as accessibility, can
help direct which datasets we want to open, how, for whom, etc., (3)
there is a fear that if we get too caught up in pushing for open data
period, we may forget why we're pushing for it in the first place.
Open data is just a means after all; it doesn't effect change on its
own. I think the only point of all this is to say "don't confuse open
data for an end in itself."

David, could you provide links to the writings of people who use
accessibility concerns to argue that open data is bad? I haven't
encountered that argument myself; it doesn't sound very persuasive
from what I've heard about it.

On Thu, Jul 7, 2011 at 6:08 PM, Michael Mulley
<[hidden email]> wrote:
> I don't think public data, in your usage, is a better conception of
> open data; I think it's a different beast.

I am not redefining open data or trying to replace it. I'm just
describing a kind of data that many refer to as a further
advancement/enhancement that builds on open data. But, like you say,
accessibility and availability are often apples and oranges. And there
are cases where, if I would have to choose, I would prefer one over
the other. Data that saves lives, e.g. health advisories?
Accessibility, preferably. Financial data, e.g. municipal budgets? I
lean towards availability.

David, I can certainly provide better definitions of accessibility.
It's an established multi-disciplinary research field, and it can
handle important questions like "Accessible to whom?", etc. But I
don't think that will get us further. Also, I've no doubt there will
always be asymmetries. That doesn't excuse the asymmetries, however,
and it doesn't mean we shouldn't endeavour to reduce them where it is
important to do so.



To introduce another reason why we shouldn't only concern ourselves
with openness but think also about what comes next: what I fear is
that with increased transparency, the government will get away with
shifting some of its burden to the citizen. Take this example from the
following NY Times piece:

"Or take conflicts of interest in medicine. Despite volumes of
research showing that pharmaceutical industry gifts distort decisions
by doctors, the medical establishment has not mustered the will to bar
such thinly disguised bribes, and the health care reform act fails to
outlaw them. Instead, much like food labeling, the act includes
'sunshine' provisions that will simply make information about these
gifts available to the public. We have shifted the burden from
industry, which has the power to change the way it does business, to
the relatively uninformed and powerless consumer."
http://www.nytimes.com/2010/07/15/opinion/15loewenstein.html?pagewanted=print

The solution is not to stop all transparency efforts; this is not an
argument against open data. Rather, it is an argument to be vigilant
and to work to ensure that the government remain responsible. Just
like many on this list fear government closing up, many others fear
government leaving certain marginalized and underprivileged people
behind. Availability and accessibility may be apples and oranges, but
they still interact, and it is this interaction that is the source of
this discussion.

James

On Thu, Jul 7, 2011 at 6:08 PM, Michael Mulley
<[hidden email]> wrote:

> I don't think public data, in your usage, is a better conception of
> open data; I think it's a different beast. And as loath as I am to
> ride into semantic battle, I think "open data" as a concept is
> confused often enough that we should be careful about the term.
> Already "open data" is a scary term in Ottawa, because "open" has been
> such a politicized term lately: semantics are important.
>
> Accessibility/"public data" means, essentially, the creation of
> applications. (The Sun article seems to think, kinda fuzzily, that
> it's a matter of file formats or some such thing; like Tracey, I don't
> think that's correct.) And, as in your homeless shelter map example,
> perhaps these applications aren't digital. Okay: I certainly agree
> that helping people access information relevant to their lives is a
> social good. I've written angry letters about reductions in library
> hours. But that wasn't open data activism. And I worry that saying
> "open data is good, but public data is better" is like saying
> "budgetary data is good, but financial literacy workshops are better":
> apples, oranges.
>
> Yes, the ultimate goal of open data policies is to help people access
> relevant information. Yes, just tossing data online -- particularly in
> Canada, where there's no significant institutional/nonprofit presence
> around open data -- will largely result in uses that are either
> commercially lucrative or at least somewhat self-interested, i.e.
> websites that appeal to the Twitter set. And no, publishing a dataset
> does not magically absolve government of all further social
> responsibility.
>
> But if we allow "open data" to balloon into "help people with
> knowledge!", we'll never get anywhere. Open data still has lots of
> benefits: for the potential it provides, it comes at an exceedingly
> low cost once cultural barriers are overcome. An ideal open data
> policy would include outreach and funding so that data gets used by
> outsiders in social useful ways. But if we increase the scope of our
> common argument beyond that, into the general provision of
> information, we confuse people about what open data is. And given that
> the largest challenge facing open data is probably marketing and
> cultural change, I think being clear about what we mean is pretty
> crucial.
>
>
> On Thu, Jul 7, 2011 at 12:54 PM, James McKinney <[hidden email]> wrote:
>> Nik, the argument is not fallacious. You are just disagreeing on
>> terms. I agree with your definition of open data. What the Sun is
>> asking for is public data, not open data. In brief, public data is
>> open data that is citizen-ready. That means ready for all citizens,
>> not just those with tech savvy. "Pure" open data is only
>> technically-savvy-citizen-ready. You write that it is not the
>> government's responsibility to provide public data, only open data.
>> The Sun's argument is that the government _ought_ to provide public
>> data, not just open data; that it ought to serve all citizens, not
>> just the tech-savvy. Much of this data is relevant to everyone,
>> therefore everyone ought to be able to use it (so the argument runs).
>>
>> You write that there's nothing wrong with relying on developers to
>> build open-data-based apps for the public, and back it up by saying
>> that the public already relies on developers to build all sorts of
>> apps. Let's have a look at these other apps. Do they target, engage
>> and empower the marginalized, the underprivileged, the poor? No; it's
>> not a hot, profitable, or easy market. But what about government?
>> Should it also ignore these people? Of course not. That's why it's not
>> enough to rely on developers. Developers do not have hearts of gold.
>> This is the argument the Sun is making.
>>
>> I find nothing objectionable with the Sun article. All it's saying is,
>> "open data is good, but public data is better; don't forget public
>> data in your open data push."
>>
>> On Thu, Jul 7, 2011 at 11:03 AM, Nik G <[hidden email]> wrote:
>>> The fallacy of this article's argument is that data isn't "truly open" until
>>> the Government builds or implements interfaces, applications & tools to make
>>> the data more accessible to non-technical audiences. The whole point of open
>>> data is to enable Government as a Platform for others to build upon, where
>>> the Government does the least possible to get the data out into the open, so
>>> that _others_, not Government, can then build visualizations, applications,
>>> mashups, etc.
>>>
>>> The key to getting more citizen-ready applications & visualizations is
>>> fostering community development, working in tandem with the local open data
>>> & open gov advocates to scale through others. There's nothing wrong with
>>> citizens relying on web-savvy developers to build apps, that's how it's done
>>> for thousands of "non-opendata" apps that we use every day on the web,
>>> mobile, desktop. The point is that our governments need to build
>>> competencies in harnessing the skills & knowledge of communities to take
>>> that open data from raw form to a citizen-ready app.
>>>
>>> City of Toronto has a great strategy in pushing to make open data as part of
>>> each department's workflow; it's visionary in that this is a cultural shift
>>> and not a point-in-time activity. But as with any initiative that's pushing
>>> the cultural, process & technology boundaries, there are cost/benefit and
>>> immediacy/relevancy trade-offs. In case of open data, as long as the data is
>>> as close to the source, original dataset, there's nothing wrong in getting
>>> it out in a machine-readable format under an open license sooner rather than
>>> later. By the way, Toronto's catalogue file formats aren't that drastically
>>> different from any other catalogue, providing your usual CSV, XML, XLS
>>> formats that are "spreadsheet-ready". If it's the geographic formats that
>>> _seems_ to be "user unfriendly" (ESRI Shapefile vs. the commonly-used KML),
>>> there are also relatively simple ways to convert those GIS formats without
>>> having to sacrifice staying close to the original data formats used by the
>>> city.
>>>
>>> Open, Rinse, Repeat is the recipe for success, not Wait Till it's Perfect &
>>> Pretty, but Outdated.
>>>
>>> I'll take open data SOONER rather than later, ANY format rather than no
>>> format, and NOT having to wait months or years till it's "perfect", but
>>> meets everyone's requirements according to everyone's technical comfort
>>> level.
>>>
>>> Nik Garkusha @Nik_G
>>> http://openhalton.ca
>>> http://port25.ca
>>>
>>> From: Tracey P. Lauriault
>>> Sent: Thursday, July 07, 2011 9:22 AM
>>> To: civicaccess discuss
>>> Subject: [CivicAccess-discuss] Toronto Sun: Toronto’s data open but almost
>>> useless
>>>
>>> Toronto’s data open but almost useless
>>> http://www.torontosun.com/2011/07/06/torontos-data-open-but-almost-useless
>>>
>>> This is the first news article I have seen in Canada to date that questions
>>> data access for citizens not just designer/developer/academic community:
>>>
>>>> “Right now, the primary audience is the local designer/developer/academic
>>>> community,” Garner says. “However, we have plans to build it out and
>>>> position it more clearly within the context of Open Government in a way
>>>> which would have more tangible benefits for a wider audience.”
>>>> It’s time for those plans to become action. As long as citizens have to
>>>> rely on web-savvy developers to do the hard work for them, the data isn’t
>>>> truly open.
>>>
>>>
>>> However, I am not sure which format the article's author wants the data in,
>>> xls is probably the lowest common denominator and many of the City's data
>>> are in those formats.  The City of Toronto Catalog has ESRI shape files,
>>> those are GIS files and there is no making those easier either.  How does a
>>> city decide on formats?  Should they be releasing data in the way that they
>>> use them in the formats used as part of a city's business processes or
>>> should they re-format the data for the public?
>>>
>>> Geogratis took the approach to release data as they use them.  This keeps
>>> their costs down by not adding any work load, it also keeps the data
>>> accurate, as conversion can introduce errors in geomatics files.  It also
>>> means that users need to know how to work with those formats and do the
>>> conversions themselves and bear the risk of errors as well.
>>>
>>> The article is also a bit erroneous, as the Edmonton Catalog has many
>>> formats that are difficult to use by lay people.  Useful for those creating
>>> mashups but not great for lay people.
>>>
>>> So interesting that it brings up citizen use but not quite an accurate
>>> picture.
>>> --
>>> Tracey P. Lauriault
>>> 613-234-2805
>>> http://traceyplauriault.ca/
>>>
>>>
>>> ________________________________
>>> _______________________________________________
>>> CivicAccess-discuss mailing list
>>> [hidden email]
>>> http://lists.pwd.ca/mailman/listinfo/civicaccess-discuss
>>>
>>> _______________________________________________
>>> CivicAccess-discuss mailing list
>>> [hidden email]
>>> http://lists.pwd.ca/mailman/listinfo/civicaccess-discuss
>>>
>> _______________________________________________
>> CivicAccess-discuss mailing list
>> [hidden email]
>> http://lists.pwd.ca/mailman/listinfo/civicaccess-discuss
>>
> _______________________________________________
> CivicAccess-discuss mailing list
> [hidden email]
> http://lists.pwd.ca/mailman/listinfo/civicaccess-discuss

[Nik G]

Re: Accessibility argument - I can assure you it comes up in the context of 'lets not rush with the data till it's "ready"', not necessarily that open data is 'bad'.

Either way, this leads to initiatives that stall. It came up as a concern as recently as last night's Burlington open data focus group.

Many in the muni space don't recognize the Progression from availability to accessibility; it's often an either or position -- as evidenced in the article. It argues that Toronto took a different, problematic path vs other cities, and now needs to correct the information availability issue... I disagree, and think its about maturity of the initiative and as  James points out it's about taking that next step and meeting further goals after the 'base' has been established.


-----Original Message-----
From: James McKinney
Sent: Thursday, July 07, 2011 6:44 PM
To: civicaccess discuss
Subject: Re: [CivicAccess-discuss]      Toronto Sun: Toronto’s data open but almost useless

Replies to David and Michael:

On Thu, Jul 7, 2011 at 4:36 PM, David Eaves <[hidden email]> wrote:
> James, there are always going to be asymmetries in resources and access.
>
> My fear is on this is that you are creating a new condition - data can only
> be released if it eliminates these asymmetries.

That is not my point, nor is it the Sun's, Michael Gurstein's, or
others that I've read. Accessibility is something that comes after
availability in the discussions of open data I refer to. All I'm
saying is "Let's not stop at availability," which I think should be
fairly uncontroversial. I'm not introducing a new condition. I'm
encouraging people to adopt, or at least keep in mind, another,
further goal. The reason people are making noise about this is that
(1) few people are talking about accessibility in the open data crowd,
(2) keeping the ultimate goals in mind, such as accessibility, can
help direct which datasets we want to open, how, for whom, etc., (3)
there is a fear that if we get too caught up in pushing for open data
period, we may forget why we're pushing for it in the first place.
Open data is just a means after all; it doesn't effect change on its
own. I think the only point of all this is to say "don't confuse open
data for an end in itself."

David, could you provide links to the writings of people who use
accessibility concerns to argue that open data is bad? I haven't
encountered that argument myself; it doesn't sound very persuasive
from what I've heard about it.

On Thu, Jul 7, 2011 at 6:08 PM, Michael Mulley
<[hidden email]> wrote:
> I don't think public data, in your usage, is a better conception of
> open data; I think it's a different beast.

I am not redefining open data or trying to replace it. I'm just
describing a kind of data that many refer to as a further
advancement/enhancement that builds on open data. But, like you say,
accessibility and availability are often apples and oranges. And there
are cases where, if I would have to choose, I would prefer one over
the other. Data that saves lives, e.g. health advisories?
Accessibility, preferably. Financial data, e.g. municipal budgets? I
lean towards availability.

David, I can certainly provide better definitions of accessibility.
It's an established multi-disciplinary research field, and it can
handle important questions like "Accessible to whom?", etc. But I
don't think that will get us further. Also, I've no doubt there will
always be asymmetries. That doesn't excuse the asymmetries, however,
and it doesn't mean we shouldn't endeavour to reduce them where it is
important to do so.



To introduce another reason why we shouldn't only concern ourselves
with openness but think also about what comes next: what I fear is
that with increased transparency, the government will get away with
shifting some of its burden to the citizen. Take this example from the
following NY Times piece:

"Or take conflicts of interest in medicine. Despite volumes of
research showing that pharmaceutical industry gifts distort decisions
by doctors, the medical establishment has not mustered the will to bar
such thinly disguised bribes, and the health care reform act fails to
outlaw them. Instead, much like food labeling, the act includes
'sunshine' provisions that will simply make information about these
gifts available to the public. We have shifted the burden from
industry, which has the power to change the way it does business, to
the relatively uninformed and powerless consumer."
http://www.nytimes.com/2010/07/15/opinion/15loewenstein.html?pagewanted=print

The solution is not to stop all transparency efforts; this is not an
argument against open data. Rather, it is an argument to be vigilant
and to work to ensure that the government remain responsible. Just
like many on this list fear government closing up, many others fear
government leaving certain marginalized and underprivileged people
behind. Availability and accessibility may be apples and oranges, but
they still interact, and it is this interaction that is the source of
this discussion.

James

On Thu, Jul 7, 2011 at 6:08 PM, Michael Mulley
<[hidden email]> wrote:

> I don't think public data, in your usage, is a better conception of
> open data; I think it's a different beast. And as loath as I am to
> ride into semantic battle, I think "open data" as a concept is
> confused often enough that we should be careful about the term.
> Already "open data" is a scary term in Ottawa, because "open" has been
> such a politicized term lately: semantics are important.
>
> Accessibility/"public data" means, essentially, the creation of
> applications. (The Sun article seems to think, kinda fuzzily, that
> it's a matter of file formats or some such thing; like Tracey, I don't
> think that's correct.) And, as in your homeless shelter map example,
> perhaps these applications aren't digital. Okay: I certainly agree
> that helping people access information relevant to their lives is a
> social good. I've written angry letters about reductions in library
> hours. But that wasn't open data activism. And I worry that saying
> "open data is good, but public data is better" is like saying
> "budgetary data is good, but financial literacy workshops are better":
> apples, oranges.
>
> Yes, the ultimate goal of open data policies is to help people access
> relevant information. Yes, just tossing data online -- particularly in
> Canada, where there's no significant institutional/nonprofit presence
> around open data -- will largely result in uses that are either
> commercially lucrative or at least somewhat self-interested, i.e.
> websites that appeal to the Twitter set. And no, publishing a dataset
> does not magically absolve government of all further social
> responsibility.
>
> But if we allow "open data" to balloon into "help people with
> knowledge!", we'll never get anywhere. Open data still has lots of
> benefits: for the potential it provides, it comes at an exceedingly
> low cost once cultural barriers are overcome. An ideal open data
> policy would include outreach and funding so that data gets used by
> outsiders in social useful ways. But if we increase the scope of our
> common argument beyond that, into the general provision of
> information, we confuse people about what open data is. And given that
> the largest challenge facing open data is probably marketing and
> cultural change, I think being clear about what we mean is pretty
> crucial.
>
>
> On Thu, Jul 7, 2011 at 12:54 PM, James McKinney <[hidden email]> wrote:
>> Nik, the argument is not fallacious. You are just disagreeing on
>> terms. I agree with your definition of open data. What the Sun is
>> asking for is public data, not open data. In brief, public data is
>> open data that is citizen-ready. That means ready for all citizens,
>> not just those with tech savvy. "Pure" open data is only
>> technically-savvy-citizen-ready. You write that it is not the
>> government's responsibility to provide public data, only open data.
>> The Sun's argument is that the government _ought_ to provide public
>> data, not just open data; that it ought to serve all citizens, not
>> just the tech-savvy. Much of this data is relevant to everyone,
>> therefore everyone ought to be able to use it (so the argument runs).
>>
>> You write that there's nothing wrong with relying on developers to
>> build open-data-based apps for the public, and back it up by saying
>> that the public already relies on developers to build all sorts of
>> apps. Let's have a look at these other apps. Do they target, engage
>> and empower the marginalized, the underprivileged, the poor? No; it's
>> not a hot, profitable, or easy market. But what about government?
>> Should it also ignore these people? Of course not. That's why it's not
>> enough to rely on developers. Developers do not have hearts of gold.
>> This is the argument the Sun is making.
>>
>> I find nothing objectionable with the Sun article. All it's saying is,
>> "open data is good, but public data is better; don't forget public
>> data in your open data push."
>>
>> On Thu, Jul 7, 2011 at 11:03 AM, Nik G <[hidden email]> wrote:
>>> The fallacy of this article's argument is that data isn't "truly open" until
>>> the Government builds or implements interfaces, applications & tools to make
>>> the data more accessible to non-technical audiences. The whole point of open
>>> data is to enable Government as a Platform for others to build upon, where
>>> the Government does the least possible to get the data out into the open, so
>>> that _others_, not Government, can then build visualizations, applications,
>>> mashups, etc.
>>>
>>> The key to getting more citizen-ready applications & visualizations is
>>> fostering community development, working in tandem with the local open data
>>> & open gov advocates to scale through others. There's nothing wrong with
>>> citizens relying on web-savvy developers to build apps, that's how it's done
>>> for thousands of "non-opendata" apps that we use every day on the web,
>>> mobile, desktop. The point is that our governments need to build
>>> competencies in harnessing the skills & knowledge of communities to take
>>> that open data from raw form to a citizen-ready app.
>>>
>>> City of Toronto has a great strategy in pushing to make open data as part of
>>> each department's workflow; it's visionary in that this is a cultural shift
>>> and not a point-in-time activity. But as with any initiative that's pushing
>>> the cultural, process & technology boundaries, there are cost/benefit and
>>> immediacy/relevancy trade-offs. In case of open data, as long as the data is
>>> as close to the source, original dataset, there's nothing wrong in getting
>>> it out in a machine-readable format under an open license sooner rather than
>>> later. By the way, Toronto's catalogue file formats aren't that drastically
>>> different from any other catalogue, providing your usual CSV, XML, XLS
>>> formats that are "spreadsheet-ready". If it's the geographic formats that
>>> _seems_ to be "user unfriendly" (ESRI Shapefile vs. the commonly-used KML),
>>> there are also relatively simple ways to convert those GIS formats without
>>> having to sacrifice staying close to the original data formats used by the
>>> city.
>>>
>>> Open, Rinse, Repeat is the recipe for success, not Wait Till it's Perfect &
>>> Pretty, but Outdated.
>>>
>>> I'll take open data SOONER rather than later, ANY format rather than no
>>> format, and NOT having to wait months or years till it's "perfect", but
>>> meets everyone's requirements according to everyone's technical comfort
>>> level.
>>>
>>> Nik Garkusha @Nik_G
>>> http://openhalton.ca
>>> http://port25.ca
>>>
>>> From: Tracey P. Lauriault
>>> Sent: Thursday, July 07, 2011 9:22 AM
>>> To: civicaccess discuss
>>> Subject: [CivicAccess-discuss] Toronto Sun: Toronto�s data open but almost
>>> useless
>>>
>>> Toronto�s data open but almost useless
>>> http://www.torontosun.com/2011/07/06/torontos-data-open-but-almost-useless
>>>
>>> This is the first news article I have seen in Canada to date that questions
>>> data access for citizens not just designer/developer/academic community:
>>>
>>>> �Right now, the primary audience is the local designer/developer/academic
>>>> community,� Garner says. �However, we have plans to build it out and
>>>> position it more clearly within the context of Open Government in a way
>>>> which would have more tangible benefits for a wider audience.�
>>>> It�s time for those plans to become action. As long as citizens have to
>>>> rely on web-savvy developers to do the hard work for them, the data isn�t
>>>> truly open.
>>>
>>>
>>> However, I am not sure which format the article's author wants the data in,
>>> xls is probably the lowest common denominator and many of the City's data
>>> are in those formats.� The City of Toronto Catalog has ESRI shape files,
>>> those are GIS files and there is no making those easier either.� How does a
>>> city decide on formats?� Should they be releasing data in the way that they
>>> use them in the formats used as part of a city's business processes or
>>> should they re-format the data for the public?
>>>
>>> Geogratis took the approach to release data as they use them.� This keeps
>>> their costs down by not adding any work load, it also keeps the data
>>> accurate, as conversion can introduce errors in geomatics files.� It also
>>> means that users need to know how to work with those formats and do the
>>> conversions themselves and bear the risk of errors as well.
>>>
>>> The article is also a bit erroneous, as the Edmonton Catalog has many
>>> formats that are difficult to use by lay people.� Useful for those creating
>>> mashups but not great for lay people.
>>>
>>> So interesting that it brings up citizen use but not quite an accurate
>>> picture.
>>> --
>>> Tracey P. Lauriault
>>> 613-234-2805
>>> http://traceyplauriault.ca/
>>>
>>>
>>> ________________________________
>>> _______________________________________________
>>> CivicAccess-discuss mailing list
>>> [hidden email]
>>> http://lists.pwd.ca/mailman/listinfo/civicaccess-discuss
>>>
>>> _______________________________________________
>>> CivicAccess-discuss mailing list
>>> [hidden email]
>>> http://lists.pwd.ca/mailman/listinfo/civicaccess-discuss
>>>
>> _______________________________________________
>> CivicAccess-discuss mailing list
>> [hidden email]
>> http://lists.pwd.ca/mailman/listinfo/civicaccess-discuss
>>
> _______________________________________________
> CivicAccess-discuss mailing list
> [hidden email]
> http://lists.pwd.ca/mailman/listinfo/civicaccess-discuss

_______________________________________________
CivicAccess-discuss mailing list
[hidden email]
http://lists.pwd.ca/mailman/listinfo/civicaccess-discuss