It is time for the government to cut bait on GILS -

The 21/2-year-old mandate for the Government Information Locator Service is being
re-examined by the Office of Management and Budget, the General Services Administration
and the National Archives and Records Administration.

Intended to improve access to federal information sources, GILS seems to have had
little success meeting that objective.

This disappointing showing is in sharp contrast to the many excellent World Wide Web
services from which the public can choose. Therein lies the cure for what afflicts GILS.

According to its advocates, GILS offers a technical solution superior to the Web.
Unfortunately, few people understand both technologies well enough to make a definitive
judgment about which is superior. But that nicety has not stopped the market from making
its decision.

As with Beta and VHS videocassette formats, we can debate which is technologically
better. But there is no argument over which has prevailed. Similarly, the Web rules.

The technology that OMB and the National Institute of Standards and Technology required
agencies use to implement GILS has not caught on. Z39.50-1992 is the international
standard upon which GILS is based. It comes from the same standards process that brought
us the Government Open Systems Interconnection Profile. And there are a number of
disquieting parallels between the GILS mandate and the now-defunct GOSIP.

For example, when GILS was made a Federal Information Processing Standard, there were
no interoperable implementations. Even a year later there were only a handful of
Z39.50-1992 implementations and little experience in their interoperation.

There was no graceful migration path from Z39.50-1988 to Z39.50-1992, as required by
GILS. As with the GOSIP e-mail standard X.400, proponents of the Z39.50-1992 standard
abandoned its installed base to achieve a political and technical compromise.

Collecting data in a GILS-compliant format helps agencies create a catalog of their
information assets. Agencies can put the catalogs on their Web sites to inform the public
of information systems not yet on the Web or not destined for it. By contrast, most Web
site maps and home pages focus on the documents in, or reachable by, the site itself.

The elements in a GILS record can also provide a comprehensive description of an
information system or collection--data that can be useful for agency employees, managers,
government customers and the public.

But the GILS mandate needs to be revamped to reflect the technological realities of the
decade. Agencies should be encouraged to put their GILS records into a Hypertext Markup
Language format and populate them with hot links to data. This will make them easier for
commercial Web search engines to identify and catalog.

The commercial Web search services are excellent and cost taxpayers and users nothing.
With their relative strengths and weaknesses, they offer niche services to those seeking
information. Yet they do so in a manner that complies with the Internet standards that
have become universally popular over the past few years. There is no reason for taxpayers
to pay for a search solution that competes with free commercial services.

If it makes business sense and meets your customers' needs, then by all means your
agency should buy, install and get trained on a Z39.50-1992-compliant server. Perhaps, in
time, Z39.50 will become a standard feature of the typical commercial Web server. If I
thought Z39.50's information discovery techniques had an advantage over Web search
engines, I might believe that the investment in GILS servers is justified. If Z39.50 does
have advantages, let its proponents market them to the commercial Web search services.

Until then, OMB should continue to suggest that agencies send their GILS records to the
Defense Department, NARA or FedWorld, where they can be made Z39.50-compliant for a
minimal fee.

Advocates for reforming GILS have conducted an exhaustive study that is available--on
the Web, of course--at

The Web address reminds me of the waning days of GOSIP when the most accessible and
timely sources for GOSIP documentation were available over the Internet. That's when I
decided to bail out and jump on the Internet bandwagon. Likewise, the time has come for
the federal government to cut its losses on Z39.50.

Walter R. Houser, who has more than two decades of experience in federal information
management, is webmaster for a Cabinet agency. His own Web home page is at

inside gcn

  • pollution (

    Machine learning improves contamination monitoring

Reader Comments

Please post your comments here. Comments are moderated, so they may not appear immediately after submitting. We will not post comments that we consider abusive or off-topic.

Please type the letters/numbers you see above