The day after a crucial poll on Java's status, Sun Microsystems (SUNW) said yesterday that it will pursue its efforts to make its Java development language an official standard, despite a vote against it by a U.S. group last month.
"I am cautiously optimistic in succeeding in our role to have Java become a standard," said James Mitchell, a Sun vice president. "We are going to follow the process, get comments, and work with member bodies to be sure they understand our position. If they understand why and what we are doing, problems tend to go away."
Mitchell's optimism was voiced in a teleconference with the press and analysts to update them on Sun's request to speed the process for making Java specifications into standards. It comes despite the June 5 "No with Comments" vote by the U.S. group working on Sun's request.
But Sun doesn't want any Java standard to bear the Java name, Mitchell indicated. "There is a tension here between a brand name like Java...and wanting to have a standard for the specs behind the product," he said. Asked if the standard could be called something other than Java, he responded, "Lots of standards have no name," mentioning the ISO 9000 quality standard as carrying a number, not a name.
Sun has asked a division of the International Standards Organization (ISO) and International Electrotechnical Commission to allow it to submit Java directly to the standards body, without going through other groups first. If approved, Sun would become the first commercial company ever allowed to go directly to ISO, although a number of organizations have such status with ISO.
In ISO/IEC jargon, Sun's wish to become a "publicly available submitter," or PAS. A PAS acts like a standards body itself, accepting submissions to improve and update a technology and passing them onto the joint technical committee of ISO/IEC for final approval.
Critics are worried that, given control of the submissions process, Sun will not make Java a truly open technology. So far, only nonprofit, membership-based organizations such as X/Open and the Video Electronics Standards Association have been approved for PAS status. Sun is the first company to apply for that distinction.
"The U.S. vote to disapprove is not unusual in responding to a standards ballot," Thomas Frost, chairman of the Joint Technical Committee (JTC 1 TAG), said in June 5 a statement. "It usually signals that the U.S. approval hinges on the issues identified in the comments."
The U.S. group's statement was generally supportive of the effort to standardize, indicating the issue was not whether to standardize Java but how to standardize the language.
"We believe we can successfully resolve this matter in the next 60 days," said Sun's Mitchell.
Mitchell's comments follow the submission of ballots yesterday by the ISO/IEC members involved in deciding the fate of Sun's request. Five countries that have not responded have until July 21 to vote. After that, Sun will have 60 days after all comments are submitted to respond.
The results of the polling will not be released to the public, only to Sun and the 27 members considering Sun's request, said Lisa Rajchel, a director of the American National Standards Institute.
Rajchel said after Sun's responses are distributed, the 27 members considering the Java issue will have 45 days to determine whether they have reached consensus on Sun's proposal.
The U.S. working group, which turned down Sun's request in June, outlined three areas of concern: the Java trademark, scope of the application, and maintenance of the Java standard.
Responding to the U.S. comments, Mitchell said the Java trademark should be handled separately from the Java language itself.
"Standardized specifications are separate from a trademark," said Mitchell. "Sun's approach was to submit only the technologies that define the Java platform," not other Sun technologies.
Other issues concern whether a for-profit company should be allowed to submit directly to the ISO, the scope of the specification itself, and how the Java standard can be maintained and updated, Mitchell added.
"No matter which route we had chosen to make Java a de jure standard, most of these issues would have come up," he added.