IP: JAVA RIVAL IN THE WORKS TOP BELL LABS SCIENTISTS INVOLVED

[Dave Farber <farber () central cis upenn edu>]

Java rival in the works Top Bell Labs scientists involved
By Lee Gomes


Mercury News Staff Writer


SAN FRANCISCO -- For the last nine months, some of the AT&T Bell Labs 
researchers credited with the development of the Unix operating system have been
 working on a secret crash project: creating a rival version of Sun 
Microsystems' Java, an on-line-oriented language that brings sophisticated 
programming tools to the World Wide Web.


The project, code-named Inferno, is of such importance that the Bell Labs group 
put a major ongoing research effort, a successor to Unix known as ``Plan 9,'' on
 the back burner, and roped in most of the unit's 10 or so scientists to work on
 it.


Ritchie mentioned project


Dennis Ritchie, a legend in the Unix world for his role in creating both Unix 
and the C programming language, made a brief reference to the undertaking 
Tuesday night in a speech at the UniForum trade show here. He discussed the 
matter further, if somewhat reluctantly, in an interview.


While noting that Java has been the beneficiary of ``a lot of hype,'' Ritchie 
said the underlying idea behind Java was compelling, and said he did not want to
 criticize Sun's work. But he said he worried that Java had become too large and
 complex; he also suggested that any Bell Labs version would be useful in a 
wider variety of machines, including future versions of television sets.


``Java does not go far enough,'' he said.


Lively contest?


Should Bell Labs' work find its way into a commercial product -- something that 
is by no means certain, if only because AT&T is expected to soon join the 
growing list of Java licensees -- it could make for a lively contest for 
high-end, Internet-oriented programming languages, a category Sun had been 
presumed to own.


That's because while Sun has for itself one of the industry's best technical 
reputations, that's even truer for Murray Hill, N.J.-based Bell Labs, considered
 one of the world's pre-eminent research organizations. The facility has 
numerous major advances in computing to its credit, as well as such fundamental 
inventions as the transistor and the laser.


And as a developer of the C language, a Java ancestor, Ritchie would be uniquely
 qualified to head a group creating a new on-line language.


Sun introduced Java last year as a way of bringing the power of full-blown 
computer programming to the World Wide Web. In an extraordinary rush of 
enthusiasm about Java, many people viewed it as a kind of one-stop, all-purpose 
software tool that would allow Web site developers to easily add such features 
as animation to their pages.


Java, though, is in fact a quite complex computer language that is beyond the 
technical ability of the great majority of people working on Web pages. And as 
that fact has sunk in during recent weeks, Sun has become more deliberate in 
describing Java. It is now careful to say, for example, that Java's main appeal 
will be to professional computer programmers developing sophisticated 
applications, and that most people wishing to publish ``content'' on the Web 
will probably use other tools.


Still, because Bell Labs has a long-term, even theoretical approach to computing
 that is presumably immune to industry fads, its interest in a Java-like 
language is yet another indication of the profound manner in which the growth of
 the Internet is affecting the thinking of nearly everyone involved with 
computers.


Ritchie did not say how far along the Inferno work had proceeded, and indicated 
no decision had been made about whether it would be taken to the marketplace as 
a product. He did say, though, that top management of AT&T has been keenly 
interested from the beginning.


Any uncertainty about the Inferno project is compounded by the fact that Bell 
Labs is about to undergo a major change, as a result of last year's break-up of 
AT&T into three companies.


Ritchie's group, along with most of the lab, soon will belong to Lucent 
Technologies, which has been set up to sell AT&T's business equipment, including
 telephones and high-end switching equipment. And the new company is still 
developing its business plan and strategy.


For its part, Sun said the interest in Java by a computer scientist of Ritchie's
 caliber was a vindication of Sun's basic ideas about the technology.


``He could have said it's a load of hooey,'' said Tim Lindholm, of the Java 
development team. ``The fact that he's taking it seriously justifies the notion 
that Java is a viable story.''


Ritchie suggested that AT&T had decided to not invite Sun to collaborate. But 
Lindholm said Sun would nonetheless like to pay a visit. ``It would be 
interesting to hear what ideas they have,'' he said. ``Java is not a done deal. 
We intend to continue developing it.''


While he was only hearing about the Bell Labs effort for the first time, 
Lindholm disputed Ritchie's technical criticism of Java, saying Sun would be 
delivering many varieties of it for use in many different systems.


Ritchie's appearance in San Francisco was in connection with the first public 
discussion of Plan 9, a new post-Unix operating system from his group that is 
named after the cult science fiction movie.


An early version of Plan 9 has been released to the technical community. Ritchie
 said t additional development on Plan 9 was put aside because of Inferno.


Published 2/15/96 in the San Jose Mercury News.