As an illustration of how deeply open-source software has penetrated the traditional computing realm, Sun adopted a version of the OpenSSH product in its own systems. In what's largely a marketing ploy against rivals Microsoft and IBM, Sun touts its openness. As examples, the company says it has no hidden interfaces to its Solaris operating system and trumpets that it supports royalty-free standards; that it shares the control of Java with many other companies; that it has released several important software projects as open source; and that it promises no nonstandard extensions will prevent customers from being able to substitute another company's software for Sun's. This claim to openness rankles de Raadt. "If Sun says they are open, they need to start acting it," he said in his Open BSD posting last week. However, Sun has taken the lead in several open-source projects. For example, the company has released software for better encryption in Web site transactions, in addition to introducing its OpenOffice competitor to Microsoft Office, its NFS file-sharing software, its Jxta peer-to-peer experimental software and its Grid Engine calculation software. The company's relations with the open-source community haven't always been rosy, though -- open-source advocates were displeased with Sun's decision to show the source code but not share the control of its Java software. More recently, it took months for Sun and the Apache Software Foundation open-source group to hammer out an agreement on working together. But for his part, de Raadt believes other companies have been more helpful to OpenBSD programmers -- Advanced Micro Devices, for example, has been eagerly soliciting open-source programmers' support to try to help the prospects of its coming 64-bit Opteron chip, code-named Hammer, which competes against Intel's Itanium line. "We have documentation for Hammer, and Hammer isn't even shipping yet," he said. UltraSparc nitty-gritty
As for Sun's UltraSparc III chip, OpenBSD needs details on how it uses high-speed cache memory and regular main memory, de Raadt said. "There are some low-level cache and memory management unit interactions that seem to be very different from the way it used to be on the older UltraSparc I and II. We run on old (UltraSparc) I and II machines," he said. David Miller, a Red Hat employee in charge of the UltraSparc III version of Linux, received documentation on the processor after signing a nondisclosure agreement with Sun. Still, he said in an email interview, it wasn't simple: to deduce Sun hardware bugs the company didn't tell him about, Miller had to decode Sun's upgrades to Solaris. Miller believes that it's "obvious" how the UltraSparc III works after looking at the UltraSparc III version of Linux. But the OpenBSD programmers have tried that strategy to no avail, de Raadt said. OpenBSD on UltraSparc III isn't a mere curiosity. De Raadt is particularly interested in UltraSparc III features that are well-suited to OpenBSD's emphasis on security -- for example, memory protections that make computers less vulnerable to buffer overflow attacks. And there are customers interested in an OpenBSD version for UltraSparc III machines. University of Alberta's Bob Beck said he is forced to buy out-of-date UltraSparc II-based E450 servers instead of newer UltraSparc III-based V880 machines for the university's SunSITE software exchange. "I want to run OpenBSD on (UltraSparc III systems), because I can do stuff with it for routing, security and traffic control that I cannot do even with a $100,000 Cisco box," Becker said in an email interview.
See Chips Central for the latest headlines on processors and semiconductors.
To find out more about the computers and hardware that these chips are being used in, see ZDNet UK's Hardware News Section.
Have your say instantly, and see what others have said. Go to the Chips Central Forum.
Let the Chips Central editor know what you think by email. And sign up for the weekly Chips Central newsletter.
