91av

REPORTS of its imminent demise may have been greatly exaggerated, but there
is no doubt about it: from New Zealand to the Shetland Islands, the Internet
really is in serious trouble. Or, at least, its most popular face, the Web, is.
And unfortunately, this is no arcane wrangle over encryption or moral panic over
pornography. This time the issues affect everyone because at the heart of the
problem lies the hypertext link, the very thing that makes the Web so
powerful.

In the bad old days before the Web, finding information on the Net was a
long-winded and often fruitless task. Researchers hobbled from one site to
another performing awkward keyword searches, picking through file indexes and
invoking arcane transfer protocols. But that all changed in 1989 when Tim
Berners-Lee, a computer scientist at CERN, the European centre for particle
physics, dreamt up a way to apply the hyperlink to the Internet, and in the
process invented the Web.

Today, a page designer can, for example, highlight a quote from a Shakespeare
play and add a hyperlink to it. Readers who click on it could be transported in
seconds to, say, a site that carries the full text of the play, or even the
playwright’s complete works. So anyone can amass huge amounts of information on
a topic simply by following a trail of hyperlinks around the world.

All of which sounds innocent and altogether worthwhile. But the arrival of
commercial companies on the Web has led to conflict. Software publishers are
suing people with hyperlinks to pirated programs, and newspapers have sued to
keep competitors from linking to their sites. In these cases, people are trying
to stretch old ideas of authorship, ownership and responsibility to fit the new
medium. But many don’t think the old rules fit. They warn that unless new ways
are found to tackle these problems, the Web could begin to unravel. “Linking is
a fundamental part of the technology,” says Juliet Oberding, a lawyer based in
San Diego, California, who specialises in Internet law. “It’s ludicrous not to
let someone link to your site.”

Falling apart

The obvious fear is that anything which stops people freely forging links
between Web sites could lead to insularity and fragmentation. “Our goal is to
promote interoperability of the Web. We’d be sad to see that fall apart because
of the Balkinisation of the Web,” says Joseph Reagle, a policy analyst for the
World Wide Web Coalition, a body that sets technical standards for the Web.

One of the most widely watched battles at present is being waged in the
Scottish islands between The Shetland Times newspaper and an upstart
Web-based service, the Shetland News. The Shetland Times runs its own
Web site where readers can catch up with the latest stories
(http://www.shetland-times.co.uk/). The News started to offer hyperlinks to
these stories alongside its own original reporting
(http://www.shetland-news.co.uk). Last October, The Shetland Times went
to court and forced the News to stop linking to its site (This Week, 16 November
1996, p 5). A final decision is expected to be made by the Court of Sessions in
Edinburgh.

The Shetland Times argues that the News is breaching its copyright.
But the News says that it is simply providing references to stories that its
readers might want to see on its rival’s site. This seems to be the first case
of its kind to go to court and it could set a precedent for other businesses on
the Web.

“What we have is a new generation of publishers thinking in the old
paradigms, and trying to change the paradigms on which the Web is based,” says
Bruce Simpson, who runs several Internet news sites in New Zealand. The Web is
built on the idea that if people put up information they are happy for anyone to
take a look—and, by implication, anyone to link to it.

Simpson himself discovered that this wasn’t the case in November. That’s when
New Zealand-based Independent Radio News threatened to sue him. Simpson runs an
Internet service called 7am, which links to a number of other news services
(http://www.7am.co.nz). One of these is called Xtra, which in turn buys daily
news stories from IRN and posts them on its Web site.

The trouble started when IRN decided that Simpson’s link to Xtra violates its
copyright. IRN thinks that if Simpson wants to give his readers access to its
news, he should pay for the privilege. But while that thinking makes sense for
radio stations, the logic is less compelling in a medium which assumes that
information is open to all.

Simpson is not unsympathetic to copyright concerns. In fact he posts a notice
on his pages asserting his own copyright over his own material. But in this
case, he says, he is not republishing any of IRN’s information. He is merely
telling people where to go to see it. Simpson argues that no one could complain
if he put up a page that told people about the IRN reports, and gave them the
Web address to type in. The only difference about a hyperlink is that it saves
readers the trouble of typing in the address and lets them click a button
instead.

So, what is a hyperlink most like? Is it like a library index card, telling
readers where to find more information about a topic? Or is it like including
someone else’s work in your own without asking permission?

The state of Georgia thinks the latter is the case. Last year, Georgia passed
a law that requires anyone wanting to link to another site to first receive
permission from that site. The law created uproar, and is being challenged in
court by the American Civil Liberties Union and others. “Links are what make the
Web a democratising and powerful medium,” says Ann Beeson, a lawyer for the ACLU
in New York. “You can be critical of an organisation or entity, and then link to
their Web site.”

Take, for example, another plaintiff in the lawsuit against the state,
Electronic Frontiers of Georgia. It is critical of some of the activities of the
local telephone company, Bell South, and on its Web page carries a hyperlink to
Bell South. It encourages readers to use the link for more information and to
lodge complaints. Beeson says that this type of political activity would be
curbed by the Georgia law.

Some of the recent controversies are not about one site offering links to
another, however, but about how the link is made. Take “Dilbert”, for example.
This American comic strip chronicles the sufferings of a heroic computer nerd at
the hands of his clueless corporate masters. A company called United Media
Syndicate controls the strip, and puts it on its Web site about a week after it
appears in the newspapers (http://www.unitedmedia.com/comics/dilbert).

Dilbert has become wildly popular on the Net, and many people began providing
links to the comic strip. That was fine with United Media—so long as the
link took the reader to the Dilbert page. But some people set up a special link
called an “inline”. A page with an inline link brings up the comic strip on that
page, rather than sending the reader on to United Media’s page. Not
surprisingly, United Media was not happy. “It’s not acceptable use when you
inline that graphic into a copy that’s out of our control. I have an obligation
to say you can’t do that,” says Jonathan Shapiro, vice-president in charge of
United Media’s Web page.

He says that if the company doesn’t exercise its copyright it stands to lose
it entirely. Besides, the Dilbert page carries adverts and the company wants
people to see them. “At the end of the day, the people who work here have to
eat, and they want to get paid,” says Shapiro.

Putting up a fight

Most people removed the inlines without a fuss. But Dan Wallach, a
postgraduate student in computer science at Princeton University, New Jersey,
put up a fight. He argued that he wasn’t violating copyright because he wasn’t
making a copy. His page never contained a copy of the Dilbert strip. Rather, the
inline link instructed a reader’s browser—the program that allows people
to zip around the Web—to go to United Media’s site, retrieve the strip,
and display it along with the material on Wallach’s page. “In the most technical
sense I’m not making the copy, your browser is making the copy,” he says. “The
issue is, am I allowed to instruct your browser to do something that the other
site doesn’t want it to do?”

In a letter threatening to sue, United Media referred to existing law, noting
that rebroadcasting transmissions of sporting events, performing music without
paying royalties and displaying a work without permission are all violations of
copyright. “I am quite confident of UM’s legal position,” wrote the company’s
lawyer, John Parker.

But not everyone shares that confidence. Joe Lester, a lawyer based in
Minnesota who followed the case, advised Wallach that no one will be sure how
the issue will turn out until it is thrashed out in court. But whoever decides
to pursue the test case would probably have to spend $500 000. Wallach,
who hasn’t got that kind of money, dropped the page. “What you think you should
be able to do and what you can get away with are two different things,” he now
says.

A different twist on this theme is taking place at Total News, a Web site
that provides links to the sites of news organisations such as CNN and Fox
(http://www.totalnews.com). In this case, Total News shows entire pages put up
by the other news organisations, ads and all, but it shows them within a “frame”
of its own. So the reader always sees the Total News logo, along with Total
News’s adverts.

CNN and some others objected, and sent a letter to Total News telling them to
drop its site. In response, Total News added a disclaimer clarifying that there
is no affiliation between the organisations. But unless the case goes to court,
which it so far has not, the legal issues will stay unresolved.

Who should be able to link to a site and how they do it are not the only
controversial issues raised by hyperlinks. Just what is the position when you
provide a link to another site? Do you become in any way responsible for what is
published at that site?

This was October’s hot potato, when the Software Publishers Association in
the US threatened to sue Intergate, an Internet service provider that connected
a woman calling herself Tapu to the Net. Tapu’s home page contained almost
nothing but links to other sites, many of which contained illegally copied
software and tools designed to “crack” the copyright protection that prevents
people making multiple copies of commercial programs. Faced with the threat of a
lawsuit, Tapu’s service provider cut her links. Dozens of other sites, however,
sympathetic to Tapu’s cause, provide mirror copies of her page complete with
links (see for example: http://www.kaiwan.com/~slayer/oof/ref.html).

Shut down or be sued

The SPA made the same threat against Community Connexion, a service provider
based in Oakland, California, which had customers with similar links to pirate
sites. But Community Connexion’s president, Sameer Parekh, refused to shut down
the sites. Parekh agrees that he would be obliged to remove pirated software
from his own system, but disagrees that he should remove references to other
sites that carry such software. The SPA should go after sites that store the
offending material, he says.

The SPA claims that providing links to the sites is “contributory
infringement” of American copyright laws. However, the association dropped a
lawsuit against Community Connexion within a couple of weeks of starting
proceedings, and says it is trying to negotiate with Parekh instead.

Federal regulators in the US are also concerned about whether a link to a
site implies some sort of endorsement of the information there. In October, the
Food and Drug Administration held a conference to discuss whether it should
police the links made by drugs companies to other Web sites. The FDA already
closely regulates what drugs companies can say about their products. It approves
prescription drugs only for particular conditions, and manufacturers cannot
publicise their drugs for unapproved treatments—even if some doctors give
them for other disorders.

The drug fluoxetine (Prozac), for example, is approved only for treating
depression, though some doctors reckon it is also good for treating eating
disorders. Eli Lilly, the maker of fluoxetine, cannot publicise this second use
unless it obtains FDA approval. But just suppose that Eli Lilly were to post a
Web page with information about fluoxetine, and included links to sites about
eating disorders which carried information about using fluoxetine as a
treatment? Should the FDA ban those links?

On the other hand, what if a quack with a useless treatment runs a Web page
with links to the FDA, the US National Institutes of Health and other
prestigious organisations to create the impression that these organisations had
endorsed the treatment? Should that be prevented? Delegates at the FDA meeting
could reach no firm conclusions on these issues even after two days of debate.
More discussion is inevitable.

If the use of hyperlinks is to be restricted by law, the companies that stand
to lose most are those running search engines such as AltaVista, WebCrawler, and
Lycos. Search engines travel the Web and automatically index the millions of
pages they visit (“Searching the Web for gigabucks”, 91av, 6
April 1996, p 30). Lots of them could potentially contain pirated software or
material that violates someone’s copyright.

Virtually necessity

“The way we look at it when running the search engine is that if the files
are publicly accessible we think it’s okay to index them,” says Jim Reinhold of
WebCrawler. Search engines are a virtual necessity for Web users, making it
possible to find quickly the one document in 50 million you are interested in.
But a law that required the companies running search engines to obtain
permission before making a link, or holding them liable for links to pirated
software, could make it impossible for them to operate, says Reinhold. So far,
he is not aware of any legal challenges to search engines specifically. But that
could change. “In the past six months, the legal issues are coming up more and
more,” he says. “We are getting more lawsuits on the Web.”

Lawyers are divided over whether existing laws are capable of resolving the
issues raised by hypertext links. One way out might be if Web developers work
out new ways of protecting intellectual property and rewarding its creators.
“Digitally signing” a piece of work might provide a partial solution, for
instance. Then, anyone who wanted to could use the work, but it would always be
recognised as belonging to the author. But that still leaves the problem of
inventing ways to pay for such information. Some have suggested automatic
payment systems using digital cash that would allow authors to collect money
every time someone looked at their work—whether the reader accessed it
directly, or via a hyperlink from another Web site.

Web developers can also avoid a lot of the legal issues simply by using
existing technical solutions, says the New Zealander Simpson. Web designers
already have the means to stop hyperlinks being set up to their pages, for
example. They can direct surfers instead to a “front door” that makes it clear
where the site’s information comes from. And if people really don’t want others
getting to their information, they can always use password protection at their
sites.

Simpson believes that people who stop others linking to their sites are not
fully exploiting the Web, and that competition will teach them the error of
their ways. “The sites that use the power of the Web are going to benefit,” he
says. “Those that struggle against it are going to fall by the wayside. The Web
is too big to fight against.”