Sun, Solaris, and a new chance to shine

With 12 million Solaris licenses now in the market, Sun's Solaris is no slouch. And while people like I talk up Red Hat's clean-up of the "certification market" [PDF] (with over 3,000 applications certified for RHEL), the chart below indicates that Solaris actually has a pretty compelling application certification story to tell.

(Credit: Sun)

The question is, "Is it enough?"

A few days ago I suggested that Sun would be wise to partner closely with Ubuntu (Read: Acquire Canonical). It seems the easiest route to continued open-source momentum as Linux vendors continue to cut into Unix. But there's a compelling story in the Solaris numbers that suggests that it may live on for a very long time.

This isn't to say that Sun shouldn't be looking to Linux for continued growth in Linux. I believe it should. The market is behind Linux and Sun could send Linux into hyperdrive. Just consider what it brings to the Linux table:

(Credit: Sun)

Not too shabby.

The future of Sun is to corral the expanding choices that open source offers, which includes Linux, as Marten Mickos underscored at the Linux Foundation's recent conference:

"We are still committed [to Linux]," Mickos said. "If we are not committed, than any one of you can take the MySQL code and fork it to make a new MySQL product, which I am sure you would do" if Sun tried to converts LAMP to SAMP.

Mickos was peppered with questions from the Linux crowd. "What is Sun's Linux strategy? What degree of freedom will the MySQL group have inside of Sun?"

"There's a realization inside of Sun that the days of the monoliths are over, but the Web continues to grow." Sun wants to participate in that growth, not by sticking to Sun-only technologies, but by offering a smorgasbord of choices to its customers, including a core set of Java/Solaris/Sparc technologies.

Amen. If Sun can get this right - if it can make choice a marrow-deep belief throughout its massive organization - then it's very possible that Sun can lead the market again. It's a big "If" today, but I like to see open-source advocates like Red Hat and Sun leading the software industry, rather than the proprietary dinosaurs of yore.

[deepwave]

I don't think Sun acquiring Canonical would be wise for Sun, simply because the two really have different focuses. Sun does well in enterprise software, helping to integrate various systems in sane ways with its Java, SQL and other FOSS offerings. Canonical however seems to be heading in the direction of the consumer market, with emphasis on the personal desktop and mobiles. Two totally different directions. Besides Mark Shuttleworth probably wants to hold onto his pet experiment for a little while longer.

[odubtaig]

The community base on the desktop is Canonical's way [i]into[/i] the enterprise. Broad community support is nothing but a bonus when you want participation and fast bugfixes (or even many-eyes bug reports).

The core reason Sun's unlikely to buy Canonical is simple. They can't control it. Look at the way projects like OO.org, Java and the newly acquired MySQL are handled. Notice anything? The one thing they all have in common (asides from the GPL) is that all submissions to these projects require the submitting party to hand over all copyright to Sun, on the condition that Sun releases the changes under the GPL. This does, of course, allow things like the now almost abandoned StarOffice to be closed source without legal issues. It's also something that would be impossible to do with a whole Linux distribution.

Besides. Sun seems obsessed with only accepting contributions if the contributer gives up all rights to their own work. This is a community crushing move which has severely hindered OO.org development and improvement. In the meantime Canonical's all about the community and Ubuntu's flourishing for it... just so long as 8.4 is an improvement on the last release :p