Slashdot Mirror
Dunc-Tank To Help Meet Debian Etch Deadline
Da Massive writes, "Debian GNU/Linux is experimenting with a new project called Dunc-Tank, which is aimed at securing funding to pay two key release managers — Steve Langasek and Andi Barth — in an effort to ensure the forthcoming Debian 4.0, known as etch, is released on time in December." Dunc-Tank is not affiliated with the Debian Project directly, and in fact was controversial on the debian-private list.
Basically, Dunc is an experimental project to try out ways of funding Debian development. Not paying for servers or bandwidth, or reimbursing expenses and flight costs, but actually paying people to sit down and do useful Debian work rather than some other day job.
Who is Dunc?There's info about who exactly is behind Dunc at the board page.
Dunc directly supports work on Debian, and is made up of a small group of people who use Debian and who want to see Debian improve. But Dunc is not endorsed by Debian, and Debian does not exercise any control over how Dunc operates.
What about other people funding Debian work?A number of other groups fund Debian work directly or indirectly, whether that be by allowing or encouraging their employees to contribute to Debian, or having Debian work be part of their actual job description. Dunc does not aim to compete with those groups, either in the tasks being worked on, or in the people being recruited, but rather to address other niches in the Debian ecosystem.
What does "Dunc" mean?Dunc is an acronym standing for "Development Under Numismatic Control" -- which could equally be called "coin-operated coding". The point of the project is to try some new possibilities of funding free and open source software development and helping people work on free software development on a full-time basis.
Really, though, the name is a reference to the linux.conf.au auction in 2003, for the t-shirt signed by the speakers, proceeds from which were directed to Electronic Frontiers Australia. To make the bidding more lively a certain individual foolishly suggested that the next Debian release would be named after the winning bidder, should the bidding go above $2000. Due to the combined resources of a table of inebriated Sun folks, Duncan Bennet won the bidding, and the right to have his name associated with the next Debian release -- which, many years later, turns out to be Debian 4.0, aka etch. So yes, this is yet another free software project that has its roots in the consumption of a little too much wine at a conference dinner.
What will the future bring?As Dunc is an experiment, we don't know what will end up happening with it. We may decide it works perfectly as is, or that it was a horrible idea that should never have been tried. In any event, we expect to review what worked, what didn't, and what should be done over the course of the first project, and have a public discussion about what to do after the release of etch.
Random factoidThis site is maintained using Joey Hess's ikiwiki.
It is licensed under the terms of the GNU General Public License, version 2.
Links: index Last edited Tue Sep 19 13:20:35 2006Dunc-Tank.
Python coder | PyQt Applications | Writer
I am sure paying these folk will get this Debian release out on time. Money always motivates people to work harder and faster.. and it always makes sure a deadline will be met...
i know the odds of actually winning is very small...
At a physics talk I heard a speaker explain how the chances of winning the lottery are smaller than the chance that a neutron star will pass through our solar sytem close enough to us to capture the earth and pull us away from the sun, thus dooming us to a dark and brief future.
I play sometimes too...
Personally I think things are perfect the way they are in the Debian and Ubuntu Camps. I use both, Ubuntu on my desktop and Debian on my servers. They both have their place and they both do their jobs well. If anything the Ubuntu people should have called themselves Desktop Debian or something. Debian developers shouldn't hang it up. They are already working on Ubuntu directly. Ubuntu just takes their work and adds a little to it.
Beer! It's what's for breakfast!
After reading the slashdot writeup and the linked page, I still don't know what they're doing. I know they're trying something new for funding, and I know how it got its name, but I still don't know what new thing it is they're trying. If it was in there, it got buried under a mass of other less important details.
I think it means they are going to raise some funds (how?) to pay some developers directly to work exclusively for some time on Etch. But if so, that's not exactly innovative; other projects have done so before (Perl foundation grants, as one of many examples), and I'm surprised Debian hasn't.
Secession is the right of all sentient beings.
Buying a single ticket is logically worth it
Depends on what you mean by "logically", I guess. Buying a ticket is logically worth it if a chance to rub your ticket and dream about what you'd do with all that money is worth $1 to you. It's generally not worth it if you calculate the stastical expected value of the ticket (odds of winning multiplied by the probable size of the pot). The expected value of a ticket is nearly always far less than its price, for obvious reasons.
However, there's another angle that may make it more "logical". The argument is that wasting $1 will have essentially zero impact on your quality of life. You could throw a $1 bill in the road every week and probably never notice any change. OTOH, winning the lottery would change your life immensely. So, if you can measure the values in terms of life impact, the $1 ticket has a near-zero cost, and the value of winning is huge, so perhaps the expected value of the ticket exceeds its certain cost.
Personally, I think that's crap. The value of $1 per week, if saved and invested over years, is decidedly non-trivial, so the argument can only work if you accept as a given your inability to save money.
It's the people who buy more than $1 that I can't really understand.
Why? I see three reasons to buy lottery tickets:
First, because the statistical expected value of the ticket exceeds its cost. This can happen when the pots get really big. In that case, there are lots of appropriate strategies, depending upon your available funds, but pretty much all of them would suggest buying more than one ticket.
Second, because you buy the low-impact/high-reward argument, and your perceived expected value is higher than your perceived cost. In that case the logical approach is to buy as many lottery tickets as will have no discernable effect on your finances. Perhaps you make and spend exactly the right amount of money so that $1 is that threshold, but it's more likely that the painless amount is higher or lower.
Third, because you like to dream, and like the excitement of having a chance. In that case, since you're basically spending money for entertainment, you just have to decide how many tickets give you the best bang for your buck. Most people would probably say one ticket, but a book of 50 really makes a nice slapping sound in your hand, so YMMV.
Personally, I can dream about what I'd do with millions of dollars without buying a ticket at all, so that's my strategy. Sure, I have no chance of winning the millions, but buying a ticket only changes that by a miniscule amount, and the buck can sit in the bank and gather interest.
Note to ACs: I usually delete AC replies without reading them. If you want to talk to me, log in.
I've been using Debian for quite some time, and to me the point of Debian is stability. I couldn't give a rat's ass when etch releases as long as it works the way Debian is meant to. Please don't force the release to meet a deadline. You'll only be hurting the users that depend on Debian to be a stable and functional system.
Personally I think the Debian people should just hang it up and start working on Ubuntu.
The Debian people are working on Ubuntu. And Knoppix. And <insert favorite Debian-derived distro here>.
If the *thousands* of Debian developers stopped doing what they're doing, Ubuntu would grind to a halt. So far Mark Shuttleworth has spent nearly $20M on Ubuntu, and all of that money has accomplished relatively little, from a purely technical perspective. What's great about Canonical's efforts is that what they've done has been focused on polishing the bits needed to make the non-developer's user experience better -- the bits that many (not all) developers tend to be less interested in.
Ubuntu isn't structured to manage the participation of thousands of active developers working on a dozen platforms. If Ubuntu were to restructure to meet that goal, (1) growing pains would cripple the project for a good long while and (2) the result would look a lot like Debian.
IMO, the status quo is better. Ubuntu takes a raw diamond and cuts and polishes it while Debian is busy squeezing carbon deposits into diamond.
BTW, I'm a Debian user, and a software developer, but not a Debian developer. I've tried Ubuntu a couple of times, but always found it to be lacking in software packages I need. I can pull those packages from Debian, of course, but there are always little issues with that, so I find it easier to stick with pure Debian (sid on my desktop, testing on my family's desktops and stable on my servers).
Note to ACs: I usually delete AC replies without reading them. If you want to talk to me, log in.
I use Debian for ? 11 years or something and have worked as a Debian sysadmin. ...
.conf file to get it up and running - apt did everything for me.
Again IMHO, using Debian involves an unnecessary amount of effort just to get it running, simply because the mentality is that the system must be shiped naked with no configuration choices made
Wow. I've only been a Debian user for 5 years, some of that time as an admin.
I use Debian precisely because it saves me from having to configure stuff. I had a need to start developing with Zope the other day, I didn't have to edit a single
Packages like gforge are utterly remarkable. The debian apt installer scripts ask you basic questions and will setup/configure/create (dbs, tables, domains, user accounts, etc) for every dependency: OpenLDAP, PostgreSQL, Exim, Apache... a fully working and configured system without touching a single configuration file.
For admin/dev stuff, Debian is remarkable. For the Desktop, you have to do some fiddling to get the final 1% functionality, but that's no problem for me. If a noob doesn't want to remember to install nm-applet to get a nice WLAN configuration applet, or which xmodmap commands will make his/her multimedia keys work - then Ubuntu is great.