that’s whacked

Indiana, and mixed messaging

Monday, Jun 4. 2007  –  Category: OpenSolaris

Indiana

I’m a little late to the Indiana party, having just recently returned from vacation, and it’s taken me a few days to try and parse the nuclear explosion of a thread that is still spawning mail on opensolaris-discuss, but here are my thoughts so far (for the two people who have asked for my opinion – ahh the beauty of blogging)

I support Project Indiana’s goals, and I think the idea of having an alternate distribution to tackle those goals is fine. If Solaris marketing wants to develop this distribution, more power to them. If Ian Murdock is to be the ‘sole arbiter’ of contentious decisions, then that’s fine too – that’s nothing more than someone filling the role of tech/project lead.

In previous postings, I have supported the idea of a reference distribution: a distribution that takes all of the various consolidations currently available on opensolaris.org and bundles them together into an installable image. I.e.: Solaris Express without the encumbered and non-redistributable components. Thinking about it more, this only worked because (for the most part) we had non-competing technologies; i.e.: no competitor to ON, no competitor to Xorg, no competitor to JDS. This isn’t true anymore (the most obvious examples being the KDE and XFCE projects happening on opensolaris.org). Putting my Sun employee hat on, I still think it’d be nice to have a non-encumbered and freely redistributable Solaris Express if only to more easily evangelise Sun’s distribution of OpenSolaris; but I think the idea of a reference distribution quickly falls apart as more diversity occurs at opensolaris.org.

I don’t support Indiana being called “OpenSolaris”, or being deemed the “reference distribution”, for a few reasons:

1. It’s unfair (anti-competitive?) to other non-Sun distributions given that Sun owns the trademark “OpenSolaris” &tm;. They aren’t allowed to use it in their distributions’ names (without Sun’s express written consent).
2. Sun calling their distribution “OpenSolaris” puts the non-Sun distributions at a disadvantage since they will forever be trying to come up with messaging explaining why their distribution is OpenSolaris, but not “OpenSolaris” &tm;.
3. It’s confusing messaging. For the past 2 years, we’ve emphasised that OpenSolaris is not a binary installable. It’s a collection of source code from which other people/organisations/companies derive distributions which people can install. This leads to part 2 of my blog post which I’ll post after I’m done with Indiana.

So, there it is. I think the goals of Project Indiana are noble and good. I think the project should be approved, and I would love to see progress happen on it. I don’t think it should be allowed to co-opt the OpenSolaris community though, and I don’t believe Indiana should be treated preferentially or differently from other OpenSolaris-derived distributions (including Solaris Nevada/Express).

Mixed Messaging

As I noted earlier, I think delivering a distribution called “OpenSolaris” is a mistake in part because it further confuses the message we’ve been trying to deliver: OpenSolaris is a technology source code base. But in truth, we’ve been delivering mixed messages for a while now. Both Sun executives and engineers have been overloading the term OpenSolaris to refer to so many things now, it’s no wonder users are confused. I don’t envy the job Sun marketing folks have in trying to distill the “What is OpenSolaris” message succintly.

So let’s all agree (for the moment) that the messaging is confusing to users. Do we make an about face and redefine what OpenSolaris is in order to clarify? Or do we crack down and try to enforce the correct definition and educate users? In order to educate users, we’d need to educate people internally – and well, that’s hard… so I can appreciate marketing wanting to just redefine what OpenSolaris is.

Hopefully Glynn won’t mind me citing him here, but he brought up a good counterpoint: Debian has the word “Debian” trademarked, but that hasn’t stopped Ubuntu from “building a better Debian”. It’s a good argument, and one that I haven’t finalised my thoughts on. I’d like to hear more opinions before I do, so if you have thoughts to offer, please do so here.