Code names and other coelacanths
Off the Beat: Bruce Byfield's Blog
I'm probably going to be answered as though I were a hybrid of Ebenezer Scrooge and Darth Vader, but can we quit with the code names for Linux distributions, already? People are taking them way too seriously.
Code names make sense if you want to keep what you're doing a secret. If you're planning a military operation, you probably don't want anyone to know until you actually hit the beaches of France. Or maybe in the case of the revised Doctor Who, when you're a producer who doesn't want the excitement to peak too soon (in which case, you pass around sheets of paper watermarked with "Not to be copied" and call what you're doing "Torchwood," then decide the name would make a great spinoff series).
But in this community? What part of "free and open source software" don't you understand?
I realize, of course, that code names have long since become one of those geeky in-jokes, like First Posts on Slashdot used to be. As such, they're not supposed to be useful. They don't even have to be particularly funny. After all, the point is that appreciating the code name shows that you are a member of the pack.
In fact, in the past, code names were frequently never made public. For example, over a decade ago, when I was on the founding team for the now-defunct Stormix Technologies, the code names for distro releases (rain and hail, if I remember correctly) were chosen not only to go with the company name, but as a company in-joke, referring obliquely to the fact that the city of Vancouver is located in the middle of what used to be a rain forest. Few people outside the company even heard the code names. They weren't supposed to.
The appendix of projects
These days, though, code names in free software are like screen savers. They're no longer needed, but survive like a coelacanth long past their heyday because people expect them. You can tell that they've outlived their time by the fact that they're now announced to the world long before the release is even in alpha. So far as they have any purpose at all, they exist mainly to suggest that project contributors are human and approachable.
An even surer proof that code names have turned vestigial is that they're become ironic -- or, as I prefer to say, just plain silly. That's why the next release of Fedora is code-named Beefy Miracle, while Ubuntu 12.04 will be Precise Pangolin (and I shudder thinking what awaits us with the release of Ubuntu 12.10, when the letter Q is reached. It's like being stuck in a surreal version of Sesame Street that never ends).
Not, as you might have noticed, that I have anything against silliness. But the silliest part is that silliness is being taken seriously.
Modern code names aren't just catchphrases dreamed up in the middle of a release to relieve the pressure of pushing deadline. Instead, they're occupying large chunks of people's time. Mark Shuttleworth, whose decision can make the whole of Ubuntu change directions on a Euro and do a triple axel, actually spends an entire blog entry going over the possible names for the next release. Over in the Fedora community, a whole campaign centered on getting Beefy Miracle accepted as the code name over some strenuous objections.
For that matter, here I am, writing about code names as well.
Of course, many people are being light-hearted about the whole idea. Shuttleworth is alliterating through his blog for all he's worth, and my colleague Larry Cafiero producing a Beefy Miracle filk.
On Facebook, distro critic Jef Spaleta is even laughing at the competition between Fedora and Ubuntu, to say nothing of himself, arguing that, Fedora is "trying to outdo the silliness of one man who imposes his silliness via the power of the purse. Though to give credit where credit is due Shuttleworth is impressively silly."
Just what we need -- competitive silliness. Like the competitive potlatching of the First Nations in the Pacific Northwet, it's a good idea gone mad.
But these examples fade to insignificance next to the people who are arguing over the appropriateness of Beefy Miracle. Like a class of English grad students, they are deconstructing the five syllables, squeezing every last milliliter of meaning from them. Does "miracle" suggest something rare? How will Hindus react to "beefy"?
Oh, and was the voting fixed? Was a mysterious cabal of insiders involved? I'm not sure which is worst: that people should believe such things, or that they actually happened. But in the last few days, such all-important matters have actually become one of the major threads on the Fedora user list. It's definitely become one of the angriest.
On the whole, I prefer the silly ones to the deconstructionists and conspiracy theorists. But all these groups have forgotten the salient feature of code names: They don't matter.
Certainly not for very long. Offhand, who can remember the code names for Fedora 8 or 11? The alliterative Ubuntu names are easier to figure out if you count on your fingers, but even they are very much of their moment, no matter how widely used they are in their time. Very quickly, they cease to matter.
Code names have never been especially good jokes, regardless of whether they're private or public. They're strained at best. But with people spending so much time on them, it's time to call a halt. Why doesn't everyone stop wasting time and do something that matters? I know crunch mode needs the occasional relief, and the end of a project deserves a little craziness, but if a diversion is needed, can we at least find a new one?
Linux namingHow about making Linux simpler for non-geeks by using some meaningful names of things instead of things like gnome, kde, xfce, nautilus, kate, vi, yum? Wouldn't Linux be more widely adopted if we hadn't built this techie only club of code names for nearly everything?
Re: Let's see...Meta comments don't count. Otherwise, you'd have caught me out
who cares?If you spend too much time thinking about code names, then you need something else in your life. If a code name is silly, who cares? If a code name is way off base, who is to say that it is way off base?
There are bigger problems in life than code names.
Let's see...Let's see, we do the codenames for publicity, and you wrote this article, yes?
So I'm thinking 'no'.
code names and confusionI am an Ubuntu user and I can not keep up with the code names. I know I am using version 10,04LTS on my net book and 10.10 on my laptop, now ask me the code names and I couldn't tell you. I guess I would have updated to either version 11.04 or 11.10 but I am not sure I like Unity. I found that 10.10 Unity crashed on my laptop, and it crashed a lot, so I nixed that version. In any case I have lost track of which version has which code name. I thought that Canical did real well in naming their Ubuntu versions by the year and month and didn't really need code names,
code namesAbout the only useful thing I've found about codenames comes in Debian when what was "testing" is about to become "stable." If one doesn't want to get hit with a large number of sudden unwanted updates, one can check the enabled repositories and change, for example, "testing" to "Wheezy."
Re:Re:You failed to state that there's only one Linux kernel and hundreds of distros. Also it's not really that important as users should use the latest versions of their software if they want them to be secure, stable etc.
Re: two words: google index"Outside of the simple fact that you are wrong"
How I love the cut and thrust of intelligent debate!
I might argue that the release number is far more reliable when you're looking something up than a name, and that the Linux kernel seems to get by just fine without code names. But, seriously, it's not code names as such that I'm commenting on, but the obsession with them.
two words: google index.Outside of the simple fact that you are wrong, every project needs a name, and a release is a project; a name is much more indexable than a number. The sillier the name, the more unique the index, the better retrievable the needed information.
I love to be able to search for 'narwhal wifi problem' not to say i have any, but if I did, itll produce less false positives than '11.04 wifi problem'.
Humans need words, not numbers. Google too.
Customers can take a free test drive of SLES for HPC on the Azure Cloud
San Francisco-based chip company announces their first fully open source chip platform.
The whole distro gets rebuilt on glibc 2.3
Ubuntu Vendor tries to solve app packaging and distribution problem across distributions.
Founder of ownCloud launches the Nextcloud project.
Will The Machine change the way future programmers think about memory?
The new Torus distributed storage system is available under an open source license on GitHub
Juries decides Google’s use of Java APIs Was Fair Use
But if you are not using the latest Linux kernel, your system is insecure.
Home routers will give room for custom firmware but still comply with FCC rules