In 1990 Bath University Computing services won a contract from the Universities Funding Council's Information System Committee to host the recently accquired ISI databases. BIDS (Bath Information and Data Systems) was up and running by February 1991. The inital four databases (the 3 Citation Indecices and the Index of Scientific and Technical Proceedings) have been joined by eight more databases acquired through CHEST so that five years later there are now twelve databases in total to choose from as well as a gateway to Blackwell's Uncover.
BIDS is available through JANET to all UK HE institutions and the Funding Councils (a couple of Scandinavian universities also subscribe) and is completely free at the point of delivery as each institution pays a annual flat rate subscription for each database they wish to offer their users. It is this ability to use the databases without additional cost as well as the ease of connecting (a link to JANET and a machine capable of emulating a VT100 is all that is needed) that appears to be the most valued aspects of the service .
An ongoing programme of re-design and upgrade continues to improve the service and to overcome, in the words of Terry Morrow, "the result of hurried initial implementation". There is an active End-User Group and a mailbase list for BIDS users. As well as adding new databases and re-defining the interface, BIDS has also expanded its services. In 1993 BIDS added a document ordering facilty using BLDSC as the supplier. However this service is not heavily used, probably due to a number of factors, namely the invisibility of the option as it is listed at the bottom of the screen, the lack of promotion by some librarians fearful of losing ILL business, the unfamilarity of academics to the concept of paying for information and possibly the knowledge of some people that the Internet is not yet secure for the transfer of credit card data (although an account option is available).
It is rarely argued that BIDS is a useless service and that certainly is not going to be the thrust of this review. However there is a debate, often vigorous as evidenced by the interchange on LIS-LINK in the latter half of February, over the balance between the information provided and the ability to get at it. For example Prof. Charles Oppenheim of de Montford University wrote:
"frankly, I find the BIDS interface so unfriendly I do believe scrapping it and starting again from scratch is the best approach." 
The present interface is a command driven menu system, designed to be accessible from any networked machine with VT100 emulation - a lowest common denominator option based on the technology available and widespread at the beginning of the decade. Others, however, believe that the interface is not a great issue for users, for example Barbara Middleton, and many find the restrictions imposed by the database compilers as frustrating as any perceived inadequacies of the interface.
Those restrictions seem most apparent in the ISI databases which make up the bulk of BIDS traffic but are by no means limited to them. Chief grumbles are the lack of distinction between authors of the same name - there is no name rationalisation at all; inability to search on any but the first author of a cited work; inability to determine which cited works are common to a number of articles and the lack of adequate thesauri, particularly for scientific and technical databases. For the most part the remedy of these problems is beyond the control of BIDS. All that can be done is lobby the database compilers, but as the Chairman of the Bids Users Group pointed out in relation to FirstSearch, a small group of UK higher education institutions do not have much influence compared to US undergraduate universities. However the fact that the UK HE sector might only constitute a small section of a user group should not preclude its representatives being as noisy as they can when reporting problems.
Yet it is a commercial reality that certain problems cannot be solved even if they may leave Academia in a grey area. For example the issue raised recently on LIS-LINK by Prof. Tom Wilson of Sheffield's Department of Information Studies, is that of conducting bibliometric research using ISI databases. Prof. Wilson questioned the ban on bibliometric studies when research is allowed and in many LIS departments this research may well include bibliometrics. Bibliometric studies can be undetaken on the BIDS ISI databases but only after the licensee has made an extra payment to ISI. In this case ISI are purely protecting their commerical rights, which considering that they undertook a deal with CHEST to provide the databases as below their commercial value, they can hardly be expected to act otherwise.
Another large bone of contention which is certainly beyond the control of those at BIDS is the slowness of the service. Personally, I have been forced to cancel all BIDS training sessions which were to be held in the afternoon as the users were finding the length of time a search took unbearable. An added annoyance for them was the lack of knowledge as to whether the computer had hung, or the search was taking a long time. This was then compounded by the users re-pressing keys in the mistaken belief that the original strokes had not be registered, so that many found themselves lost in the system or having paged through their search results in one fell swoop. One definite advantage of having the planned WWW interface compared to the present interface is that it would be possible to tell to a certain extent that the search etc. was active. However this problem does not reflect on BIDS as it is a world-wide problem of traffic exceeding the capacities of networks.
To return to the design of the present interface - can it really be as nasty as Prof. Oppenheim implies? It certainly takes some getting used to as anyone who has had to train end-users can testify, but that does not in itself mean that it is a bad interface.
Certain elements can cause confusion until the user is comfortable with them. The division into menus and the listing of these menus at the bottom of the screen is perhaps the single most confusing element for new users. The BIDS team have recognised that the position of menus on the screen can make a huge difference to how the databases are searched. Terry Morrow quotes figures relating to the use of the Contents /Issues option in both the ISI databases and the BL Inside Information service. Contents information was accessed only in 0.6% of searches on ISI compared to 58% of Inside Information searches where the Contents/Issues menu was actually listed at the top of the Search menu options on the opening search screen instead of listed at the bottom of the screen. One would have hoped that BIDS would have made more use of this evidence by now and redesigned layout accordingly.
The need to change menus after searching in order to display results rather than select a numbered option, whether the search took place in the Search menu, Citation menu, Issues menu, can also cause much confusion even among fairly regular non-LIS users. Again it is the lack of Display as a given option in the main screen that seems to lead to this disorientation felt by users as they hop about in and out of menus - once of course they have located the menus listed at the bottom of the screen. In particular, the loss of search results from swapping between menus for displaying can be very frustrating, for example the inability to go back to the lists of issues available for the selected journal in the Issues menu after displaying a contents page from one issue without re-conducting the original search. This feature has however been accepted as necessary for review and development.
Other niggles that are often mentioned by new users is the slight inconsistency in field names between some databases. For example, what is the difference between a corporate address in the ISI databases and an address in the Ei databases?
The lack of on-screen examples of search types is another source of complaint from new users, but in BIDS defence I would say that their context sensitive help is far more useful than a few examples cluttering up the screen. However in common with the positioning of the other menus, the bottom of the screen position of the help command does not encourage its use which is a great pity as it is an excellant resource. The email help has also been praised by those who use it for its speed and effectiveness of service .
These are niggles, however. They are not to be construed as major criticisms. They are the type of items that come up on the BIDS wish-list, and I do not feel overall that Prof. Oppenheim's comments are justified. As Harry East found in his survey, the majority of regular users are happy with the service _and_ interface as provided. It is possibly only the Librarians and the trainers of end-users who come up against novice users (some of whom may well find getting around an OPAC difficult), who find elements of the interface frustrating. Frankly, I find the BIDS interface eminently more preferable to something like FirstSearch, which is a cluttered and messy screen telling you more than what you wish to know at the same time as not telling you what you want to know; or, to take an example of a cleaner screen, RAPID: [http://edina.ed.ac.uk/RAPID/about.html], which manages to list most options in the menu at any one time, but has the advantage of having far less options or menus to list in the first place and whose help system is more cursory and does not list how to exit the system beyond the initial screen. BIDS, after all, has been designed from the start with input from LIS professionals and does have an active User Group which lobbies for changes which have made the interface almost unrecognisable to users who have not been near BIDS since its early days. It also has an large number of daily users making on average 9,000 sessions per day so something must be right.
Many who are discontent with the present interface have welcomed the planned implementation of a WWW interface by the start of the next academic year. This will almost undoubtably attract more users and doubtless be more attractive to the casual user. However will it be able to carry the same functionality as the present interface and is BIDS in danger of sacrificing service to prettiness? The WWW is stateless, i.e. unlike a ftp session where it is possible to make several commands in one session with each command changing the state of the session, the Hypertext Transfer Protocol (http) treats each command as a separate and distinct session and there is no mechanism for carrying state from some session to another. In other words a search using a http would be treated as an individual session - the use of previous sets and the ability to set year/document type etc parameters for a series of searches would be compromised. The BIDS team are going to have to work hard to overcome these disadvantages.
The move to a WWW service should also be a cause for concern with regards to access. While it is necessary to move the lowest common denominator up a notch or two from time to time as technology generally improves, it would be a mistake to think that even all academics in all traditional universities, let alone new universities or colleges of higher education, have ready access to machines that can handle anything beyond a basic text only web browser. Here at the OU, we are particularly aware that students do not always have the access to the best, or even average, networking facilities. This situation is bound to change and in the meantime BIDS have promised continued support for the traditional interface, a commitment that can only be warmly welcomed and encouraged. However the WWW interface must take slow network connections into account when designing its pages if only as a gesture towards keeping network load as low as possible.
BIDS does have its problems - mostly historic or commercial - however they are for the most part minor. While I know many will disagree with me, I do not believe that BIDS is the useless, unfriendly and unhelpful beast it is made out to be. Users need training whatever the interface is - I have had to train people how to use Netscape which is held up as one of the most intuitive and user friendly pieces of software around. Therefore I really do not believe that a WWW interface with reduced functionality which may well be easier to train users on, is inherently better than the present command driven interface. With a bit of practice and more importantly, a knowledge of effective search techniques and methodologies, the end user does find BIDS a friendly, helpful and useful tool. As librarians and information specialists, we should be more worried about the lack of knowledge of search strategies in many students and researchers than debating pretty interfaces for a useful service, and we should devote energies to the former in preference to the latter.
1: H.East, E.Sheppard, Y.Jeal. A Huge Leap Forward, CCIS Policy Paper No. 5, BLR&D Report 6202, May 1995 also see the email from Harry East (H.East@is.city.ac.uk), 'Dear Old Bids' posted to LIS-LINK@mailbase.ac.uk on Tues, 20 Feb, 1996, archived at: [http://www.mailbase.ac.uk/lists-k-o/lis-link/1996-02/0364.html]
2: T.M.Morrow. BIDS - The Growth of a Networked Bibliographic End-User Database Service, Program, Vol. 29, No. 1, Jan 1995, p36
3: Prof. Charles Oppenheim (CHARLES@dmu.ac.uk) 'Re: Bibliometrics, dissertations & the BIDS interface', posted to LIS-LINK@mailbase.ac.uk on Fri, 16 Feb 1996 14:41:51 GMT; also archived at [http://www.mailbase.ac.uk/lists-k-o/lis-link/1996-02/0286.html]
4: Barbara Middleton (B.A.Middleton@newcastle.ac.uk) 'Re: How bad is bad? BIDS Software and history', posted to LIS-LINK@mailbase.ac.uk on Mon, 19 Feb 1996 09:48:24 GMT0 ; also archived at: [http://www.mailbase.ac.uk/lists-k-o/lis-link/1996-02/0315.html]
6: Tom Wilson (T.D.Wilson@sheffield.ac.uk), 'Bibliometrics and dissertations' posted to LIS-LINK@mailbase.ac.uk on Thu, 8 Feb 1996 12:21:24 +0000; also archived at: [http://www.mailbase.ac.uk/lists-k-o/lis-link/1996-02/0136.html]
7: T.M.Morrow. BIDS - The Growth of a Networked Bibliographic End-User Database Service, Program, Vol. 29, No. 1, Jan 1995, p36
8: Barbara Middleton (B.A.Middleton@newcastle.ac.uk) 'Re: How bad is bad? BIDS Software and history', posted to LIS-LINK@mailbase.ac.uk on Mon, 19 Feb 1996 09:48:24 GMT0 ; also archived at: [http://www.mailbase.ac.uk/lists-k-o/lis-link/1996-02/0315.html]
9: H.East, E.Sheppard, Y.Jeal. A Huge Leap Forward, CCIS Policy Paper No. 5, BLR&D Report 6202, May 1995 also see the email from Harry East (H.East@is.city.ac.uk), 'Dear Old Bids' posted to LIS-LINK@mailbase.ac.uk on Tues, 20 Feb, 1996 and also archived at: [http://www.mailbase.ac.uk/lists-k-o/lis-link/1996-02/0364.html]