Several people have written in to ask what I make of the recent Mobily editorial, in light of my coverage of Debian and Ubuntu previously (I made, in fact, a very similar argument back in May – and here‘s me arguing on behalf of Debian in December of 2004). Mobily’s piece, in case you haven’t yet read it, makes the simple but persuasive argument that by more or less abandoning its desktop efforts, Red Hat has unwittingly opened the door for Ubuntu to come in and eat its lunch. It has spawned an interesting discussion, from David Berlind’s thoughtul piece here to Nik Cubrilovic’s dissenting view here to Ubuntu’s Supreme and Benevolent Dictator for Life Mark Shuttleworth’s polite and humble disavowal of the remarks.
Where does one fall, is the question. Considering the extremes – Mobily’s Ubuntu is the bane of Red Hat’s existence to Shuttleworth’s Mickos-like diplomacy – who’s to be believed? Well, I tend to think that they’re both right. And they’re both wrong. Let me explain.
Mobily’s case is simple, and easily defensible because of that:
Now, the key sentence: I became a user of Red Hat Linux for my desktop machine (and yes, it was a bit of a challenge!), and a couple of months later, when I had to choose what distribution I should use for my server, I chose the one I was most accustomed to: Red Hat Linux.
The desktop, in other words, was the original entrypoint for the server and was where Red Hat initially got its footing. I would more or less agree with that statement. Where I break with Mobily is his assertion that “by abandoning their desktop users, Red Hat has effectively shot itself in the foot.” As Shuttleworth notes, and I’ve discussed in our billion dollar debates, Red Hat’s built itself a strong and sustainable business that is no longer dependent (or even related, some would argue) to the desktop. And more to the point, its entrenchment within the enteprise affords it an inertia that will be difficult for any player – Ubuntu, Solaris, whomever – to overcome. The case could be made, incidentally, that the real risk from Ubuntu is not to Red Hat but Novell/SuSE, but that’s an argument for another time and post.
I don’t totally disagree with Mobily, however – I think the argument simply needs some nuance and subtlety. Rather than contend that what gets deployed is what’s used on the desktop, which is an argument I can’t convince myself of, I’d simply point to the desktop’s role in building community visibility and momentum. When I built the case that Ubuntu would emerge as a credible enterprise Linux player, I did so on the basis of its intrinsic advantages from a community perspective. Just as Red Hat enjoys certain advantages financial and otherwise over Ubuntu, so too does Ubuntu successfully differentiate itself from Red Hat in its package management architecture and the community behind it. One reason for that community strength is indeed the desktop; because it’s usable, because it’s polished, it sits in front of a lot of developers. Developers that would like to improve that experience even further, and contribute time and effort in packaging. Those community packaging advantages will, IMO, emerge as a true differentiator for Ubuntu as it attempts to make inroads into the enterprise.
As for the first South African in space, Mark’s comments were almost to the letter what I expected to hear. Like MySQL’s Marten Mickos, Shuttleworth can play the role of diplomat when it suits him – and in this case it’s the right move. Embracing Mobily’s comments would be premature at this point, because candidly Ubuntu is not (yet) in a position to overtly challenge the enterprise strength of Red Hat, hence the Kung Fu style humility. Does that mean it never will be? No indeed. And I think Mark’s contention that Ubuntu’s goals are so different than Red Hat’s are belied by things like this. As pirast, one of the commenters on Mark’s post, reminds us, Linus was similarly humble when in his business ambitions: “Really, I’m not out to destroy Microsoft. That will just be a completely unintentional side effect”
The comparison to Mickos is apt (no pun intended) in another way as well: I fully expect to see Ubuntu embrace a MySQL-like approach to the enterprise market. In baseball terms, it would be described as taking what the pitcher gives you. If you get a pitch on the outside part of the plate and try to pull it, odds are good that you’re going to ground to short. Take the pitch the other way, however, and you might hit anything from a bloop single to a towering home run. MySQL’s great at this, actively embracing the Corolla to Oracle’s 747 comparisons, and going after a (volume) market that the big guys have never had any interest in. Ubuntu, I think, has a similar opportunity in front of it. While Red Hat and SuSE are focused on what my colleague likes to call the Charles Schwab/JP Morgan problem – where vendors treat every customer as if they were a Fortune 50 business – Ubuntu might just slip in through the back door and solve the problems for the rest of us. The margins won’t be as good, but there are a lot more non-JP Morgans than there are JP Morgans.
In other words, I think that Mobily’s general point does have merit, even if the supporting evidence is not the desktop argument originally proferred. His argument, however, need to be tempered with a dose of Red Hat reality, which Shuttleworth is happy to provide, saying “my own view is that Red Hat will continue to do well in the specific areas that they have targeted.” I think he’s right, and the really interesting question is whether the specific areas they have not targeted will end up being more important.