Who is Godot in the Java world?

Peter Judge: Sun's increasingly isolated position in the industry must leave the Java community wondering what will happen if the standard eventually has to stand on its own

Sitting in the Samuel Beckett conference room with the Java Community Program executive committee, I wondered who was taking the role (or rather, non-role) of Godot.

Themed conference rooms are an industry cliche, but if you are going to do it, then naming them for Irish writers strikes me as a pretty good one, especially if you are the JCP's hosts distributed software company Iona, in whose Dublin headquarters we were.

However, these room names are a tough act to live up to, and we had just come from a joint press conference with Intel and Iona, held in the Oscar Wilde room down the hall. And, much though I enjoyed it, I'm afraid it was not long on Wildean wit. "So Centrino doesn't support 802.1a or 802.11g? To lose one wireless standard may be regarded as a misfortune. To lose two looks like carelessness." I wish I'd said that, Oscar.

Now, in the Beckett room, the stage was surely set for bleak, eloquently speechless despair, for what the critics have described as "a terrible dreamlike vacuum, overcome by an overwhelming sense of bewilderment and grief".

Strangely enough, the players bore little resemblance to the cast of Waiting for Godot. The JCP are an optimistic, jolly bunch of standards-makers from different vendors, always laughing and joking to each other.

The JCP recently completed a 22-month legal battle to wrest a greater measure of control of the Java specification from Sun. "I had a full head of hair when I started that," said the leader of that battle, Donald Deutsch, Oracle's balding vice president of standards strategy and architecture -- to roars of mirth from his JCP buddies.

The group includes long-time former IBMers, including one of the original architects of the SNA network architecture, which ruled the world before TCP/IP, and a former Bell Labs man who led the early stages of development for Unix System V release 4.

These are the kind of people, in short, who should be out there in public now, to reassure the world that IT knows its history, and to restore an image that was all-but-destroyed by the brainless dot-com kids of the year 2000 who, in Oscar's words, gave us "the full benefits of their inexperience".

Looking round the table, journalists scouted for a fight, but got little for their trouble. A row between IBM and Sun might be tearing the Java world up (over IBM backing the open-source Eclipse software development project but these people liked each other too much to show anything but humour.

But still, I couldn't help wondering if we were waiting for Godot -- and who would be taking the role? In the original play, of course, Godot never arrives.

The most obvious gap at the table was IBM, of course, but the JCP people refused to dither about IBM's non-appearance, like Vladimir and Estragon in the play. The IBM representative had to leave when the formal proceedings ended and couldn't stay to face the press, unfortunately.

Perhaps the absence of Microsoft was more significant. The company is pretty unlikely to show at a Java event, but it was, of course, invited. A reference to Microsoft as an "absent friend" sent the group off into most un-Beckettesque peals of laughter.

Perhaps the closest to Godot would be Sun. Sun is the creator of Java, and many commentators see Godot as an absent God-figure in the play. And the delegates did show a bit less humour around Sun.

Sun's representative on JCP was not there to take part in the discussion -- like the IBM member, he was called away. Instead we had Sun's Onno Kluyt, who manages the JCP project office and does not represent Sun on the board.

The JCP people seemed a little weary of dealing with Sun, though deeply grateful for the way the company bankrolls the Java movement. Where it costs hundreds of thousands of dollars to join other standards makers such as the Object Management Group, it is virtually free to join JCP, thanks to Sun.

But they must, if they are honest, be wondering what will happen as Sun's future evolves, and how the Java process looks to a company which is still a hardware-first company, even though it is learning to accept Linux and work with Intel machines. For most people blade servers are an Intel-first phenomenon, but Sun launched its blade servers SPARC first. Sun, like Apple, is driven by a visionary, whose vision may not always be the most practical move.

It must be a little odd to be a software appendage to that kind of outfit, and the JCP has very good reasons to want to make its process as independent as it can from the original founder of Java.

To move to a political metaphor, whenever Sun has got involved in software it has attempted to behave like a feudal monarch, with divine right over the software. It was that way with its Unix partnership with AT&T in the 1980s, and it attempted to lord it over Iona in the object field, when it owned a quarter of that company.

The JCP legal effort, led by Deutsch, has painstakingly moved Java on to a constitutional monarchy. The old king, Sun, still has a veto on the server edition and enterprise edition Java standards, and holds the patents for the software.

If, for whatever reason, the addled monarch becomes unable to fulfil its role in future, the Java world would have to either elect a new constitutional monarch (say IBM?) or else turn itself into a self-governing republic.

The drawback there is that members would have to pay more without a sugar-daddy. And Java Republic sounds too much like a coffee shop to me.