All the electronic forms of data access are in their infancy. There are enormous issues to be addressed in at least three different areas – economic, technological, and design – and there are some worrying signs that bad decisions are being made.

For example, the New York Times is a typical Newspaper Site with a web design that is unnecessarily slow (frustratingly so unless you have a fast machine and high-speed access), has a search engine that is crude at best, and on top of that is charging $2.50 per archived article for access even though photos, charts, and graphs are not available (and they are still not paying writers for electronic reprints, preferring to appeal a court decision to the Supreme Court). In contrast, London's Daily Telegraph lets you search by section &/or date as well as keywords, and delivers the results free, as well as routinely linking to older articles and external sites. Unfortunately, the New York Times model seems to be the more common, at least in America.

On a more specialist level, Lexis-Nexis has a fast response but an interface that while commendably simple is remarkably unsophisticated. Temporary subscriptions are available, at $69/day or $129/week for full news access (the restricted one seems hardly worth the cheaper price; the legal-opinions service is separate). Subscription rates seem to be negotiable, presumably on expected use; my local university library decided it was too much, though I do not know the figures. The amount of data available is huge, though it seems somewhat less amazing now that so much is on the web anyway, but the system appears to be in transition from one with proprietary software that only lawyers with UNIX training could love ("to find documents concerning employee drug tests, you might use this search request: drug W/5 test OR screen! W/10 employ!") to one with a web-based natural-language front end. For now, in the News section, you select a database from pull-down menus, add an optional date range, and search for one or more key words, which simply have to appear in the same document. The potential for more complex searches is enormous, and the worry is that the adequate may freeze out the excellent, or at least delay its introduction.

EBSCO, which is available free to users of my local public library system (in Santa Cruz, CA; try calling your reference desk) is rather like Lexis lite. It covers far fewer newspapers, and thus is less useful for finding editorials and the like, but provides a solid overview of news reports. And the price is right.

MAGS and the related databases, generally available through university libraries, provide better coverage than Lexis, let alone EBSCO, when it comes to specialist journals, but far worse than either for general-interest publications. They suffer from a noticeable lag time and seem to be better considered as a source of semi-random discovery.

Most of the standard Web Search Engines are still focused on the perceived need to get the latest data as fast as possible. In part, this is because it is what they can do; NewsTrawler, a nifty site that lets you submit the same search to many different papers at the same time, explains that they do not perform date-related searches because "most news sources have varying date formats for limiting searches and some don't have them at all" and promises that the engineers are working on the problem. Meanwhile, it is all too common for a search to result in thousands of hits, a huge majority of which are irrelevant at best and offensive at worst. In my experience, the main function of such a search is to locate the rare site at which some individual has applied intelligence to impose order. In other words, I rarely find my questions answered directly, but I often find the place to ask; the search sites are more like the humble library catalog (which directs you to a book that has its own index) than the encyclopedic indices some people claim them to be.

As a delightful example of what can happen at a search site, there is something just a tad off-putting about searching on AltaVista for 'human cloning' and finding, at the end of a list of links to recent news stories, the following helpful suggestions, shown in a partial screen-shot:

Clicking the first one actually led to a reasonably good selection of books on the subject; I was rather hoping for something more exotic ...

A more serious problem with the search portals is a lack of discrimination As a perfect example, I just searched for 'human cloning' at Excite and the second site match was described as "Human Cloning and Re-engineering / Informs readers of work on human embryo cloning, and discusses the moral implica..." which sounded interesting, but it turns out that the site was last modified on 3/28/96 and was never that substantial anyway.

Much more useful are the static collections of sites, but even these need careful filtering. For example, Excite has one on cloning that ought to be useful and within limits is, except that it leads off with the wonderful "Americans for Cloning Elvis - Sign a petition to get scientists to do something useful - clone the King" which unfortunately was suspended on July 9, 1998 because too many people were signing on.

Which brings me back to where I started, at Yahoo!. Some people really dislike this portal, particularly for what it does if you use its customizing facility (which I never have), but I find it unfailingly useful. It collects websites and organizes them into categories (though they claim it is not true that the name is an acronym for Yet Another Hierarchical Officious Oracle), allowing you to search within categories, throughout Yahoo! or on the whole web (for which it now partners with Google and links to other search engines), and also carries news reports from several sources on its own website, thus reducing the need for external links and generally speeding up the process. It also applies what seems to be human intelligence to selecting items within its extended news categories. I can think of no other reason that the 'Magazine Articles' section of their Biotechnology_and_Genetics sub-category of Science news would include, as the first entry in four weeks, a New York Times review of a book on Mendel that includes the following pertinent quotation:

What drives our desire to decompose our complex, integrated wholes into sums of idealized and simple biological atoms? Why have we so unassumingly linked our very natures to the colors of Mendel's peas? Mendel might have been innocently simplifying a complex world. But Mendelism has much to answer for. Here hangs a great tale, untold so far in the history of genetics and of modern notions about heredity. Its absence may be the most notable aspect of Henig's book.

For the intelligent neo-Luddite, there remains the matter of Printed Materials. If (to choose a non-random example) you wanted to review the coverage of the Human Genome Project announcement on June 26, 2000, it would undoubtedly have been a good idea to frequent a large city newsstand in the last week of June, purchasing every available paper; getting them from any library is time-consuming and the photocopying even more so, and you still have to type in anything you want to quote. When the standard Periodicals Indices come out, they are probably the best places to start looking for archived materials, though they remain somewhat cumbersome.

There are also several academic journals dealing with Bioethics, which I have not investigated and should.