Common Data Interchange Format

[David Mandel]

On 17 Nov 1995, Alastair Duncan wrote:

> Hello everyone,
> 
 ...
> >> This is correct, the standard is called SDTS (Spatial Data Transfer
> >> Standard) and will be mandatory for (as I understand it) all government
> >> authorities in the US down to, but not including, tribal government level.
> >> As from something like the middle of next year all new data will have to
> >> be available in this format and the agencies will have to begin making
> >> their old data available in this format. 
> >> A standard closely based on SDTS is being put up in Australia and New
> >> Zealand, and I believe that the ISO is examining it.
> >> This is excellent news for GISers as any GIS supplier which wishes to sell
> >> anything to US agencies will have to make this standard available in their
> >> packages - goodbye data transfer problems (we hope).
> 

I don't know what is happening with SDTS or if it is good or bad.  
To date, the standards documents I have read scare me.  The standards
look too general and too vague.   I'm not sure how a software writer 
would implement them.  In particular, it seems like so many options are 
allowed that writing a general SDTS reader will be a very major thing.  
On the other hand, I'm and outsider and I've only looked at a bit of 
the documentation.  Also, correct me if I'm wrong, but I believe USGS is 
working on SDTS libraries to assist software developers.

Ideally, it would be nice to have something better than a data transfer 
standard.  Given the growth in multi vendor (software and hardware) 
networks, especially the internet, it would be better if everyone 
used the same internal formats.  That way GRASS and LTxx and MapInfo and 
Arc Info and "Dave's Really Nice Total Solution Program (DRNTSP)" could 
all work together at the same time on the same data.  The need to 
constantly import and export data between different vendor's software 
packages is locking GIS people into whatever solutions they can get from 
a single vendor.  This tends to kill freeware and small commercial vendors.
I haven't read the OGIS documents, but I think they have been thinking 
along these lines.

On the other hand,  sometimes I wonder if it is theoretically possible to 
create any "good" GIS data standards.  There is such a difference of 
opinion among professionals regarding the information content that a GIS 
needs to have.

So much for my comments on GIS philosophy.

Dave  Mandel

 ==============================================================
 David Mandel, Linux Activist             dmandel at transport.com
 Internet Provided by:   Transport Logic         (503) 243-1940
               Serving the Pacific Northwestern USA
 Portland  *  Vancouver  *  McMinnville   *   Salem    *   Bend
 ==============================================================