[Xastir] Openstreetmap?

[Gerry Creager]

A couple of points...

Brad Douglas wrote:
> On Sun, 2007-10-07 at 23:08 +0100, Dave H wrote:
>> O.k thanks for that - I'm no geo-whatever expert - in fact most of the
>> acronyms 
>> floated in here mean very little - i suspect to many this side of the
>> Ocean. 
> 
> I am a geo-whatever expert. ;-)  Most of acronyms used here are used in
> the Geographic Information Systems (GIS) and geo-informatics industries.
> It's a small, large industry in the sense that it is everywhere, yet few
> have heard of it.

Glad to have you.  I do a little in the industry, too, with the 
technical committee of the Open Geospatial Consortium.  I'm currently 
active on the Defense-Intelligence working group and Sensor Web 
Enablement.  I'm also working with the somewhat fractious Coordinate 
Transformations ad hoc group, which I think will achieve a working group 
status in December.

>> It seems to me your so lucky in the US that your public-tax-$
>> investments in 
>> geo-data collection - the data seems  to be handed back freely back to
>> you.. 
>> Certainly in the UK either we have no of our own or its damn-secret or
>> we 
>> have to pay a second time.
> 
> This is true that the US has been inherently blessed in the past with
> good quality data.  This is changing at a fast (and IMO, alarming) rate.
> There really hasn't been truly freely available orbital data since
> LANDSAT-7 and the SRTM shuttle mission.

Depends on your sources.  Data here in Texas are good, free, and readily 
available.

> Street and feature data extracted from Census data is really a byproduct
> of the Census, complete with varying amounts of error.  This error can
> be easily  demonstrated on this page a friend did on the subject:
> http://casoilresource.lawr.ucdavis.edu/drupal/node/433

Nice site.  Had not seen that before, but it's interesting. 
Unfortunately, Census, who uses the old DRG/STDS mapping info from USGS, 
also munges the data for their purposes.  I really wish they'd not made 
their data widely available and so discoverable since they don't 
maintain good (or adequate or even minimal real) metadata.

Unfortunately, a simple affine isn't always satisfactory, nor is the 
affine transform matrix consistent for more than a small area (with 
"small" being somewhat relative).  Wish it was true but to correct all 
of Tiger, you'd not only have to do all the county offerings, but 
actually would end up subsetting some and doing unique transformations 
to get them all right.

> I've suggested that at a minimum, OSM should include metadata specifying
> projection parameters, but it falls on deaf ears.  It would be much
> better if users either uploaded data consistently (and rejected
> outliers) or the system reprojected upon upload.

I'd be able to come up with a minimum metadata list, but I fear I'd hit 
the same response as you already have.  At a minimum I'd love to know 
what datum is being used and then transform to a common datum, ITRF2000 
or WGS84 (2005).

> OSM may be a useful last resort for Xastir in areas where better data
> does not exist.

Unless you're thinking you'd get better data in Europe with OSM than you 
can find now, OSM looks too sparse to be of a lot of use, anyway.

>> Xastir in the UK is often poor looking simply due to lack of decent
>> map sources 
>> and probably zero overlays. Either we don't have them or some buggers
>> got 
>> copyright over things we paid for once already. We get desperate or
>> use outlines
>> for lack of much else.
> 
> Aside from Census data, that is largely the case here in the States,
> too.

Ordinance Survey tell me (and I asked specifically about this at the 
summer OGC meeting) they are working on a policy statement and fee 
structure to provide data of this sort to certain users at solely cost 
recovery.  Should happen this calendar year.  And radio amateurs are now 
added to the list.  I confirmed this 2 weeks ago at the Fall OGC meeting.

> I don't know how things of this nature function in the UK, but here we
> are able to request data from local municipalities.  I've had no trouble
> getting needed data for projects that local governments have collected.
> 
> Have you tried asking various levels of government for a Shapefile of
> roads? Try asking for street center-lines, first.  They are generally
> high accuracy.  Be specific of what you want so that you aren't creating
> work that they don't need to do...and they might respond to you
> favorably in the future.

Lots of municipalities here and over on the other side of the pond have 
a fee-recovery system.  Sometimes I can get their data by asking, 
sometimes by paying, and sometimes I have to use other means, like going 
to their state or the Feds.  I expect you'd see the same thing in UK.

gerry
-- 
Gerry Creager -- gerry.creager at tamu.edu
Texas Mesonet -- AATLT, Texas A&M University	
Cell: 979.229.5301 Office: 979.458.4020 FAX: 979.862.3983
Office: 1700 Research Parkway Ste 160, TAMU, College Station, TX 77843