Dear All,

Thank you to everyone who offered suggestions and assistance in finding an
answer to this "puzzle".

This SUM is rather long since I received a wide variety of suggestions in
response to my question about opening 1980 STF1A and STF3A data files
downloaded from the Census website. I have only begun to try these
suggestions but will send a SUM anyway, in case anyone out there would like
to give it a try as well.

Several people suggested buying the data already processed from Geolytics at
www.geolytics.com
I'm not ready to resort to this--yet.

Several other websites that have data and info on the data were suggested:
sedac.ciesin.org
        homer.ssd.census.gov/cdrom/lookup

Checking at a University Library (Suzzalo at UW) in the government
publications section was suggested. I know that libraries are often
repositories for Census data.  Often the data is still in its original
format, though.  This data also may be restricted to students, staff and
faculty for academic purposes only.

A couple people mentioned that the data file is UNIX.  Using the UNIX2DOS
utility before running crlf.exe was suggested.

Several people suggested using statistical packages and other software:

        SAS or SPSS (might require some programming to make it work)
                Text editors such as Textpad and UltraEdit-32 (both
downloadable free/shareware)
                In UltraEdit-32 create end-of-record markers,
                Import this text file into Access and when the "import
wizard" opens,
                                define the positions of each field. (An SPSS
syntax file, which
                                will open in Word, was attached with the
field names, record
                                lengths and field descriptions.) I will
forward a copy of this file to
                                anyone who would like to give this a try.

Several more suggestions:

Use a utility called "Extract" that will put the data into comma-delimited
format. This utility is available at ftp.census.gov/pub/extract

If the records are fixed-length, they can be defined in a template in INFO
and the data copied into the template (using INFO's GET with the "copy
ASCII" keywords).

Import them into ArcView and then use Xtools to export out to an Excel file.

In Arcview add the file as a new table and then in an Avenue script, parse
each record using the Avenue Extract request (not related to the Census
Bureau utility)

Pkzip has a suite of products that offers an option, which in DOS will
automatically put in a carriage return as it unzips the file.

I hope I didn't't miss any suggestions in this summation.  When (and if) a
solution as to how to turn this original data into a usable format becomes
evident, I will be sure to forward a follow-up summation.

Again, thanks to the many who offered their ideas and assistance.


Nancy West
GIS Analyst
House Democratic Caucus
(360) 786-7843