ACRL News Issue (B) of College & Research Libraries 160 Managing automation for results: The role of the campus computing center By Ellen G. Miller Director, Library Systems D evelopm ent University o f C incinnati How librarians and data processors can collaborate to provide the best service to the end-user. T h i s is the second in a series of articles on p ra c ti­ cal m anagem ent issues facing the planners and im- p lem enters of lib ra ry in fo rm atio n systems. T he first article discussed using tim e productively (edit­ ing 500,000 O C L C records) w hile participative system selection (by library, d a ta processing staff, fa c u lty an d stu d en ts) proceeds; it a p p e a re d in C & R L N ew s, Novem ber 1984, p p .532-39. A central requirem ent for m anaging the p la n ­ ning and im plem entation of any autom ated library inform ation system is obtaining quality d a ta proc­ essing (DP) support. In the rush to figure out w h at systems best m eet local needs and find financing, how ever, this vital aspect m ay be overlooked or at least u n d erestim ated . Yet th e fact is th a t a u to ­ m ated systems require D P expertise. The m ore li­ b ra ry sites, integrated functions, records, queries, u p d ate transactions, staff involved and grow th ex­ pected, the m ore expert th a t expertise m ust be. T he key questions, then, for planners and im ple­ m e n te d are: W h a t D P support does m y library need, w h a t optional sources do we have, and how can w e m anage it effectively? To ignore these is­ sues, assum ing th a t som ehow they will w ork out, is to ta k e u n a c c e p ta b le risks. All lib ra ry ty p e s— college, university, school, special, governm ent— m ust find D P expertise. This article discusses th a t need in general w hile focusing on academ ic li­ braries in p articu lar. DP services needed Typically, libraries need m any kinds of services from th eir com puting center (CC) during the p la n ­ ning, im plem enting, and operational stages: • E v a lu a tio n of h a rd w are /so ftw are/telec o m - m unications offered by vendors of library systems. H aving DP staff atten d dem onstrations and be p a rt of site visits speeds up the lib rary ’s ability to w in ­ now am ong vendor offerings. • P reparing the w ritten Request F or In fo rm a­ tion (RFI) and later Request For Proposals (RFP). D P and library staff can w ork effectively in evolv­ ing docum ents th a t best m eet library needs. Sec­ tions in each docum ent on h ardw are/softw are/te- lecom m unications are m andatory. • E valuating vendor proposals. As anyone who has been through th a t process knows, evaluating v en d o r proposals is a t best a tim e -co n su m in g , m ind-expanding learning experience for the evalu­ ation team . At w orst, it is frustrating and time- w asting, e.g ., flipping am ong several volumes of vendor docum entation, never finding an answer th a t is allegedly there. DP staff see questions th a t require followup tele­ phone calls to the vendor’s D P experts; they are com petent to evaluate inform ation gained through those calls. T heir analyses of the hardw are/tele- com m unications configuration bid by the vendor will help prevent the unexpected and unw elcom e 161 Cr: UC Information Services L ib ra ry and data processing s ta ff being trained for B L IS by B iblio-T echniques instructors. h a r d w a r e u p g r a d e s r e p o r t e d b y Boss a n d M cQ u e en .1 Those upgrades ap p a re n tly w ere due to “lo w b allin g ” estim ates of custom er disk storage an d C PU requirem ents in o rder to secure th e con­ tra c t. • G ettin g specific o p eratio n al in fo rm atio n from libraries cu rren tly using th e v en d o r’s system. I b e­ lieve th a t no lib ra ry should select a system w ith o u t h aving key lib ra ry staff tak e site visits to see th e top tw o contenders’ systems in action. D P staff m ay join those visits, b u t also m ig ht be able to get in fo r­ m atio n on th e telephone from those lib raries’ com ­ p u tin g centers a b o u t operations an d vendor sup­ p o rt. • H elpingselect th e system. B oth functional and D P specifications m ust be m et in o rd er to select th e system. This m eans th a t both lib ra ry and D P staff need to be on th e team m aking th e final selection, w hich w ill be fo rw ard ed as a reco m m en d atio n to th e p a re n t organization. • P roviding expertise d u rin g c o n tra ct neg o tia­ tions an d finalization. E very co n tra ct w ill contain a section on h ard w are/so ftw are/teleco m m u n ica- tions c o n fig u ra tio n re q u ire m e n ts. R egardless of w h e th e r th e cam pus C C , th e vendor or a th ird p a r ty is p ro v id in g e q u ip m e n t, th e lib ra ry m ust have this p o rtio n of th e c o n tra ct carefully review ed by D P experts. W hy? To assure th a t th e co n fig u ra­ tio n is m ore th a n a d e q u a te to h a n d le file sizes, tran sa ctio n loads an d estim ated grow th. • H elping set up im p lem en tatio n calendars. Im - 1 Rich ard W . Boss an d Judy M cQ ueen, “A uto­ m a te d C irc u la tio n C o n tro l S y stem s,” L ib r a r y Technology Beports 18 (M arch-A pril 1982): 171. p lem en tatio n is th e process th a t brings together several strands of tasks. T hey include site p re p a ra ­ tion for individual lib ra ry sites (electrical, rem o d ­ elling, cabling), co m p u ter room p re p a ra tio n , o r­ d erin g h a rd w a re /so ftw a re , designing cabling links an d telephone line load balancing, designing ac­ ceptance tests, an d scheduling th e C C staff special­ ists needed d u rin g softw are installation. C C exper­ tise is needed to identify th e place of each task and stra n d in an overall im p lem en tatio n p lan . • D esigning screens. Some system s, e .g ., th e B ib lio -T ech n iq u es L ib ra ry In fo rm a tio n System (BLIS), p e rm it extensive design of screens an d screen sequences so as to m eet different end-user needs. T h a t’s useful w h en lib rarian s feel th a t the needs an d strategies of, for exam ple, a tw o-year technical college stu d en t v ary from those of a m ed i­ cal researcher. D P personnel m ust p a rtic ip a te in screen design com m ittees, w hich should also in ­ clude facu lty an d stu d en t m em bers, because th eir know ledge w ill help others u n d ersta n d th e p ro ­ gram processes going on b ehind those screens. • O verseeing eq u ip m en t installation an d in itia l­ iz atio n in lib ra ry sites. T h e C C staff w h o have w orked w ith th e lib rary th ro u g h o u t th e p lan n in g process have th e know ledge to oversee th e vendor, cabling personnel (w hich m ight be from th e C C , physical p la n t, or a su b co n tracto r), an d lib rary staff. T he goal is th e correct eq u ip m en t located in th e rig h t place th a t com m unicates w ith o u t e rro r to th e cen tral processing u n it (C P U ). • O p e ra tin g th e system. A fter being tra in e d by th e vendor, it is th e task of C C staff to o p erate the lib ra ry ’s system. T h a t m eans th a t th e system is to be d ep en d ab ly up a t agreed-upon hours, w ith p ro p er 162 security and backup being taken regularly. It also means th at the CC helps the library solve prob­ lems. In summary, DP support needed by libraries for autom ated inform ation systems ranges from edu­ cation, advice and review of documents to system operations. Sources of DP support T here are five m ajor sources of DP expertise available to most libraries: 1) parent organization com puting center, 2) vendor of selected system, 3) outside service bureau, 4) consultant, and (5) li­ brary staff after proper training. Here we concen­ trate on the first case, but a few comments on the other options are in order. The two m ajor criteria for deciding w here to go for DP expertise are 1) the degree to which the li­ brary wants to control its autom ated system; and 2) w hat it can afford. A related question is the degree to which it is ready and able to m anage its DP ex­ pertise. One of the attractions of “turnkey” systems is th a t they take m any responsibilities off the li­ brary’s shoulders. But as a leading vendor noted nonetheless, “A turnkey system implies th at the li­ brary will operate the com puter.” In reality, the library will always have more re­ sponsibilities m anaging its DP expertise than it first expected. My experience suggests th at the library has more leverage over its DP supplier the closer it is organizationally. In other words, a library th at uses its p aren t organization’s com puting center should be better able to get the dependable, quality service it needs than if it relies on the vendor; there are more levers w ith which to apply pressure. However, if a library has no parent organization C C —and th a t’s different from not w anting to use it—then the library seriously contem plating auto­ m ation must look outside. Service bureaus abound in the Yellow Pages. American Libraries classified ads list several consultants. If funds are very tight, th e n a d e p e n d a b le v o lu n te e r m ay be a n o th e r source. Staff can be hired or trained to take on cer­ tain jobs, but there needs to be an expert source of DP knowledge to whom they can turn locally; for example, the University of Florida, which has set up a sizeable DP staff w ithin the library, turns to both the North Eastern Regional D ata C enter, on w hose co m puter the N o rth w estern U niversity Technical Inform ation System (NOTIS) runs, and Northwestern University, for assistance. Some ven­ dors m aintain the software remotely as well as pre­ pare no-cost enhancements (Carlyle and Biblio- Techniques). Merely having a parent organization CC doesn’t settle the issue. Many factors are taken into consid­ eration, such as past library use, parent organiza­ 2Lyndon S. Holmes, “System Acquisitions and Vendor Expectations,” Library Hi Tech News 1 (June 1984): 112. Holmes is vice president of C L Systems, Inc. tion policies concerning CC availability to subunits such as the library, who pays for w hat, and the de­ gree to which library plans have the support of the parent organization’s top m anagem ent. Assuming th a t the CC is available, the real decision-making begins. The library wants to be assured th at the CC will give its production system—which must run de­ pendably about 19 hours per day, 365 days per year—top priority. It has to have confidence that the com puting center, as a fellow service unit, will treasure the library’s lifeblood. Why? Because the library cannot afford to have backup m anual sys­ tems; it must go fully w ith the autom ated system and not look backwards. The issues are life and death for the library, yet it often fears th at to the CC it is just another customer. Trust and confi­ dence in the parent organization CC by the library is a prim ary issue. Two campus service units Developing trust and confidence between the li­ brary and the campus com puting center m ay take some special effort. Differences in reporting struc­ ture, prim ary client group, and resource alloca­ tions may require frank discussions. Often the CC and library report to two different vice presidential levels, w ith the CC under an ad- m inistrative/finance VP and the library under the provost. As the keeper of a glamorous technology, the CC has often been much more successful in ob­ ta in in g re so u rc e s — b u d g e ts , p e rso n n e l slots, space—than has the library. Since m ost cam pus co m p u tin g began in the 1970’s w ith the autom ation of adm inistrative ser­ vices, e.g., payroll and student registration, the CC often fell into an attitude th at academic users, including libraries, were less im portant. Berry as­ serts th at in those days, DP personnel would “tell the user th at he not only did not know w hat he w anted, but was also incapable of understanding anything th at would run on a com puter.”2 3 The dif­ ferent ways th at the two service units—CC and library—defined “service” often epitomized those differences, w ith the com puter center concentrat­ ing on serving adm inistrative user units and letting end-user academics find their own way. In con­ trast, the libraries’ prim ary user group are faculty and students.4 It is for these kinds of reasons th at so m any campus library inform ation systems have been set up w ith little or no participation from their CC. W hile understandable, the waste is im- 3Dennis Berry, “C om puter Services: Is This a C ontradiction in Term s?” C ause/E ffect 6 (Mav 1983):5. 4For a discussion of which organization, com­ puting center or library, has the values to direct growing campus m icrocom puter use, see Alan E. Guskin, C arla J. Stoffle, and B arbara E. B aruth, “Library F uture Shock: The Microcomputer Revo­ lution and the New Role of the L ib rary ,” College & Research Libraries 45 (May 1984): 177-83. 163 m ense. Those D P staff have expertise th a t is d i­ rectly relev an t to th e lib ra ry in systems p la n n in g , im p lem en tin g , an d operations. If m an ag e d effec­ tively, these tw o cam pus service units can collabo­ ra te to provide superior systems to end-users an d to lib ra ry staff. T h a t co llab o rativ e philosophy w as th e basis for th e U niversity of C in c in n a ti deciding several years ago th a t its C o m p u tin g C e n te r (U C C C ) w o u ld as­ sist th e lib raries in all phases of lib ra ry a u to m a tio n , in clu d in g facilities m a n a g e m e n t. T w o m a jo r a d ­ v a n ta g e s a r e e s ta b lis h in g c o s t-e ffe c tiv e h a r d - w are/so ftw a re /te le c o m m u n ic a tio n configurations an d utilizin g expertise. B oth are c o n trib u tin g to th e sm ooth in stalla tio n of th e B iblio-T echniques L i­ b ra ry In fo rm a tio n System in N ovem ber 1985. As of e a rly 1985, a c o n fig u ra tio n c o m b in in g stan d -a lo n e (for lib ra ry use only) a n d sh ared (used c a m p u s w ide) resources is in p lace. S h are d r e ­ sources used for BLIS include on e-h alf of a 12-m eg IBM 3033N m a in fra m e located in th e U C C C co m ­ p u te r room , 3350 disk devices, a line p rin te r a n d a C O M T E N 3 6 9 0 c o m m u n i c a t i o n s c o n t r o l l e r . S tan d -alo n e e q u ip m e n t ac q u ired by th e libraries includes ta p e drives a n d re n te d ALA p rin t tra in . T h e co m b in atio n of sh ared an d stan d -a lo n e d e ­ vices pro v ed especially cost effective in p ro v id in g access to BLIS by a m axim um n u m b e r of UC users. In ad d itio n to th e 132 d ed ica te d term in als in li­ b raries, offices a n d d e p a rtm e n ts can access BLIS th ro u g h th e cam pus teleco m m u n icatio n s n etw o rk co n tro lled by th e C O M T E N . Sixteen d ia l-u p ports also p ro v id e access fro m p erso n a l c o m p u te rs in hom es an d offices. These th re e m ethods w ill p ro ­ vide access to a fa r g re a te r n u m b e r of end-users th a n th e lib ra rie s’ b u d g et alone could su p p o rt. T h e second m ajo r a d v a n ta g e of w o rk in g w ith th e U C C C has been expertise. T h e D ire cto r, L i­ b ra ry Systems D ev elo p m en t, has been on lo an to th e libraries for fo u r years. A senior systems engi­ neer a n d senior p ro g ra m m e r have been w ith th e p ro ject for 2 years an d are now experts in several m ysteries, in clu d in g O C L C records. T h eir w ork in stallin g N o rth w e ste rn U niversity’s T echnical I n ­ fo rm a tio n System (NOTIS) has en ab led staff to ed it a n d d e-d u p over 600,000 O C L C b ib lio g rap h ic reco rd s in 14 m o n th s. T e c h n ic a l specialists a n d o th e r U C C C staff h av e p a rtic ip a te d as needed. Us­ ing existing U C C C staff expertise saved th e p a re n t o rg a n iz atio n from h irin g an d su p p o rtin g d u p lic a ­ tive skills. Facilities management agreement I believe th a t th e key to effective m a n a g e m e n t of a cam pus C C by th e lib ra ry for th e pu rp o se of r u n ­ n in g a d ep e n d ab le, q u ality in fo rm atio n system is h av in g com m on sh ared expectations. W h e n expec­ tatio n s a b o u t services, tim efram es, p ro b lem d etec­ tio n , problem -solving an d costs are sh ared m u tu ­ a lly , th e n all p a r tie s h a v e a c o m m o n p o in t of reference. In those in ev itab le tim es of difficulty, e. g ., w h en th e system is d o w n for re p e a te d or extended periods of tim e , those sh ared expectations help resolve th e division of la b o r needed to solve problem s. T hey help red u ce finger-pointing. T h e U niversity of C in c in n a ti lib raries a n d co m ­ p u tin g c e n te r (U C C C ) h a v e w r itte n a facilities m a n a g e m e n t ag reem en t th a t sets fo rth sh ared ex­ p ectatio n s a b o u t o p e ra tin g BLIS. T h a t ag reem en t The key to system quality is having common shared xpectations. is a u g m en ted by a sep a rate d o cu m en t concerning p ay m en ts to U C C C for use of c e n tral site e q u ip ­ m en t a n d softw are. T h e facilities m a n a g e m e n t ag reem en t w as d e­ veloped iterativ ely an d jointly. T h e process began w ith a c a n d id discussion of concerns on b o th sides u n d e r th e gu id an ce of th e D ea n an d U niversity L i­ b ra ria n a n d Vice Provost for C o m p u tin g an d I n ­ fo rm a tio n T echnology. T h e Systems M an ag em en t C ouncil, com prised of th e heads of all five UC li­ b ra ry jurisdictions, an d key su b o rd in ates laid out th e ir expectations. D ra fts w ere review ed by lib ra ry a n d C C m an ag ers an d by th e Systems M an ag e­ m en t C ouncil. T h e earlier an d sim ilar ag reem en t d ra w n up by Johns H opkins w as very helpful. Points of interest from th e U niversity of C in cin ­ n a ti facilities m a n a g e m e n t ag reem en t for o p e ra t­ ing BLIS are: I. Goals: A. P rovide facu lty , students an d staff online ac­ cess to p u b lic catalo g an d circu latio n /reserv e fu n c ­ tions 7 days a w eek, 19 hours p er day. B. B ased on ex p erien ce, keep fine tu n in g th e h a rd w a re /s o ftw a re co n fig u ratio n so as to m eet ser­ vice level objectives. C. Set up p ro b lem -d e te ctio n m ethods th a t spell o u t steps to be tak en by th e libraries an d by U C C C , th u s using everyone’s tim e m ost effectively. I I . D e fin itio n s , e .g ., w h a t lib r a r y u n its a re served. III. Considerations about: A. C e n tra l site, e .g ., assure e q u ip m e n t security. B. R e la tio n to B ib lio -T e ch n iq u es, I n c ., e .g ., U C C C shall advise B iblio-T echniques of so ftw are failures th a t are beyond th e scope of responsibility of U C C C . F u rth e rm o re , U C C C w ill co o rd in a te so ftw are su p p o rt for IBM p ro d u cts (OS/VS1 an d re la te d p roducts). It is expressly u n d erstood th a t B iblio-T echniques assumes p rim a ry responsibility for th e resolution of s o ftw are-re lated problem s for th e a p p licatio n an d th e VM/VS1 p o rtio n of th e sys­ tem . e 164 C. U C C C service level objectives, e.g.: 1. O p e ra te BLIS from 7 a .m . to 2 a .m . seven days p er week. 2. M ain tain a telephone “ho t lin e” d u rin g the above hours. 3. In th e event of h a rd w a re or softw are failure, take all reasonable an d expedient m easures to co r­ rect th e failure, notifying th e libraries an d Biblio- T echniques as quickly as possible. 4. C o o rd in ate corrections of h a rd w a re failures w ith th e a p p ro p ria te service vendor, p er existing service contracts. 5. Set up tran sactio n logging. 6. B ackup BLIS an d th e d a ta base regularly, b u t not to exceed once every 24 hours. 7. M onitor response tim e of online transactions, on both a ro u tin e “snapshot” basis via a statistical re p o rt an d for special diagnostic purposes. 8. Schedule p rin tin g of reports an d o th er docu­ m ents w ith th e libraries. 9. E nsure th e integrity of p rogram s an d d a ta m a in ta in e d by BLIS. 10. C onfigure th e 3033N such th a t: a t least 50 % of th e 3033N is allocated for BLIS; an d to p p rio rity is given to processing BLIS transactions. 11. W ith th e libraries, p la n for teleco m m u n ica­ tions grow th to accom m odate ad d itio n al h a rd w a re an d d ial-u p term inals. D . L ib raries’ responsibilities 1. T ra in staff in all un its for all shifts to use m u tu a lly -d e v e lo p e d p ro b le m d e te c tio n /s o lv in g procedures. 2. M ake every effort to solve problem s locally, using m utually-developed procedures. O nly w hen th ey are u n ab le to proceed fu rth e r in th e proce­ dures w ill lib raries’ staff co n tact U C C C . 3. W h en re p o rtin g problem s, staff w ill state how fa r they got in th e procedures an d re p o rt all d a ta re q u ired by th e procedures in o rd er to assist U C C C in analyzing an d solving th e problem . E. M anagem ent 1. Id e n tify key U C C C an d libraries persons re ­ sponsible for BLIS perform ance. 2. U C C C an d th e libraries m eet as freq u en tly as needed b u t a t least q u a rte rly to review system p e r­ form ance. Summary Regardless of th e source of D P expertise, success­ ful c o lla b o ra tio n b e tw e e n th e lib ra ry a n d th a t source is based on th e lib ra ry ’s w ill to m anage. C lear shared expectations a b o u t services, finances, an d tim efram es com bined w ith m axim um lever­ age should en h an ce th e lib ra ry ’s ability to offer end-users d e p e n d a b le , h ig h q u a lity system ser­ vices. A w ritte n facilities m an ag e m e n t agreem ent is one tool th a t helps set up shared expectations b e­ tw een tw o service units, th e lib ra ry an d th e com ­ p u tin g center. W h en those tw o service units are p a r t of th e sam e p a re n t o rg an izatio n , e .g ., higher ed u catio n , th e lib ra ry should consider m anage- 165 m en t strategies a n d tools th a t w ill best assure sys­ tem access an d p erfo rm a n c e to su p p o rt faculty, students an d researchers. T h e academ ic lib ra ry has a vested in terest in m a n a g in g its cam pus c o m p u t­ in g c e n te r effectively. ■ ■ Building a strategy for academic library exhibits By Lucy S. Caswell C urator, L ib ra ry fo r C o m m u n ic a tio n an d G raphic Art. O hio State U niversity Education as a primary goal fo r library exhibition programs. D e s p i t e th e fact th a t exhibitions are la b o r in te n ­ sive a n d thus m ake heavy d em an d s on th e m ost ex­ pensive re so u rce, m ost a c a d e m ic lib ra rie s h av e som e ty p e of exhibit facility, ra n g in g from b u lletin b o ard s used to display th e jackets of new books to custom -designed cases for ra re books. W h ile few w o u ld d isp u te th e effectiveness of b u lletin boards in boosting c irc u la tio n , such efforts are no t u n d e r co n sid eratio n here. T h e term exhibition as used in this article refers to displays of a significant n u m b e r of item s org an ized in such a w a y as to convey in fo r­ m a tio n to th e view er. T h e purposes of academ ic li­ b ra ry exhibitions w ill be briefly review ed p rio r to suggesting uses of exhibitions as lib ra ry in stru ctio n tools a n d exam ining th e c o n trib u tio n of such exhi­ b itio n p ro g ra m s to th e acad em ic co m m u n ity . T h e tra d itio n a l reasons for exhibits in college an d university lib raries a re often tak en for g ra n te d . S a n d ra P ow ers suggests th a t th e m ajo r goals of an exhibition p ro g ra m are ed u c atio n , increased use a n d p u b lic re la tio n s .1 W ith in th e acad em ic lib ra ry S a n d r a Pow ers, “W h y E xhibit? T h e Risks V er­ sus th e B e n e f its ,” A m e r ic a n A r c h iv is t 41 (July 1978) :302. ce rtain ly ed u c atio n should be th e p rim a ry aim . Ex­ hibits m ay be re la te d d irectly to classroom projects (a d a g u e rreo ty p e exhibit for th e history of p h o to g ­ ra p h y class) or to c o n tin u in g e d u c atio n p rogram s (m edieval m an u scrip ts displayed for a conference on p o p u la r religious c u ltu re in th e M iddle Ages). In stru c tio n a l goals m ay also be m o re general (to il­ lu stra te th e dev elo p m en t of film ad vertising from 1910 to 1930), b u t exhibit p lan n ers should alw ays be able to re la te th e ir show to th e academ ic e n te r­ prise. As m en tio n ed previously, increasing circu latio n by calling a tte n tio n to exhibited m aterials has long been accep ted as an a p p ro p ria te ra tio n a le for exhi­ bition. W ith in an ongoing exhibit p ro g ra m in a col­ lege or university lib ra ry , it is im p o rta n t to re m e m ­ b e r t h a t e s s e n tia lly th e e n t i r e u n d e r g r a d u a t e p o p u la tio n changes over a fo u r y ear p eriod. Be­ cause of this, re p e a t exhibitions m ay be a p p ro p ri­ ate, especially displays of special m aterials w hich m ay n o t com e to th e a tte n tio n of p o te n tia l users in o th e r w ays. U n d erg rad u a te s are som etim es hesi­ ta n t to e n te r w h a t ap p ears to be an in n er sanctum called “special collections” or “ra re books ro o m .” If selected m aterials from these areas are exhibited in