Re: [Talk-us] Legislative districts, Land-use zoning, etc.

[Ray Kiddy]

On Wed, 21 Oct 2015 08:19:20 +0200
Frederik Ramm  wrote:

> Hi,
> 
> On 10/21/2015 04:46 AM, Ray Kiddy wrote:
> > To me, OSM is a tool which is ideal for relating various information
> > layers across a multi-dimensional substrate. This substrate is a
> > two-dimensional geography, which is defined geographically. To me,
> > it seems perfect for things like borders.
> 
> OSM is first and foremost a community of people curating a data set.
> This process works best with data that is verifiable on the ground,
> because if two community members disagree over something, the dispute
> can be resolved by simply looking at the place. Also, the mapper
> (surveyor) is the ultimate authority in OSM; we map what *is*, not
> what some government says should be.
> 
> We do have a few items that go against these principles, most notably
> borders. They are not easily verifiable, and they are items where the
> authority lies elsewhere - where OSM can only ever be a copy of some
> master data being defined by a government, instead of being the
> authoritative source. OSM is certainly not "perfect" for collecting
> and curating such information; this is a fact and not a matter of
> personal opinion. Having these borders in OSM is already a compromise
> where the usefulness (high) has been weighed against the suitability
> of OSM as a medium (low).

I am seeing the truth in what you are saying now.

First, I am still somewhat new to the OSM game. But also I am
interested in the use of it for borders for, for example, school
districts in the US. And I am seeing that (in line with Richard's
suggestion, different e-mail), I may want to investigate doing that in
a separate database connected to OSM. And I created just such a
database several weeks ago, so yes, that makes sense. I am currently
writing software which keeps track of the relations and which
periodically checks their integrity.

> > It is very true that, as you say, OSM "excels at holding information
> > that users can see, verify and update." I think it is also true that
> > OSM excels at relating abstract themes in a multi-dimensional space.
> 
> I can't process the use of "multi-dimensional" in this context. OSM is
> not multi-dimensional, it is 2.5-dimensional at best, and affixing
> bits and bobs of extra information to some objects doesn't make it
> multi-dimensional. OSM certainly does not excel at relating abstract
> themes - the contrary is true, OSM is about concrete stuff. As soon as
> we veer into the less concrete - for example, public transport
> relations instead of steel tracks on the ground - we hit the limits
> of our editing tools, and of most people working with OSM too. Yes we
> do that (public transport relations) but we certainly don't "excel"
> at it.

I meant "dimension" in terms of themes. So a map (2 d) with a layer for
average family income, a layer for electricity usage and a layer for
foliage coverage is a 5-dimensional map. Like that.

> > And OSM is many, many others things as well. Many others would
> > define it differently and all of those would also be valid and
> > useful.
> 
> > All of our viewpoints are valuable, and it is more clear that this
> > is true when we describe our viewpoints as viewpoints, not as norms.
> 
> I think this lovey-dovey relativism doesn't go anywhere. To me, it
> smacks of "well, the scientific method is one way to look at physics
> but of course there are many others that are equally valid and
> useful". OSM is certainly not whatever anyone sees in it, and
> certainly not all these views are equally valid and useful.
> 
> Bye
> Frederik
> 

"Lovey-dovey" :-) I like that. I usually have been accused of not
being, shall we say, "lovey-dovey". Perhaps I am just trying to be
politic and have sung the pendulum too far.

The points I am seeing from you all make sense, so I stand corrected.

thanx - ray


___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us

Re: [Talk-us] Legislative districts, Land-use zoning, etc.

[stevea]

Ray Kiddy writes:

 > It is very true that, as you say, OSM "excels at holding information

 that users can see, verify and update." I think it is also true that

 > OSM excels at relating abstract themes in a multi-dimensional space.


I agree.  There are hundreds or thousands of clearly themed maps 
based upon OSM data, all and any of which are useful for that 
particular narrow slice of query that a map consumer wishes to see 
(literally, visualize).  Building hundreds, thousands or millions of 
these visualizations based upon the rich data found in OSM truly is 
what OSM is about:  the ends DO justify the means in this case!  In 
short:  why bother building a rich database or don't bother to ask it 
to provide beautiful answers?  That would be silly!  We can, and we 
do.


And Frederik Ramm replies:

I can't process the use of "multi-dimensional" in this context. OSM is
not multi-dimensional, it is 2.5-dimensional at best, and affixing bits
and bobs of extra information to some objects doesn't make it
multi-dimensional. OSM certainly does not excel at relating abstract
themes - the contrary is true, OSM is about concrete stuff. As soon as
we veer into the less concrete - for example, public transport relations
instead of steel tracks on the ground - we hit the limits of our editing
tools, and of most people working with OSM too. Yes we do that (public
transport relations) but we certainly don't "excel" at it.


On the contrary, OSM is absolutely multi-dimensional:  "bobs of extra 
information" (in the form of our super tool, free-form tagging) DO 
make it multi-dimensional.  That's the beauty of an abstract 
dimension:  it can be defined to be what you want it to be.  Often we 
start with the two dimensions of "earth's surface" then we choose a 
richly-defined theme to be the third (or include a fourth or fifth). 
This is simply abstract thinking applied, and to say that a dimension 
must be "space" (as in 2-space or 3-space or "2.5-space at best") and 
space ONLY is so very limiting.  Space is a good place to BEGIN using 
the 2 dimensions of "earth's surface," but after that, OSM is so 
wonderfully useful PRECISELY because we use it in "creative, 
productive, or unexpected ways" (just like our Main Page says). 
Those other ways might be abstractly defined as "multi-dimensional 
extensions of a geographically-defined database."  After that, as it 
is said, "the sky is the limit."


We COULD excel at public transport relations (real things, "findable 
on the ground"), we just don't quite yet.  OSM only having 
partially-implemented or not-quite-perfect public transport routes is 
not an existence proof that public transport routes don't belong in 
OSM or that they overly challenge the editing skills (or tools) of 
the project or its participants.


I reject the assertion that a public transport route is "less 
concrete" than, say, a drinking fountain.  Public transport routes 
have platforms, signs which display their timepoints, schedules, a 
beginning and end, etc.  They are a real, not abstract things, and 
OSM not only reflects this, we have done so with sane growth from 
public_transport=v1 to v2 in a way we should be proud of.  Sure, we 
have much more growth and data to enter to be an impressive and 
definitive source -- we are still a growing project.  Let us not 
dismiss this real, useful and actively growing subset of our data as 
"less concrete" or even its only faintly-hinted-at next logical 
conclusion of "these are unworthy data, so let's purge them."  This 
smacks of "changing the rules of the game in the middle of the game." 
Yes, we've done this before (e.g. old license to ODBL), but the 
process is painful, only works when we are honest and forthright that 
that's what we're doing, and most of us agree to do so.



 > And OSM is many, many others things as well. Many others would define

 it differently and all of those would also be valid and useful.



 All of our viewpoints are valuable, and it is more clear that this is
 true when we describe our viewpoints as viewpoints, not as norms.


I think this lovey-dovey relativism doesn't go anywhere. To me, it
smacks of "well, the scientific method is one way to look at physics but
of course there are many others that are equally valid and useful". OSM
is certainly not whatever anyone sees in it, and certainly not all these
views are equally valid and useful.


I don't want to put words in Frederik's mouth, but what I think he is 
getting at is that OSM is not a dumping ground for whatever we want 
it to be.  Yes, that is true, and a good point.  Do we need to manage 
what goes into and doesn't go into OSM?  Yes, of course.  Our core 
tenets (e.g. "on the ground verifiable") guide us well here.  But if 
we are going to change the rules mid-stream, let us say so and not 
pretend we are not.


SteveA
California

___
Talk-us mailing list
Talk-us@openstreetmap.org

Re: [Talk-us] Legislative districts, Land-use zoning, etc.

[Mike Thompson]

Many borders, particularly international borders, are prominently marked
("monumented") (e.g. [1]), and thus are verifiable on the ground (and
sometimes the monumentation is so prominent it is visible from imagery).
It is what is physically monumented on the ground that is the legal border,
from [1]: "If some of the original markers were off by a few dozen -- or a
couple of hundred -- feet here and there, which was inevitable given the
conditions in which the crews worked and the technology of the time, it
doesn’t matter because it is the position of the monuments on the ground,
not the 141st meridian, that is the de facto boundary by treaty."

Legislative districts on the other hand, because they can change (every 10
years in US), are not monumented directly, and therefore would be very
difficult to verify on the ground.  In the US one would have to look up the
official text description of the district, then look up the census blocks
it references (" most redistricting was based on whole census blocks.
Kentucky was the only state where congressional district boundaries split
some 2010 Census tabulation blocks." [2]), and then head out to the field
to observe the features that the census blocks reference. Census blocks are
generally defined by streets, rivers and other physical features [3].

Mike

[1]
http://www.adn.com/article/20140727/trail-monuments-men-border-crews-cut-20-foot-swath-alaska-yukon-line
[2] https://www.census.gov/geo/maps-data/data/aboutcd.html
[3] https://en.wikipedia.org/wiki/Census_block
___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us

Re: [Talk-us] Legislative districts, Land-use zoning, etc.

[Frederik Ramm]

Hi,

On 10/21/2015 04:46 AM, Ray Kiddy wrote:
> To me, OSM is a tool which is ideal for relating various information
> layers across a multi-dimensional substrate. This substrate is a
> two-dimensional geography, which is defined geographically. To me, it
> seems perfect for things like borders.

OSM is first and foremost a community of people curating a data set.
This process works best with data that is verifiable on the ground,
because if two community members disagree over something, the dispute
can be resolved by simply looking at the place. Also, the mapper
(surveyor) is the ultimate authority in OSM; we map what *is*, not what
some government says should be.

We do have a few items that go against these principles, most notably
borders. They are not easily verifiable, and they are items where the
authority lies elsewhere - where OSM can only ever be a copy of some
master data being defined by a government, instead of being the
authoritative source. OSM is certainly not "perfect" for collecting and
curating such information; this is a fact and not a matter of personal
opinion. Having these borders in OSM is already a compromise where the
usefulness (high) has been weighed against the suitability of OSM as a
medium (low).

> It is very true that, as you say, OSM "excels at holding information
> that users can see, verify and update." I think it is also true that
> OSM excels at relating abstract themes in a multi-dimensional space.

I can't process the use of "multi-dimensional" in this context. OSM is
not multi-dimensional, it is 2.5-dimensional at best, and affixing bits
and bobs of extra information to some objects doesn't make it
multi-dimensional. OSM certainly does not excel at relating abstract
themes - the contrary is true, OSM is about concrete stuff. As soon as
we veer into the less concrete - for example, public transport relations
instead of steel tracks on the ground - we hit the limits of our editing
tools, and of most people working with OSM too. Yes we do that (public
transport relations) but we certainly don't "excel" at it.

> And OSM is many, many others things as well. Many others would define
> it differently and all of those would also be valid and useful.

> All of our viewpoints are valuable, and it is more clear that this is
> true when we describe our viewpoints as viewpoints, not as norms.

I think this lovey-dovey relativism doesn't go anywhere. To me, it
smacks of "well, the scientific method is one way to look at physics but
of course there are many others that are equally valid and useful". OSM
is certainly not whatever anyone sees in it, and certainly not all these
views are equally valid and useful.

Bye
Frederik

-- 
Frederik Ramm  ##  eMail frede...@remote.org  ##  N49°00'09" E008°23'33"

___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us

Re: [Talk-us] Legislative districts, Land-use zoning, etc.

[Ray Kiddy]

On Fri, 16 Oct 2015 10:10:00 -0500
Toby Murray  wrote:

> OSM is not the ideal tool to be a dumping ground for all GIS data. It
> excels at holding information that users can see, verify and update.
> Boundaries like this cannot be seen or verified by anyone except the
> government agency that originally made them. So the data gains no
> benefit from being in OSM and in fact makes it more difficult to
> update both the boundary data as well as non-boundary data in the
> vicinity.
> 
> Yes, we do have national/state/county/city boundaries. Some people
> aren't happy about this either... but they are tolerated because they
> enable geocoding functionality.
> 
> Toby

Toby, I am sorry but I think you are making a normative statement here
where it is not useful to do so.

Your statement above would have been perfect if you started it with "To
me, ...".

To me, OSM is a tool which is ideal for relating various information
layers across a multi-dimensional substrate. This substrate is a
two-dimensional geography, which is defined geographically. To me, it
seems perfect for things like borders. There are many kinds of "themes"
that appear in thematic maps, and administrative entities are a useful
theme.

It is very true that, as you say, OSM "excels at holding information
that users can see, verify and update." I think it is also true that
OSM excels at relating abstract themes in a multi-dimensional space.
And OSM is many, many others things as well. Many others would define
it differently and all of those would also be valid and useful.

All of our viewpoints are valuable, and it is more clear that this is
true when we describe our viewpoints as viewpoints, not as norms.

cheers - ray


> On Fri, Oct 16, 2015 at 1:19 AM, Charles P. Lamb 
> wrote:
> > Looking thorough the OSM Wiki Map Features entry there didn't
> > really seem to be features defined for such things as legislative
> > districts, polling place districts, and land-use zoning. I don't
> > think such features are peculiar to the USA. Am I missing something?
> >
> > Thanks,
> >
> > Charles P. Lamb
> > New Jersey
> >
> > ___
> > Talk-us mailing list
> > Talk-us@openstreetmap.org
> > https://lists.openstreetmap.org/listinfo/talk-us
> 
> ___
> Talk-us mailing list
> Talk-us@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/talk-us


___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us

[Talk-us] Legislative districts, Land-use zoning, etc.

[Charles P. Lamb]

Looking thorough the OSM Wiki Map Features entry there didn't really 
seem to be features defined for such things as legislative districts, 
polling place districts, and land-use zoning. I don't think such 
features are peculiar to the USA. Am I missing something?


Thanks,

Charles P. Lamb
New Jersey

___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us

Re: [Talk-us] Legislative districts, Land-use zoning, etc.

[Toby Murray]

OSM is not the ideal tool to be a dumping ground for all GIS data. It
excels at holding information that users can see, verify and update.
Boundaries like this cannot be seen or verified by anyone except the
government agency that originally made them. So the data gains no
benefit from being in OSM and in fact makes it more difficult to
update both the boundary data as well as non-boundary data in the
vicinity.

Yes, we do have national/state/county/city boundaries. Some people
aren't happy about this either... but they are tolerated because they
enable geocoding functionality.

Toby

On Fri, Oct 16, 2015 at 1:19 AM, Charles P. Lamb  wrote:
> Looking thorough the OSM Wiki Map Features entry there didn't really seem to
> be features defined for such things as legislative districts, polling place
> districts, and land-use zoning. I don't think such features are peculiar to
> the USA. Am I missing something?
>
> Thanks,
>
> Charles P. Lamb
> New Jersey
>
> ___
> Talk-us mailing list
> Talk-us@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/talk-us

___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us