OpenSolaris – A New Era

June 16, 2005 9:41 am General

It was really nice to see being released this week – it’s easily another milestone that Sun can be immensely proud of. Initially I was pretty skeptical about the project’s success, since it was yet another huge dump of code that Sun was throwing over the wall. Yet another ‘sink or swim’ moment. Historically, there’s definitely a few cases where we’ve sunk.

Right now, from my perspective, OpenSolaris is as yet unproven. Despite how successful Solaris has been in the past, OpenSolaris is a completely different ball game. There’s a huge number of potential hurdles in place – from basic community contribution in terms of lines of code and participation in the various existing processes like Architectural Review Committee, right down to documentation, marketing and general advocacy of the product. It’s very clear, right from the start, the project needs to enlist community champions.

Fortunately though, OpenSolaris already has many of these – both within Sun, and outside. Anyone who’s capable of this is bound to have a following of OpenSolaris fanboys before too long.

What already turned my eye is the fact that for once Sun is listening, and I think we’ve been guilty of not doing so in the past. It’s listening to the engineers, the partners, and the customers. It’s ultimately doing the right thing – we can’t afford to do anything else.

  • blogs.sun.com has been a huge success, so much so that there was no big PR job around the OpenSolaris release – just a collection of diverse blogs about OpenSolaris from clueful people like Jim Grisanzio, community manager of OpenSolaris, Claire Giordano, OpenSolaris marketing, not to mention all the developers. These are real people, without which the project couldn’t survive. I’ve said this before about other projects.
  • I’m just can’t believe how cool it is that the OpenSolaris guys set up a BitTorrent for the download – that’s a *huge* step, and I’d very much like to see this happening for other projects that Sun supports.
  • Even better, is the release of the Sun Studio Tools. Okay, so they’re not open source, but are released under a pretty flexible license – Provided that You are a participant of the OpenSolaris community, You may use the software for Research and Instructional, Individual, and Commercial Use to design, develop, test, or otherwise engineer software. That’s fricken amazing.

So what does this mean for GNOME? At the recent GUADEC conference in Stuttgart, there was much conversation about the performance and profiling of GNOME. The key message from Robert Love’s Optimal GNOME Programming talk was the fact that we should use the tools that are already out there. Now we have that opportunity. Download Solaris, download Sun Studio 10 and benefit from those tools.

From my perspective, as part of the desktop team within Sun, it brings other challenges. It gives us an opportunity to change our work processes and I think we as a team probably need to do better job in keeping the desktop up to date, and relevant. That brings a number of hurdles –

  • The current version of GNOME in Solaris [and Nevada] is 2.6 – I’d very much like to see this being updated to 2.10 before the release. I think we can make it happen and I’ll be pushing that direction, but 2 months is a pretty tight schedule – it’s not just integration of core code, it’s the 800+ patches we have, the user online help, the localization and branding. Watch this space.
  • How do we coordinate the ongoing development of GNOME with respect to OpenSolaris? Are we likely to build up a community around the desktop, or do we point them to the GNOME community instead? If we do, it seems sensible that we work more closely with the community – develop, QA, document and localize all on the latest community code. Integrate and brand stable releases into OpenSolaris, which in turn gets pushed into Solaris releases. Rinse and repeat. Will there be official releases of OpenSolaris, or just a series of snapshots?
  • Will the OpenSolaris community want to have snapshots of unstable development GNOME made available?
  • Can we hide our developers from the mundane and sometimes legacy processes so that they can focus on the code and making useful contributions?
  • Ultimately, what do we release that is useful to OpenSolaris? Does it make sense to mirror a bunch of community code with our patches supplied? Do we just make our patches available? Do we shift our entire build process and CVS outside the firewall and into OpenSolaris?
  • How will OpenSolaris and Nevada interact in the first few months? What are the rules for what changes we can make? Do we have any more flexibility with OpenSolaris? Can it be a test bed for changes like Fedora is to RHEL?

There are lots of unanswered questions, but Wow! I’m excited. First step is starting a OpenSolaris community for the desktop/GNOME – and go from there.

For everyone involved in OpenSolaris, from the developer, manager and lawyer, Thank You! – words that aren’t said enough in open source and free software.

Comments are closed.