University of Waterloo
                    Electronic Data Service Task Group
                     Minutes of Meeting April 6, 1994


Present:  Albert Kemp
          Sue Moskal
          Doug Morton (Chair & Sec.)
          Richard Pinnell
          Shabiran Rahman
          Debbie Tytko
          Carol Vogt

CC:       Bruce MacNeil
          Mike Ridley

1.   Minutes of Previous Meeting
     none

2.   Business Arising
     To be covered below.

3.   Additional Agenda
     7.   Storage of data - Carol Vogt
     8.   Demo of 1991 Census on CD-ROM

4.   CANSIM/EPAS
     Carol Vogt reported that the disc drive had been ordered but
     she was unaware of the status of orders for the CANSIM data
     or the EPAS software.

5.   Census and GSS data
     5.1  Inventory      Sue Moskal has received a list from
     Laine Ruus of the 1991 Census data that has been
     distributed.  A comparison with the lists distributed last
     week confirms that there are gaps in our holdings
     5.2  Description   As data is received, it appears that
     descriptions come with them.  The data is such that
     abstracts are unlikely to be appropriate.
     5.3  Acquisition    Carol Vogt indicated that more files are
     ready to be FTPed.  Albert Kemp and Debbie Tytko will get
     together with Jan Willwerth to train on the process
     5.4  Census Geographic Data    Richard Pinnell has spoken
     with Laine Ruus about the census geographic data.  The data
     were in ArcInfo format and have all been distributed, the
     consortium did not purchase the Street Network files nor the
     Enumeration Area files.   Richard has a proposal going to B.
     MacNeil and M. Ridley to purchase the Street network files
     in Mapinfo format.

6.   User Interface
     The interface was discussed briefly, Shabiran Rahman
     distributed the "About the EDS" page she is developing. 
     Committee members will look at other data service Gophers to
     see how others have done it.

7.   Data Storage
     Carol Vogt discussed the impending storage problem with
     respect to the Census (and other) data.  She has exhausted
     the avenues open to her for finding space for the 15-20
     Gigabytes of data we expect to have coming in.  Currently
     the Census and GSS data are stored in several locations,
     making access virtually impossible.

     In the past, data came on tape and stayed on tape. It was
     cheap, reliable, and access from CMS was relatively easy.

     With the changeover to Unix, there is one tape drive,
     attached to Watserv1, which is not as easy to use as on CMS.

     Attic has been suggested as a storage option. Attic is a
     hybrid technology where data are stored on disc and, after
     some time, transferred to Exabyte tape for longer term
     storage.  In theory an anonymous FTP site could be set up
     for EDS data.   In practice, the Exabyte system is not as
     reliable as desired and the licence would need renegotiation
     for storage of this much data. 

     A move to a system similar to Attic, using optical discs,
     might be more reliable but would be more expensive.

     Carol suggested that the cheapest long term solution may be
     to put all the data online (the way CANSIM is being handled,
     though the software would not be the same).

     It was decided to approach Mike Ridley and Bruce MacNeil as
     soon as possible to begin working on a solution, which is
     likely to be expensive.

8.   E-STAT
     Richard Pinnell reported that there will be a demonstration
     of E-Stat (a subset of Census data with mapping
     capabilities) on May 5; time and place to be announced.

9.   Next Meeting
     April 13, 1994,  1:30 pm  in LIB 407   **NOTE DIFFERENT
     ROOM**