JDS Closer to Open
August 31, 2006 General Comments Off on JDS Closer to OpenWe’re finally one step further to really being “open” – Dermot’s mail this morning gave me my morning stiffy.
We’re finally one step further to really being “open” – Dermot’s mail this morning gave me my morning stiffy.
Good to see the Adobe guys making some good choices in terms of API use, though would be great to see them move towards GStreamer.
Good to see new opportunities are being created around OpenSolaris and related technologies – people like Ben moving on to new jobs, Apple hiring for DTrace hackers, and many other bits and pieces happening in the wings. Open source projects have the ability of creating superstars, and nice to see that happening closer to home as well.
Ahh yes, it’s lovely to see a bit of life back in the corporate world, IBM bashing Sun over OpenSolaris and Sun mocking HP among the notable ones last week. Bring it on.
However, I guess I was slightly disappointed to see the response from various Sun people. I can absolutely appreciate that OpenSolaris is everyone’s baby, and those people involved can be really proud of the long hours they have put into it, but I think there’s a lot of value to be learned from Dan’s comments. While I don’t agree with some of the points he made, it’s one perception of what others may think about the project and something that we should take as constructive feedback, however aggressive those comments are.
The truth is, there may never be another open source project like Linux, that caught on as quickly as it did. There was an obvious gap in the market and the project was still very early in development. Sun, and OpenSolaris, may well have missed that specific gap. However, do I think that there’s no room for OpenSolaris? Hell no, and the various innovative features [DTrace, BrandZ, Zones, SMF, FMA] that have gone into Solaris 10 is a good reminder of that. Choice is good. Building a community over night to the same fit as the current Linux community was always going to be hard. You have to find your niche. Ubuntu found it by providing regular 6 monthly releases tracking the latest and best code the open source community could offer, and provided a mechanism for a much wider distribution. It’s a challenge that I’m sure the OpenSolaris community will take up.
Dan’s right, the code is still stored behind the firewall. Yes, that *is* a problem that we need to fix. However, in a rather weird twist, that’s one of the things that Sun got right. As frustrating as it currently is to push back code [and really you should be talking to one of the people who have done this from outside], it gave the opportunity to discuss what development model works best for *everyone*, not just those internal to Sun. Better still, we got the people writing the code for the various distributed SCM’s to get involved in the discussion about which one to use. It’s nice to see open collaborative progress.
Rich Green described OpenSolaris as a run-away success. While he may have been talking about numbers of contributions or adoptions, I’d like to think it was the fact that we’re trying for the first time to encourage our 1000’s of Sun engineers to interact with a wider community, working towards a common goal. That’s not an easy thing to do – you wouldn’t believe how many people don’t completely grok open source development. I’m absolutely sure you’ll see the very same in just about every company developing proprietary software, including IBM. But that’s the fun part, and certainly makes my life more interesting.
Thanks for the feedback Dan!
Despite what others have written, I’m *delighted* to see Sun dropping the project names for Java SE 6 and 7 – it shows a great intention to actually be a community from the start, rather than keeping the clique project names around. Also good to see the feedback forums that Tom blogged about, and I hope it will prove to be a good channel back to what we should be doing though I have my doubts of making it a very public forum of casual readers, rather than a public mailing list of ‘vested interests’. Good to see the outreach and the intention to do the right thing from the very start – fingers crossed for a bright future.
Rob pushed me into giving a short presentation at the Wellington LUG the other night. While the turnout was pretty small, all peering over a line of monitors in a gaming facility, I think my talk went okay. I gave a short presentation about OpenSolaris, along with my slides from my OSCON lightning talk.
What I really wanted to push home was that working within a wider community can be an amazing amount of fun. Too often I see LUGs just keep to themselves, help others in the mailing list or through meetings, but never actually contribute to the upstream community – having the chance to work with absolutely amazing people, on amazing technology and being able to figure out what skills you can personally contribute. Not sure if that message got through – who knows.
So finally decided to join the Portland mailing list today, with gentle prodding by Joe. For a long time I’ve agreed in principle with what Portland project is trying to achieve, though wildly disappointed with the implementation (I still am). Which is why I was delighted to read Waldo’s mail announcing the DTL Technical Workgroup conference call. From my point of view, the specs are definitely where we should be spending all our availalbe time right now – they’re often hard to read, ambigious, and have a weird mix of specification and implementation. They also suffer from the fact that it’s not entirely clear what commitment there is for the various interfaces they export.
What is worrying is the intense desire to have something out yesterday (akin to what’s happening in the LSB too), with the current approach of using a bunch of scripts seeming like a train crash waiting to happen – if you don’t believe me, ask one of the guys who’ve sat for years on a Sun ARC committee on how easy it is to maintain compatibility with them. In my previous experience in Sun, the desktop team have taken 5+ years to slowly mature the various set of GNOME interfaces – even now, there’s only a relatively small set that we’re actually committed to. Does that hurt our ISVs? Perhaps, though I think constantly changing them will hurt them more.
It’s been a rough day at Sun for many, many people. While many say it’s a good opportunity to do other things, for me, well, Sun always feels like a family [most of the time], so it really fucking sucks losing people – although I obviously understand the business needs of the company to have to resort to this.
For all the absolutely amazing people who got laid off today – it was a horrifying shock to the system, you rocked my world, and it was a pleasure to have worked with you. All the best for whatever you get up to in the future and I hope to continue to hear from you.
It’s times like this where I feel lucky, exceptionally lucky to be continuing to do the job I do, and have the opportunities that come with that job. It’s a suitably motivating factor for getting Sun (and SUNW) where it should be, and trust me, that will happen soon.
Be excellent to each other!