Agenda for HOLLIS Liaisons Meeting #85
April 14, 1993
Lamont Forum Room, 9:30 - 11:00 a.m.
1. Announcements: Tracey Robinson
Notes from the March Meeting
LIBINFO discussion list revisited. It has been just over a year since OIS founded the LIBINFO electronic discussion list. "Traffic" on this list includes official meeting minutes, announcements from library units and OIS, and reposted messages from other discussion lists such as PACS-L, USMARC-L, and BI-L. Recently, some subscribers have expressed concerns that the traffic on this list is growing heavy with "noise" (non-relevant postings). Tracey asked liaisons whether this "noise" warranted a change in LIBINFO's scope or configuration. One possibility would be to split LIBINFO into two different listservs; a list restricted to Harvard subscribers where official minutes and other administrative information is posted; and a list with open subscription that distributes reposted messages and other "FYI" information of less interest to some subscribers.
[Editor: Recently, LIBINFO subscribers were asked their opinion on this issue. An overwhelming majority reported being happy with the existing scope even though not all postings were relevant. However, most subscribers had no objection to splitting LIBINFO into separate administrative and FYI lists.]
OIS increasingly relies on LIBINFO for the distribution of minutes and other types of announcements and wants to encourage as many Harvard staff subscribers as possible to stay "tuned" to LIBINFO. A decision about the future of LIBINFO will be made this spring. Staff with suggestions should contact Julie Wetherill in OIS.
Networked PC access to HULPR. During the presentation in February on connecting your PC to HULPR over the network, Tracey presented the billing options for the remainder of fiscal 1993. Beginning in March, OIS will bill for HULPR access via networked PCs on either a transaction basis or a fixed monthly rate basis. March billings occur at the end of the month and OIS will survey users before then for their billing preference. Units that want to access HULPR via a network connection should contact Ciata Victor in OIS.
Recon cleanup activities for Widener records. OIS has developed "cleanup" processes to compare Widener Recon records against the OW bibliographic file and the HU unlinked item record file. OW records that match with incoming Widener Recon records are deleted and physically removed from OW -- OW is beginning to shrink and is now below the one million record mark. Regular monthly matching will continue to shrink OW over the life of the Recon project. Incoming Widener Recon records are also compared with existing unlinked item records in HU - - matches lead to the item record being linked to the Recon bibliographic record. Hopefully, this process will reduce the number of unlinked item records in the file.
Recon update. Karen Carlson Young, Recon Project Manager, provided a summary of recent Recon Project activity.
Through early March, OCLC has searched 111,000 titles, including 39,000 in February. To date, 33,000 records have been added to HU. The discrepancy between the number searched and added reflects a lag time between processing and loading. As OCLC productivity increases and the loading of records into HU occurs more frequently, this discrepancy will diminish.
The Recon projects for the Law School and Kummel Library (Geology) are scheduled to begin in May. Gutman Library's project has been moved up from May to April because OCLC has sufficient staff to begin earlier than planned.
Karen has started visiting libraries to plan serial recon activities. She will be gathering information from individual units in order to prepare schedule for library participation. Staff with questions about Recon activities are encouraged to contact Karen Carlson Young at the Recon Project Office (496-4011).
The Harvard-Yale-MIT cooperative serial project. Allen Bourque (Cabot Library) summarized for liaisons a pilot project between Harvard, Yale University, and MIT to share serial subscription information and to provide document delivery services for journals in the sciences.
Background of the agreement-- During the summer of 1992, representatives from the three institutions met to discuss a mechanism for sharing information about serial subscriptions and coordinating serial cancellations for journals in the sciences. All three anticipated that future cutbacks would be required in their serials budgets and were concerned about the impact these cuts would have on the availability of journals to their students and researchers. After several months of negotiation, much of which occurred via an electronic listserv discussion forum, the institutions agreed to develop a trial document delivery service to supplement science journal cancellation decisions.
Description of the service-- The document delivery service envisioned by the institutions supplies free copies of science journal articles with a three day turn-around time for receipt. Participants fill out a special document request form and send it via fax to the institution holding the article; delivery of articles is also done by fax. In the future, requests will be processed at an Ariel workstation when the technology becomes available at all three institutions (MIT has it; Harvard and Yale are scheduled to join).
Allen explained that at Harvard, this new service is currently a HCL initiative which incorporates HCL and FAS science libraries. "We look forward to the time when ILL departments in other parts of the University Library system can make requests for science journal articles using this mechanism," Allen added. There are three local service "nodes": Wolbach, Gordon McKay, and Cabot Library. Wolbach and Gordon McKay handle their own ILL requests; Cabot handles its own requests and those for all other Harvard science libraries. In order to facilitate this service, OIS has scripted access to the catalogs of Yale and MIT from the HOLLIS staff menu. Staff can access Barton (MIT) or Orbis (Yale) by typing "telnet barton" or "telnet orbis" from the staff menu. Eventually, both catalogs will display as choices on the menu. Cabot has a staff terminal in their reference area and staff will help patrons access Barton or Orbis on request.
A one-year trial project. This new document delivery service began operation on 16 February and participants expect to let it run for a year and then reconvene the planning group to review the project. Allen noted that another round of serials cancellations would also trigger a review. So far, "traffic" has been lighter than expected. Participants are keeping careful track of requests and turn-around time to insure delivery is reliable and that there are no violations of copyright laws. If you have questions about this service, contact Allen Bourque or Marcia Chapin (495-5354) at Cabot Science Library.
Presentation: Spine labels from HOLLIS. The capability to produce spine labels from HOLLIS MARC records will soon be available to Harvard library staff. OIS has a mainframe-to-microcomputer spine label function that is scheduled for release in April. This presentation was intended to acquaint liaisons with this new function so they can help their units decide whether it is a function they will want to implement. It included a demonstration of the microcomputer portion of the function and a discussion of the design decisions behind spine label formatting. What follows is a summary of the presentation and liaisons' comments. There is also more information about the implementation plans. A full description of the online portion of the spine label function appeared in the March 1993 HOLLIS Newsletter.
Why choose Vernon? During the presentation, Jill Thomas from the Botany Libraries questioned whether the Vernon program was the best choice of microcomputer software for this new process. Since few members of the original working group were present, a complete answer to Jill's question was not possible at this meeting. However, Derek Katz has since put together a short history of the design process for this new feature, including the reasons why the Vernon program was the Group's first choice. This excellent short history -- which isn't that short -- is appended to this newsletter.
Also note that the list of Working Group members published in the March HOLLIS Newsletter was incorrect. The "real" participants were: Kathy Anderson, Lisa Biblo, Karen Carlson Young, Derek Katz, Daniel Lobdell, Beata Panagopoulos, Jon Rothman, Julian Stamm, and Jill Thomas.
The PC process. The procedures for saving and sending MARC records from HOLLIS to an electronic mail account were described in the March 1993 HOLLIS Newsletter. What follows is a summary of what happens to these records on your PC. There are no references to specific commands used in this process; a full set of commands and instructions will come with the HOLLIS Spine Label Kit.
First, the operator needs to "pre-process" the downloaded file of HOLLIS MARC records into a format acceptable to the Vernon Spine Label Program. That is the job of the software that comes with the Spine Label Kit. The "pre- processed" records can then be formatted and printed by the Vernon Spine Label Program.
Example 1 below illustrates of a HOLLIS MARC record (AAA2644, Widener location) that has been sent from HOLLIS to an electronic mail account.
Example 2 illustrates the same HOLLIS MARC record after being "pre-processed" by the spine label program into an OCLC MARC record.
The highlighted line in Example 3 illustrates the MARC record once the operator has loaded it into the Vernon program. Record AAA2466 happens to be the first of 11 spine label records in a single set.
The Vernon software offers many options for editing spine label records, creating new labels, or deleting existing labels. Most of these options are available from the "Main Menu" display illustrated in Example 3 above.
When is the #@!& function going to become available? The short answer is "hopefully in late April." The long answer is that this new function was originally scheduled to debut in March, but some unexpected technical problems appeared that OIS is still working out. OIS staff are working with the producers of the Vernon program to iron out these problems and although we are optimistic that everything will be ready to go for late April, the debut may slip a few weeks into May.
Do I need authorization? Each unit interested in using this new function must request a Spine Label Kit and online security authorization from OIS. A request form was appended to the March HOLLIS Newsletter and is also appended to this issue. Ignore this if you have already filled out and returned a request form.
How do I get the HOLLIS Spine Label Kit? Fill out the request form appended to this newsletter. If you have already sent in your request form, ignore these instructions. How do I get the Vernon program? The HOLLIS Spine Label Kit will provide instructions on how to obtain the Vernon program.
What do I need to produce spine labels from HOLLIS? You need the following:
How much will this cost? There will be a one cent per record charge for every MARC record stored and sent from HULPR. Units also must purchase the Vernon Spine Label Program, which retails for approximately $300.00 per copy.
How do I know that I can download files to my PC? To use the spine label function, you need an electronic mail account to which HULPR can send the file of MARC records. This e-mail account might be located on a large multi-user system, such as a HUBS account on OIT's IBM mainframe, or a HUSC account on the Science Center VAX. Or, the e-mail account might be on your local area network (DaVinci Mail and Microsoft Mail are examples).
You also need an IBM or compatible PC that has dial-up or network access to your electronic mail account. A dial- up connection can be a PC with a modem attached to a telephone line, a PC connected to an ISDN telephone line with a data option, or a PC connected through a datajack to a terminal multiplexer (MUX) . A network connection involves a PC with a network card connected via a local area network to Harvard's High Speed Data Network (HSDN). In all of these cases, there will be some type of communications software that enables you to transfer the file of MARC records from the e-mail account down to your PC.
How can you tell if any of the above circumstances are true for you? If you access your electronic mail account from a PC now, then you probably have the means to download the file to your PC. If you access e-mail from a HOLLIS staff terminal (as with HUBS or HUSC) then ask your LAN Administrator, Automation Coordinator, or PC troubleshooter about whether any of the PCs in your unit can provide dial-up or network access.
How do I download files? Contact your LAN Administrator if you are unsure how to download records to your PC. If you have a HUBS, HUSC, or other multi-user e-mail account, contact the administrator of that service for more information. Or, contact the Network Assistance Line at 495-9388.
Taking the Spine Label Function for a test drive. Before the new Spine Label Function becomes generally available, OIS will establish one or more "beta" test sites where the process will be heavily tested and refined by actual users. Lamont Library and the Law Library are tentatively scheduled to be beta test sites in April. Near the end of beta testing, other interested libraries will have the opportunity to try this new function -- taking it for a test drive -- before becoming registered users. OIS will announce the details about this "test drive" event when we get closer to the implementation date.
Yes, a gopher can be a burrowing rodent from the family Geomyidae but it can also be a software tool that allows you to browse the abundant information resources available on the Internet. A gopher arranges resources in a sequence of menus; by choosing a menu option, you instruct the gopher to connect with the remote resource provider and retrieve data for display. Gopher uses a client-server architecture, meaning that on your end there is gopher client software that can "talk" to a gopher server at the other end -- the site offering the resources. These resources can be library catalogs, databases, electronic "phone" books, bulletin board systems, and other information stored on remote computer systems. By using a gopher, you do not need to know how to communicate with the remote system holding the resource and you do not need to know the address of this system -- gopher negotiates all this for you. Each gopher site knows about other gopher sites, so once you connect with a gopher server, there should be a menu selection allowing you to travel to other gopher servers. Most of these gopher sites are maintained by academic institutions and will contain information resources of interest to those sites. Anyone who uses a gopher will see the same kind of menu setup and use the same simple commands to move around.
The most efficient and effective way to access gopher systems is to install gopher client software on your computer (assuming your computer has access to the Internet). There are free gopher clients for just about any type of computer: IBM/PC, Macintosh, UNIX, X Windows, VAX/VMS, VM/CMS, and probably others. See the Gopher sites and software section below for a location where you can FTP gopher software.
If for some reason you do not have a local gopher client installed, you can telnet to one of several public gopher clients (that is, telnet to a computer system on which gopher client software is running that is available to everyone). From this public client, you can dig your gopher holes with glee! (See the Gopher sites and software section below for a partial list of public gopher sites.) However, where you are telnetting from is very important and will determine how well gopher performs for you! Why? Gopher prefers that your local computer emulate a VT100 terminal when connecting. Most people who try accessing a public gopher will be coming from a computer device that can provide VT100 emulation and therefore the public gopher site should work for them. On the other hand, gophers do not prefer the IBM TN3270 emulation. At Harvard, anyone who telnets from HUBS or from the menu on a staff HOLLIS terminal is actually using the IBM mainframe at 1730 Cambridge Street as the computer device that connects to a public gopher. Unfortunately, the IBM mainframe offers TN3270 emulation and does a poor job emulating a VT100 terminal. What this means is that you can connect to a gopher from an IBM terminal but don't expect it to work well (or at all).
Luckily, there is (or rather, there was) at least one IBM mainframe site, University of Minnesota (pubinfo.ais.umn.edu or 184.108.40.206), that offered a publicly accessible gopher client that would accept TN3270 emulation from IBM mainframe users (like those telnetting from HUBS or the staff HOLLIS terminal). And, apparently many Harvard library staff members took advantage of this opportunity and used this gopher to connect to other gopher sites on the Internet. In the end, the public "traffic" on U. Minnesota's mainframe was so heavy that they were forced to close their mainframe gopher to outside access.
U. Minnesota's 3270 gopher appears to be the only one of its kind that was available to the public -- OIS and OIT are not aware of an alternate site providing the same service. Staff who used the 3270 gopher from their HOLLIS terminals will have to find an alternative method for connecting to a gopher. A microcomputer attached via a local area network to the Internet is one option. Another would be to obtain an e-mail account on a non-IBM multiuser system, such as the VAX at the Science Center (a HUSC account). It is unlikely that any time soon another 3270 site with public access will arise -- the gopher technology works poorly in the IBM mainframe environment. Those adventurous enough can still try to access a gopher from their IBM terminal, but performance will be very poor. OIS will keep staff posted in the event another alternative is identified. Contact Derek Katz or Julie Wetherill if you have questions.
The most elegant way to access a gopher is to use gopher client software appropriate for your local workstation. This software is freely available via FTP from many gopher sites, including boombox.micro.umn.edu (220.127.116.11), in the directory /pub/gopher. Your workstation must have an HSDN connection and an IP address to use client software, but it is well worth the effort, since using the appropriate client provides you with a familiar gopher interface. Ask your LAN Administrator or other technical support personnel for further information and assistance.
If for some reason you do not have access to a local gopher client, or you want to try out gopher before deciding to install a client on your workstation, here are some public gopher sites. You must be able to telnet to access these sites. Login as "gopher" unless noted otherwise.
If you decide to try one of these sites, please make your attempt during off-peak hours if at all possible (after 5pm) - - this will help reduce the traffic load on the sites and probably improve your response time as well.
Notes and Reminders
Jeff found a bug. This problem occurs when you attempt to update or add an 020 field to an HU record that contains just a subfield z and no subfield a. HOLLIS is expecting to find a valid subfield a ISBN before allowing you to add an 11-digit number in subfield z. This bug is annoying but does not present a critical problem. Until OIS can plan a solution, staff will need to watch out for the occurrence of 11-digit ISBNs (a relatively rare event), especially in BF records that they try to migrate or merge into the HU database. If you need to add an 11-digit ISBN to an existing 020 already containing a subfield a you should have no problem. Until the bug is fixed, if you need to add a 11-digit number and the 020 does not or will not have a subfield a, you will have to add it to a note field (5xx). If you have questions, contact Robin Wendler in OIS.
Fiscal 1994 RLIN rates. RLG recently released a memo summarizing their pricing plans for fiscal year 1994 (September 1993-August 1994). Anyone who needs a copy and has not received one should contact Robin Wendler in OIS.
Linking unlinked item records begins. Using the HU database call number indexes, OIS has developed a process to match up unlinked item records with bibliographic records in HU -- including newly added Recon records. OIS recently made a first pass through the file of unlinked item records using this new process and was able to link up 10,871 item records out of a total of 68,428 unlinked item records. This process will work for item records belonging to any library -- not just Widener, but it will not work for unlinked items that carry a "charged" status.
However, after OIS makes a few refinements, the process will run on a roughly monthly basis and unlinked items that were missed the first time around may be linked at a later time. OIS does not expect that this process will link all unlinked item records, but it will go a long way in reducing the total number of them over time. Contact Charles Husbands in OIS if you have questions.
Change to MLOX command. OIS has made a change to the MLOX command in HULPR that saves 948 field(s) (RECON project information) from the source record and deposits them on the target record.
This small change will eliminate the need for operators to rekey 948 fields that do not get "mlox'd" over to the target record during the process. The 948 field, introduced at the beginning of RECON (late 1992), contains notes about the source of RECON information. A description is available in the December 1992 issue of the HOLLIS Newsletter.
This change is being made in HULPR now since it seems better not to wait two weeks for the next liaisons meeting for it to be officially announced,and so that staff do not have to rekey the 948 fields as part of the MLOX process. This change should have no other affect on the MLOX process. If you have questions contact MacKenzie Smith in OIS (495-3724).
HOLLIS Phone Chain errata. The telephone number for Roberta Schwartz at the MCZ Library is incorrect on the 03/03/93 issue of the HOLLIS Phone Chain. It should be 495-3946. Please note this correction on your copy of the Phone Chain.
Response time tests. Response time tests are scheduled for the first Tuesday of each month. Upcoming tests are scheduled for: 6 April, 4 May, 1 June, and 6 July. Volunteers should mark their calendars appropriately.
As the frequency of cataloging notification and other, similar reports increases we have heard from some units that they would prefer to get them "electronically" (i.e. via email, or via FTP, but not on paper). How many of the units that receive these reports already have an e-mail account? How many have access to the HSDN (Harvard High Speed Data Network) for using FTP? How many have a PC, Macintosh, or other device to which they can download electronic files? If there is sufficient interest in getting these sorts of reports electronically, and if most libraries are prepared to get them this way, then OIS could investigate this option for the future. Please come to the April liaisons prepared to give your library's preferences on this matter!
Tape loading of HOLLIS records to OCLC using FTP. Libraries can send copies of their HOLLIS records to their bibliographic utility (OCLC or RLIN) by using the TAPE command. This has the effect of adding their holding code to the utility copy of the record. OIS batches these HOLLIS records and sends them to the utilities on a tape once a month. Starting next month OIS will be sending TAPEed records to OCLC using the Internet's FTP (file transfer protocol) facility instead of by tape. As a result of this change we can send HOLLIS records to OCLC every week if we like. Do libraries who use the TAPE command want their records loaded into OCLC on a weekly or on a monthly basis? If you have a preference let us know! We'll be asking for your opinion at the April liaisons meeting.
Recon Quarterly Report. Included in this issue of the Newsletter is the Retrospective Conversion Project's 1st Quarterly Report for 1993. In addition to its publication here, the report was sent to all Recon Liaisons, members of the ULC and to the members of the Recon Oversight Committees.
Within the next few weeks we will set up a demonstration DeskJet in OIS which people will be able to use for evaluation purposes. Please watch LIBINFO and HOLLIS EMS for further information. To make it easier for units considering replacement of their Canons (or just adding additional printers), OIS is investigating price and availability and will put together a document describing how best to order this equipment.
As many of you are aware, the IBM 3151 terminals have always had problems with printing certain letter-diacritic combinations. In conjunction with our investigations of alternative printers, OIS has also tested printing from the IBM 3151. These tests show that the problems are inherent in the design of the IBM 3151's cartridge. The ALA Cartridge Users Guide which accompanies the 3151 cartridge includes a table (Table 1-1 on page 1-2) containing what IBM considers to be legitimate letter-diacritic combinations. Although other combinations can be entered in HOLLIS, these do not print correctly. The HP DeskJet 500 described above can be used with the IBM 3151. The results will be no better than -- but no worse than -- with the Canon BJ80. OIS will, of course, bring this problem to the attention of IBM and will keep you posted if there is any change. If you have any questions, please contact Ciata Victor or Heather Reid in OIS.
New guidelines for calling in network problems. In order to provide better and more timely service in the event of HOLLIS network and equipment problems, OIS is making a change to the way network problem calls are handled. If you determine that you need to call OIS about a problem, telephone 495-3724 as in the past but do not ask for Ciata Victor. Instead, please identify your call as a network or equipment problem and an OIS staff member will either direct your call to the right person or take information from you about the problem.
How do I determine if I should be calling OIS? Please refer to the new OIS contacts list that is appended to this issue. If you have any questions, contact Heather Reid in OIS.
When to turn off your PC? A recent posting to the LIBINFO discussion list suggested that there would be a "big environmental benefit" (mostly energy savings) if people would turn off their microcomputers when not in use. While this may be true, frequent shutdown of micros can lead to damage of their components. Derek Katz from OIS posted a response to LIBINFO, paraphrased below, that concerns this issue.
There has been some debate following the message about saving energy by powering down microcomputers. The upshot of all this seems to be:
Units should consider their own needs, but also understand that most microcomputers consume relatively small amounts of electricity such that turning them off frequently saves little. And, powering a computer up and down frequently (several times a day) does more harm than good. Perhaps a policy of turning micros off at the end of the day would be a good compromise. As for HOLLIS IBM terminals, there is no need to turn them off at night, although screens should be dimmed if the screen saver option is not activated. Contact Ciata Victor in OIS if you have further questions.
Last spring, some of the parties interested in producing spine labels from HOLLIS formed a group to design a process for extracting call number information from HOLLIS, moving it on to a microcomputer, and printing spine labels from it. By June 1992, the nine person group (chaired by Jon Rothman of OIS) had come up with proposals for the new HOLLIS commands SPINE and SEND, an e-mail based mechanism (analogous to the HOLLIS SEND mechanism) for exporting records from HOLLIS, and a PC program for processing the records, and producing labels from them. These proposals were taken to HAAC, which approved the new HOLLIS commands and the export method, but suggested that we find an existing PC program that could produce spine labels, instead of writing our own.
We then queried other libraries, and searched journals and catalogs for such a program. Most of the users that responded used the spine label program from Vernon Library Supplies (VLPS), although a couple of libraries used Word Perfect macros or home-grown programs. All of the other commercial programs were disqualified, either because of four-digit price tags, or because they came as part of a larger system, and would only print records from their own cataloging or acquisitions modules. We demonstrated a couple of them anyway.
The choice seemed to come down to VLPS and the OCLC Micro Enhancer (ME). The group came down in favor of VLPS for the following reasons:
A PC program to convert HOLLIS records to OCLC MARC was written last July, while mainframe development was delayed by Recon project developments until this winter. Further delays have been caused by the need to alter both mainframe and PC programs to make them compatible with each other, and by problems importing larger sets of records into VLPS. The programmer who wrote VLPS is currently working with HOLLIS data to resolve the latter problem.
The output from this PC program should be usable with any spine label program that can import OCLC MARC records. The program could also be compiled for the Macintosh operating system, to produce records to imported into a Mac spine label program, should such a thing exist.
Please call Derek Katz (5-3724) if you have questions about the VLPS program, or about the PC program for converting HOLLIS records to OCLC MARC.