USGS Exploring a Primitive Spatial Data Transfer Standard (PSDTS)

==== Russ

Map Data for Dummies.

Standard data formats are the problem.  I've been a scientist for about 35
years in the Air Force, CIA, NOAA and USGS.  I would play with data.  I used
to call it cryptology.

What I wanted as a scientists was not more sophisticated tools.  If it is a
series of lats and longs, I just wanted to look in and see it.

If you could do it in a word processor, then it would work.   The trick was
to get it down to where people could work with it.

I wanted to know where the states go.  Three months later, I still couldn't
find it.  Today, I still don't have it.

In the seven years I've been at USGS, I've never seen it.

If you are a cartographer or a GIS programmer, this is no problem. But for
most people looking at a LandSat image, this stuff makes no sense.

What is UTM?

I have a standing policy that I would buy anyone lunch if you visit.
If you could show up in the morning with your data, I could always display
it by lunch.  I have yet to buy anyone lunch!

When it comes to a map, rather than primitive data, the lines are the data.
What we really want to know is where the points are, where the lines are.
Where they are and what they are.

The problem with GIV is that you had no way to attribute.

If you have a table of lats and long.  This is all the data you need to
construct any map.  Polygon assignment of lines.

What happens when you move a point or cut a line.

Anyone who wants to use GIS is going to have a GIS system.

A lot of the stuff we have been publishing has to do with line weights,
colors.  You want to find the latitudes and longitudes of Route 29 on top of
a satellite image.

SDTS is very difficult for a County government that does not have a PC.

Now it doesn't solve the problems of the GIS community or metadata.  The
issue is, how do we get the primitive data in a form everyone can read.

"Primitive" - basic, underived information

General format should be in ASCII.  First thing is, people should be able to
look at it.  It has to be in lat and long.  Coordinate systems don't work.
When you actually reproject them, they are wrong.

Put the attributions with the data.  

For most people, the overhead of separate attribute data is too hard to learn.

There are two ways to attribute: the way DLG3 does it and the way everyone
else does it.  The tag method works better because you have things like
"intermittent" which apply only to streams, not roads.

Geography professors want this tool.

Q: [Martin] What do we do with SDTS?

This is not an official action.  SDTS will be there. What we are proposing
is map data in this primitive form.

Q: It is really amazing to hear you proposing this.  There are going to be
translators - again and again.

Have you looked at SDTS?

Q: Yes.  But for you to say that metadata is not important.

I didn't say that.

Q: The reason SDTS was created was that we don't have meta data we need.

This has nothing to do with metadata.  It is a standard for only one
purpose.  Look at DXF - it works with architectural drawings, but not with maps.

All these formats were developed for special projects.  This one is too.

I don't think formats are the problem.

Q: We don't care how many lines of code you have to write. People don't want
to write code.

USGS does not provide tools that anyone can use in schools.  Complicated
database formats does not solve the problem.

-o-