Editor's Note: When Life Hands You Lemons...Sep 01, 2006, 22:30 (4 Talkback[s])
(Other stories by Brian Proffitt)
How You Can Improve Your Machine Learning with Spark REGISTER >
By Brian Proffitt
It was, without a doubt, one of those weeks. You know the kind, when everything seems all planned out and managed and then Life reaches out and smacks you on the head and reminds you about the folly of thinking you can actually manage It.
The first smack came when I decided to implement a software fix a friend of mine sent me after he lavishly described his "eye-popping" font configuration tweak. He sent me the instructions to improve the anti-alias settings from an article he's writing for Linux Magazine, and I was eager to try them. After making the settings changes, I did something I rarely do: I restarted X.
Oh, did I mention I was using Kubuntu at the time, and that this
was actually happening on August 22? Before we all heard about the
xorg-server break. Needless to say, when I got the "No screens
found" error message, I was instantly sure it was my friend's
fault. After about an hour of trying to reconfigure X, and having
no luck--which completely blew my mind at the time, because I was
sure I'd killed the font changes--I decided to chuck it and
reinstall the Ubuntu packages. (Made easier by the fact my
Oh, did I mention I'm moving soon, and most of my software is in boxes? All I had readily on hand were the giveaways from LinuxWorld: specifically, Fedora Core 5 and Freespire 1.0. I've used Fedora before, so I thought I'd try Freespire and at least get a new software review out of it. Since there will be a review very soon, I won't give away the details... but I found myself wishing I'd gone with FC5. Especially when, on the morning of the 23rd, I learned this was a universal Ubuntu problem and that I could have made a very quick fix and recovered my Kubuntu machine in mere minutes. (Good timing there, Ubuntu crew.)
By Friday last week, I thought I'd managed to catch back up with my grand plans. I'd sent a few questions to one Linus Torvalds for the fifteenth anniversary article I wrote last week. From his last kernel announcement, I knew he was off on vacation, and I figured I could write the article without his direct input if I had to. Which I did. And I think it was rather good, save for one teensy, tiny point. According to Linus' reply, which I received the day after my last Editor's Note, there are some questions about whether August 25, 1991 is actually the Official Birthdate of That Which is Linux.
Now, I try to be an honest fellow and own up to my errors, and this will be no exception. Here was Linus' reply: "A lot of people consider the birthday to be October 5th, since that's when the first announcement with actual pointers to source went out (version 0.02, since 0.01 was never really announced)."
Okay, was August 25, which is the date of Linus' first message about Linux to the comp.os.minux newsgroup even close? Sort of.
"Others consider it to be September 17th, which is when the unannounced 0.01 version was distributed (there's no announcement, but you can see it in the tarball that is still available: I think the last date stamp in the tarball is 9:30AM on September 17th, 1991)," Linus wrote.
So, we have three possibilities: August 25, September 17, or October 5. Which one does Linus holds as the real anniversary date?
"Me, I don't know. The September 17th date seems to be the most correct one, although since I only told people in private about it, and since I don't have any archives of my own emails, it doesn't have any good announcement to go with it," he replied.
I'm still keen on the August date, mind you, since the first announcement seems an auspicious way to mark the beginning of the project that would become the Linux kernel. But, I figure it's his project, Linus can set the beginning wherever he darn well pleases. Who knows? Perhaps this discussion will be elevated to the great vi vs. emacs flame wars, and I will be forever blamed for starting the argument.
In seriousness, the primary reason I'd contacted Linus was to ask him a question that's been on my mind for quite some time. 15 years is quite a while for anyone to be on one project, and it made me wonder, is there ever going to be an end point for Linus' involvement in the Linux kernel?
Clearly, he's made no mention of retirement, and I am not wishing it, but milestones like this do tend to make people think about the past and the future, so I put the question to him: "15 years is quite a haul, so where do you personally see yourself in the future development of Linux? Do you ever imagine a time when you will transition the kernel to others, or is this your baby from beginning to end?"
Here's what he had to say:
"Well, I obviously have transferred a lot to others. I've always fairly actively tried to push out the work that I'm not personally interested in to others, and the end result means that not only am I personally involved with just the kernel (with all the rest of the infrastructure being done by others), but even when it comes to the kernel, I try to make pretty damn sure that I'm not in the way of others.
"In fact, over the last few years, I've even tried to make sure that the whole 'process' (which used to be very centered around me personally) is more distributed. That was brought on by some real concerns about five years ago about me not scaling in accepting patches, and the use of Bitkeeper to allow the whole development process to be much more distributed.
"So I'm still very much involved, but at the same time, as far as I'm concerned, the whole point of open source is that nobody really controls it. I would absolutely hate it if people saw it as purely my baby from beginning to end, and I don't think they do. The thing is very much set up so that anybody else can take over maintenance not just from a legal standpoint, but also from a tools/infrastructure standpoint, and I'm around as somebody who most people trust, even if they don't necessarily always agree with me."
So, in the end, a renewed lesson on the strengths of open source. A pretty good ending to a pretty strange week.
Next week: building a Linux machine from cardboard boxes, duct tape, and a permanent marker.
0 Talkback[s] (click to add your comment)