It's standard operating procedure for corporate PR departments to regularly tell the world that X or Y feature will solve world hunger, establish world peace, etc. It's less standard for the tech press to slavishly repeat such press releases and uncritically treat them as canonical truth. Or, at least, it should be.
Take the case of Canonical's recent pronouncement that it has ended decades of dissonance between competing Linux package management solutions. The lack of thoughtful scrutiny of the claims by the tech press beggars belief. Fortunately, a swelling chorus of critics is rising to put the claims in context, separating the wheat from the chaff in Canonical's attempts to unify Linux distributions.
It's all good
Maybe this whole thing isn't that surprising. As one open source insider confided to me, "There is literally zero critical analysis of so much that comes out of open source land." The reason, Apache Software Foundation board member Jim Jagielski adds, is that "by adding the magic buzzwords 'open source' and 'community' to any offering, people will assume it's true."
In other words, because intentions are presumably good, "the right thing to do" is simply to regurgitate open source PR schlock.
In the case of Canonical's press blitz, who doesn't want to believe that the packaging wars have come to an end? That peace, love, and a unified Linux community has been achieved in our time? And so we see headlines like these:
- Ubuntu bids to eliminate Linux fragmentation by making Snap packages available to all (So generous!)
- Goodbye rpm and deb. Hello Snaps! (Don't let the door hit you on the way out!)
- Ubuntu Snappy-Based Package Format Aims to Bridge Linux Divide (I weep with joy!)
Much of the "analysis" then goes on to parrot Canonical chief Mark Shuttleworth's claim that "different communities and developers started asking Canonical if they can port [Snaps] to their distribution." The problem, however, is that it's not true. At least, not in the way it's being presented.
Or maybe it's not all good
Perhaps the most persuasive counter to the Canonical PR machine comes from Adam Williamson. Though hardly unbiased -- Williamson works for Red Hat and is part of the Fedora development team -- Williamson actually represents the very communities that Canonical claims can't get on board the Snaps train fast enough.
Yet here is Williamson's assessment of the news: "The press release and the stories together give you the strong impression that this thing called Snappy is going to be the cross-distribution future of application delivery, and it's all ready for use today and lots of major distributions are buying into it." Unfortunately, he goes on, "This is, to put it diplomatically, a heaping pile of steaming b------t."
Well, that doesn't sound promising.
One problem with Snappy, Williamson continues, is that for all its "unification," it's actually a single-vendor project. Every contributor works for Canonical and any outside contributions require assignation of rights to Canonical. In other words, this "collaboration" between diverse distributions is really a one-way highway into Canonical, a Hotel California of code contributions.
That's why Canonical's hints of broad community support for Snappy are suspect. According to Williamson, "The sum total of communication between Canonical and Fedora before ... this press release was that they mailed us asking about the process of packaging snappy for Fedora, and we told them about the main packaging process and COPR. They certainly did not in any way inform Fedora that they were going to send out a press release strongly implying that Fedora, along with every other distro in the world, was now a happy traveler on the Snappy bandwagon."
Oh, and to make matters worse, "the server end (the 'app store' bit of the equation) is closed source, and Canonical have been refusing to tell anyone how to run their own 'app store'." So, in essence, Snappy is designed to make Canonical the center of this happy new universe, with all roads (and snaps) leading to Canonical.
Is it even necessary?
Even if we discount all the self-serving PR guff that Canonical has spewed with this release, there's still an open question as to whether the effort is even worthwhile. The short answer is "probably not," as Kyle Keen of Arch Linux details.
Solving the technical problem of unifying different package management solutions isn't actually very interesting, because it completely misses the point: "Distros exist because users can't agree on how exactly we want software. A universal package will not be good enough for some percentage of them." In other words, if the distributions wanted to have everything the same, that would have already happened years ago. A common package manager isn't going to solve much of anything, particularly one so heavily controlled by one vendor.
This is an intractable problem and certainly not one that is aided by the chimera of some grand bargain between distributions about package management. Canonical has not done the hard work of resolving cultural differences between distributions, thinking that a package management is purely a technical issue. It's not, and even if it were, the company would still need to do a heck of a lot more lobbying to garner support for its approach, rather than press releases that pretend at support that doesn't actually exist.
But this isn't really Canonical's fault. It's doing what most companies do: painting a big picture and hoping enough partners/customers will believe.
No, really the problem is with a fawning tech press that accepts anything that sounds like "open source community" as gospel truth, when a more critical analysis is needed.