To the Editors of The Crimson:
As a graduate student in the Department of English and American Literature and Language, as well as a part-time programmer for OIT and for Harvard Computer Services at the Science Center, I read Susan Glasser's excellent article [9/14/88, page 13] on the new HOLLIS system with great interest. HOLLIS is terriffic; I will be using it all the time. I must take issue, however, with some of the boosterism reflected bvy remarks by library officials suggesting that the system is "absolutely state-of-the-art." Whatever the state of the art is, it is not HOLLIS. HOLLIS cannot do keyword searching; HOLLIS cannot search both the "Older Widener" and Union Catalogs at the same time; there is no provision in HOLLIS for saving selections for later printout (although one may line up for a terminal with a connected printer); HOLLIS has only primitive logic capabilities to narrow searches. Library officials have said that these features are on the way--but they're not here now.
A major problem with HOLLIS is that of access (a problem with all user-oriented computing at Harvard). HOLLIS is not presently accessible from campus networks (or is so far not documented as such). From a terminal in the Science Center, it is impossible for a student or faculty member to integrate catalog information into a text file or a mail message. In the last year, I received in my computer mailbox cataloging information from a librarian at the library of the University of Toronto; but I suspect it will be a long time before I can send a computer mail message to a librarian at Widener and receive a reply with cataloging information.
Glasser writes that Library officials confess "that there is no way to predict how many people will use the newly installed terminals or how the technology will affect the research habits of the community as a whole." Why not? Didn't the staff at OIT check with other universities that have been using this kind of new-fangled technology for the last ten years?
Furthermore, HOLLIS has a few problems with its user interface. HOLLIS uses grody IBM terminals which do not have keyboards organized in the manner of the commonly accepted industry standards. Anyone used to an IBM PC or an Apple Macintosh will be surprised to find that the key she expects to delete the previous character does not do so. And there is no on-line help explaining the keyboard (what do these extra keys DO!?). Also, HOLLIS does not take into account what is surely the most common task dial-in users will be performing with the system; to wit, dial-in users will be logging a transcript of their session to a file for later editing or printing. But HOLLIS, unfortunately, in its VT100 mode, moves the cursor for every display of cataloging information! (It could be scrolling the information instead.) What this means is that anyone seeking to print collected displays will have to edit the file first to get rid of the cursor-moving escape sequences. What gives? After years of modifying software OIT had to get from another university, can't the most trivial ease-of-use features be integrated into the system?
In short, what really bugs me is not necessarily the capabilities of the system, but the attitude of the developers that it is "state of the art." They tell us that they can put 150,000 entries on line per year. With 8,000,000 more volumes to go, that's about 50 years. Let's be real, OIT.
I eagerly await the computer cataloging system that the Harvard community needs. John Norman, GSAS