Sean's Technological Journeys

Dear Jim,

First, thanks for your thoughtful trackback. I am glad that my blog garnered some attention from Sun employees; I wasn't really expecting that! For those of you who haven't been following the discussion, the original post on my blog is here, and Jim's reply on his blog is here.

I would like to hear a bit more explanation on the title of your post, "No Coup Needed". I'd like to hear your side of the story. My blog post was intended to demonstrate how, with a little work, OpenSolaris can really meet peoples' demands of a desktop OS. I was simultaneously trying to cheer on the community to make the amount of work (on the user's behalf) less and less, while trying to point out where it currently falls short, in the hopes of getting OpenSolaris contributors to focus on making those issues more seamless. If all of these goals were fully realized, therefore, OpenSolaris would be in position to peacefully convert Linux users to OpenSolaris, thereby causing a "coup" of sorts (but without armies and assassinations and so on). My title, though, was really asking a question to Sun, and to the OpenSolaris community at large.

What I'd like to know is, why does Sun expend man hours on OpenSolaris features which directly benefit desktop end-users, but not servers and the like? I have always been curious about this, because it seems that OpenSolaris reaps the most revenue for Sun by being deployed on large servers (and those servers might also conveniently run Sun hardware). I may not be seeing the entire picture of how trying to catch up to Linux usability on the desktop is directly benefiting Sun.

On the other hand, of course, I recognize that there *is* an OpenSolaris community out there, which consists of non-Sun employees who have their own agendas; and they may be contributing stuff that benefits desktop users. This is a nice way to go, since Sun only has to accept their contributions, which doesn't take nearly as many man hours.

If by "No Coup Needed" you meant that Sun just doesn't need to _try_ to overtake Linux, but it will do so eventually in time, I think that's also a distinct possibility. At this point, it's a simple matter of calculus: if the rate of change measuring the growth of the Linux community is faster than the growth of the OpenSolaris community, then OpenSolaris will likely have fewer features, fewer testers, fewer contributors, and fewer well-integrated FOSS packages than Linux distributions. If, on the other hand, velocity is in OpenSolaris' favor, then we need only keep doing what we are currently doing, and in time, OpenSolaris will have more desktop adoption than Linux distributions.

Of course, it's also nice to recognize that in the Free Software world, there is no need for aggressive competition in the sense of attempting to extinguish a competitor, or reduce them to obsolescence. Instead, simply trying to make your community as strong as possible, while coexisting peacefully with competitors, seems to be the common theme. Linux _is_ a competitor because the kernel is inextricably different between OpenSolaris and Linux; furthermore, any user of a Free Software operating system must choose between Linux and Solaris. So in that sense, since Linux and Solaris serve the same function as being the core of the OS, one might argue that it is important to seek out and retain new community members, even at the expense of converting them away from the other alternatives. But the GNU userland, I believe, is a potential powerful ally, if properly adopted into OpenSolaris.

At this point in time, I think I will remain neutral and sit on the fence. I really don't know which interpretation of the future of Linux and OpenSolaris is the most likely. I run both Linux and OpenSolaris, and I will not disregard either community as being obsolete, inferior, or defunct in any way. I have tried OpenSolaris distributions in the past, and I did not consider it to be a viable alternative for a desktop OS; however, since I tried 2008.11, I have revised that opinion. We will see what happens in the future. I hope that we will soon have choice for every aspect of the Free Software stack: until OpenSolaris, there is an alternative for every component except for the kernel. Now there are no single-source components of the system, and the alternative -- OpenSolaris -- is shaping up quickly.

Thanks,

Sean