R-86-857 - 4/24/1986WHEREAS, the City has duly advertised for bids to lease a Data
Conversion System for the Library; and
WHEREAS, Xa:0 submitted the lowest
and best bid; and
WHEREAS, the Council wishes to accept the bid of p)��A
, and to authorize the execution of the
necessary documents, Now Therefore
BE IT RESOLVED BY THE COUNCIL OF THE CITY OF ROUND ROCK, TEXAS
That the bid of ,,
accepted as the lowest and best bid, and the Mayor is authorized and
directed to enter into a lease agreement with
for a Data Conversion System for the Library.
RESOLVED this 24th day of April, 1986.
ATTEST:
JO
NE
LAND, City Secretary
RESOLUTION NO. 57R
is hereby
MIKE ROBINSON, Mayor
City of Round Rock, Texas
RETROSPECTIVE CONVERSION BIDS
Bid specifications were mailed to: AMIGOS, Auto- Graphics, BIBLIOFILE, Blackwell
North America, Brodart, Computer Company, GRC, LSSI, Maceta, MARCIVE,'MITINET,
OCLC MICROCON, SAZTEC, UTLAS, LASERSERRCH (Ingram), and the Library of Congress
We received bids from UTLAS, AMIGOS; Computer Company, GRC, LSSI, MARCIVE,
Brodart, and the Library of Congress. There was a mail return from SAZTEC. .
Bib specifications were designed by Bob Walton, Automation Consultant, Texas
State Library.
In this situation, it is generally difficult to determine which is the the least
expensive bid as very different methods are used." In consultation with the
Automation Consultant, three vendors were preferred. These included LSSI, GRC, and
MARCIVE, in order 'of preference. In our discussions, we felt that processing our
records in -house was going to be the most cost and time efficient approach as
problems could be solved relatively _quickly, rather than waiting for printed
reports. The concept of mailing out the shelflist (inventory record) was quickly
vetoed. As it turns out, this is the most expensive approach, as well.
LSSI is being recommended for the following reasons: -
*Variety of search options, making it easier to find a matching record
*Use of laser videodisk which is extremely fast •
*Representatives in Texas
*Lease of all equipment at an extremely reasonable rate
*Ease of immediately modifying records copied from laser videodisk
*Keeping shelflist on site/ Avoiding' copying and refiling shelflist (34,000
cards) ,
*Ability to finish one'section completely without waiting for reports
*Choice of entering "no- hits" locally or sending them out to a vendor with a
greater number of records
DESCRIPTION OF PROJECT
History and brief explanation of "Retrospective Conversion"
In any automation project, the essential first step is to translate a library's
holding from a paper format into a machine - readable, or computer - readable, format.
The industry calls' this "retrospective conversion ". It can be also called data
conversion ".
In the past, before the advent of the PC's, a library would send out its shelf
list or photocopy it, and a firm would enter the data. The library would get a 9
track tape to place on its purchased computer, and then go hack and change any data.
This left the reliability of the data in limbo. Another choice was to get a copy of
the records in print format, at a per item cost in most cases, to verify that the
description did, indeed, match the shelflist card, yet data in the record count not
be changed immediately. ,
With the arrival' of the PC, a has been to do in -house conversion. When
this option was first offered, the library typically would _purchase or lease the
equipment and software, and have temporary staff and volunteers do the inputting of
.data. This would usually consist of inputting an "ISBN" or "LC° number only. When
a diskette was full of these numbers, the library would send the diskette to the ,
firm,. which would match it against its Library of Congress records. A "hit" would
be added to a tape, and "no-hits" would be the responsibility of the library to
resolve. Also, the individual library's call number and purchasing information
would not be included, and would need to be added at a later date.
•
As the PC, gained greater acceptance, some of the vendors began to offer the
.. option' of including not only ISBN or LC input, but also call number input. As the
technology in computing improved, laser technology allowed the industry to put all
- of the Library of Congress` holdings on videodisk, and as of the past few months,
CD -ROM (Compact Disk- Read Only Memory). To go with this, software was developed to
allow the library to copy the LC record onto its own diskettes, modify the record
and' include call numbers, special subject headings, purchasing /donation /memorial
information; and location codes. All of this used to be done in stages, for each of
the 34,000 records. It can now be done in one step. -
Work flow using LSSI system
The library would lease two IBM -PCs, monitors, cables, videodisk players,
controller - cards, and software for about 6 months. The library would provide the
labor through the CETA program in the summer, volunteers, and an outside contract
for temporary data conversion labor. Labor would be covered under the bond issue and
can be - thought of in the 'same light as hiring a carpenter to change walls in a
building as -part of a renovation bond issue. Data conversion, as indicated above, .
is an essential part of the automation project, and is cheaper to do in - house or
through a temporary contract, rather than sending out the entire shelflist or a
photocopy of the shelflist.
Costs using LSSI C-
„
The library looked into the cost difference of purchasing the IBM - PCs rather
than lease, but there was only a $500 difference between the cost of the bid without
equipment as opposed to cost with equipment.
° The cost for 6 months of use of the LSSI software and hardware will be $5,000.
Other costs include $150 for diskettes, $50 for loading the diskettes onto tapes.,
and about $4,300 for contracting labor for data conversion for a 6 month period,
about 20 hours per week. The total for the proiect is about $10,200, with $5,000 to
be paid in advance to LSSI, and $50 at a later date for loading two tapes. Labor and
diskettes will be purchased locally.
GRC
REASONS FOR REJECTING GRC AND MARCIVE
The Library. Director and Automation Consultant were pleased with the use of
•CD - ROM by GRC. ',Their main concerns were:
1) limitations to searching for a record by author /title only. ' An additional '
search option of ISBN and LC could be quicker, with author /title search being an
option if ISBN and LC failed; and,
2) the relative newness to the market of CD - ROM technology would indicate that
6 - 10 months down the road the hardware would be less expensive, and there would be
some computer architectural changes which might make searching of multiple CD -ROM
units quicker; and,
3)` related to 2) above, the current expense and slowness of 4 CD -ROM players
"daisy- chained” in order to get the speed offered by LSSI. In the LSSI system, the
entire Library, of Congress data base is on one platter. With the CD -ROM units, the
data base is on 4 CD's. The other, slower,option, is to make 4 passes through one
shelf list drawer to pick up all the records an the 4 CD's.
The library would be interested in this option at a future date for maintaining
its new records, but would rather wait until PC and CD -ROM purchases drop further.
Costs for this project would have been about $2300 more for software lease and
purchase of one ROM reader. Purchase of an IBM -PC would have been extra. It also
appears that translation ,Of diskettes into tape format would have been about $400
more than ;„LSSI's, service. Both require some staffing, as configured above'in the
LSSI discussion.
MARCIVE:
.
The MARCIVE firm offered two alternatives. One was to use their software, the
. other to use our word processing software. The project would have cost about $7700
plus staff labor as indicated in the LBSI configuration, as well as purchase of an
• IBM - compatible machine. This project uses the method refered to earlier in that
ISBN and LC numbers, with local call number, is inputted, records are matched, and a
database created. Any changes to the record could not be made until the database is
loaded onto the library's hardware. Also, records which a "no -hit" could be
inputted by the library by filling out an original cataloging form, which would then
be inputted by a MARCIVE' staff member. This step seems cumbersome as some staff
member would need to recopy on a coding sheet what is already on the shelflist card,
send this information onto Marcive,, and then someone else input the data. It is
more time and cost efficient, to do this on a PC and transfer the information onto a
diskette which in turn would be directly inputted onto a tape. An extra step is
saved.
The Library is interested in sending to MARCIVE a•copy of its ISBN and LC
numbers for records. which do not match the LSSI LC -MARC data base. Estimated cost
is $1360.
REJECTION OF OTHER BIDS
Other, bids were rejected for at least one of the following reasons. It is our
assumption, based on the Library's prior experience with ordering cataloging
information from Baker and Taylor and Marcive, that of the 34,000 records, BO% will
require original cataloging or modification of a record, primarily in the pagination
and publisher area.
* Shelflist required to be sent out,'often resulting in a bid exceeding $20,000
*Inconvenience due to a lapse of time between sending records and getting a
"hit" o r "no -hit" list
*Need to correct records at a later date, including the addition of local notes
' *Need to resolve through a separate contract the problem of original cataloging
*Purchase of hardware, which the library is not ready to do at this point -
*Use of a microfiche based system requiring someone to look up a number for each
book, send out the information on a diskette,, and then wait for records
*Stipulation on ownership of database (no clear title to use records as the
library preferred).
Specific details on each vendor is available.