Slashdot Mirror
Conflicting Goals Create Tension in OSS Community
An anonymous reader writes "Mark Shuttleworth, of Ubuntu, has a post up meant to clear the air and clarify the project's place in the Debian community. He's specifically referring to comments made by Matthew Garrett earlier this month." From the post: "A little introspection is healthy, and Debian will benefit from the discussion. Matt is to be credited for his open commentary - a lesser person would simply have disengaged, quietly. I hope that Matt will in fact stay involved in Debian, either directly or through Ubuntu, because his talent and humour are both of enormous benefit to the project. I also hope that Debian developers will make better use of the work we do in Ubuntu, integrating relevant bits of it back into Debian so as to help uplift some of those other peaks - Xandros, Linspire, Maemo, Skolelinux and of course Etch."
Don't confuse debian with "The OSS Community". They are really not the same, and there is no such thing as "The OSS Community".
Conflicting Goals Create Tension in OSS Community
Yeah, anyone who's ever gotten even remotely involved in wikipedia could have seen this one coming a mile away. This is why, at work, you have "project managers", that have final say (and yet, also take the burden and responsibility of making decisions).
Push Button, Receive Bacon
So far the system has worked well for the end user.
And that is a good thing. Linux is moving along at a great pace, even with the little spats here and there. I love that there is alot of different ideas, with people to push them through.
I posted this first to Mark's blog, but I'll repeat it here:
This is a very well-written summation of the issues.
To paraphrase a comment from a message board I visit, "[Debian|Ubuntu] can't be everything to everyone."
Debian provides a wonderful base for many other distributions, not just Ubuntu, and it is a rock-solid platform for servers. It runs on many different architectures, and can be used on machines from a handheld up to a massive server. This is one of its greatest strengths, but also one of its greatest weaknesses.
Ubuntu, on the other hand, is far more focused than Debian is. Starting with the general base (the plateau, as Mark called it), it builds a strong distribution targeted to only 3-4 architectures (counting SPARC), which opens many more options. This is no different than many other distributions have done. For example, Knoppix is another version of Debian with customizations on top of it for a specific platform (or platforms).
Ubuntu can't be everything to everyone, because everyone has different needs and goals, and Ubuntu has a specific focus. Similarly, Debian can't be everything to everyone, because it is a more general distribution, a jack of all trades (and master of none).
IMHO, Gnu/Linux on the desktop still kinda sucks right now, but it is advancing rapidly. This makes me want to upgrade my distribution to get the latest and greatest, because it fixes features I really want (multimedia these days). We are quickly getting to a place where most the needs of average users will be well met. Then I won't mind if Debian is a little behind. It's like Windows XP being good enough that most people don't really care about upgrading to Vista. I can't wait until we are in that place, and I hope that then, the impedus to move forward so rapidly is lightened enough to relieve some of the stress on the Debian devs, allowing them more time to work through some of these issues.
The two opposite sides of the scale appeared in consecutive posts.
The thorough discussions apparently remove the risk of mistakes associated with conformity, called "groupspeak" by some consulting firms.
However, when all is said and done, the code for a function needs to be stable. At what point does the free-for-all become a liability?
*nix projects a somewhat splintered image. There is a group of users who are unhappy with the other two closed OS vendors, and are surveying the state of affairs. I at least am baffled trying to objectively rate all the variants out there. Does anyone know of a comprehensive feature chart that allows prospective users to scrutinize the specs for their favorite purpose across most of the builds?
We all know what MS is about. Apple's entire existence has been positioned as "the Friendly Branded OS". I have remarked that I will ease into one of the OSS builds. But which one? Red Hat? Debian? uBuntu? Xandros? When I go reseaching, who is a neutral source?
I am quite satisfied that we don't need Every Last User on OSS. There are net jokes about AOL users, and the stereotype exists for a reason. But for the midline user who wants to promote OSS, what if ALL the variants remain incomplete because of the flamewars?
My first Journal Entry ever, in 8 years! http://slashdot.org/journal/365947/aphelion-scifi-fantasy-horror-poetry-webzine
Project managers do a lot more than "review submitted code/content and decide whether it should be included in the production version of the product." They drive the process, not just filter it.
I would say that the line, "Ubuntu could not exist without Debian," is not an opinion. It is fact. Does this mean that Ubuntu owes anything to Debian? Not really. Other than the GPL obligations, there really isn't anything that Debian could or should ask for in return from Ubuntu.
Anybody who has worked with Debian already should have a deep and profound respect for the fact that Debian is plain and broad. When you sit down at a Debian computer, you are seated before a gateway to what might be the most customizable distribution in existence. All of the packages are roughly as far away as "$ sudo aptitude", and it is all but guaranteed that no matter how complicated or convoluted the package you want is, it will be downloaded and installed, along with dependencies, and you don't have to worry about a damn thing. (If you've ever compiled your own VLC or GIMP, you know what I'm talking about.)
The problem is that people would like to see specialization in Debian. Debian is not for specialization. It's for everybody to make what they want. Taking that away from Debian compromises the entire goal of the project...
~ C.
You take a straightforward, uncontroversial statement (Shuttleworth's blog entry) that practically everyone agrees with. Then you publish a headline saying there's a "conflict", and pretend there's a huge row going on.
Pretty soon you've got a heated argument going on, mostly between people who haven't read the statement that allegedly started it all.
What does it all prove? That Slashdot isn't "stuff that matters" any more, it's stuff that draws mass readership. Just what we were trying to get away from when we first started reading Slashdot ...
I read the blog post (yes, I know reading the actual articles is unusual for Slashdot, I'm sorry), and as I see it Shuttleworth's point is mostly that Debian should focus on the work on the unstable branch because, plainly, that's what working best.
Maybe he's right. Debian's never been succesful at meeting the scheduled release dates. If Ubuntu is capable of delivering better desktop releases, and in soon perhaps also better server releases, then what's the point of struggling and perpertually flamewaring to do the same inside Debian? Food for thought. Maybe a Debian developer would like to comment on that?
By the way, he openly admits that the unstable branch is vital for Ubuntu which could explain why he thinks that it is better to focus on it.
> What I want out of Linux:
>
> 1.One GUI.
>
> 2. Ability to play DirectX games.
>
> 3. Double click driver and application installs. "Fire and forget"
>
> 4. No preaching. I don't really give a rat's ass about what is free and what isn't.
As your points, especially the last one, make it abundantly clear, what you want is not Linux but a free[*] clone of Windows. Nothing wrong with this, of course, but what does it have to do with Linux and why any of the developers working on Linux [desktop] should care about what you want?
[*] I presume the cost is the only thing which keeps you from just using the original right now
The main - and subtlely articulated - point that I gathered from this:
Shuttleworth is ackowledging that many of the Ubuntu users/booster club members are thinking out of their ass. Cruise over to the Uubuntu forums (or any of the unbearable "I just installed Ubuntu" threads on Digg) and you'll see a blatant ignorance of Debian. Not of its existence necessarily, but of Debian's immense role in the Linux world for all of these years. Mark knows it, the Slackware folks know it (but don't want to deal with those sysv scripts), but the "my laptop spins like a floating cube with Ubuntu" crowd don't always get. And their brash attitude is a bit of an embarrassment.
Mark Shuttleworth is not in a position to tell other projects how to manage a project without conflicts. I recall that just before the Dapper release some German Kubuntu developers threatened to leave the project because Canonical refused to communicate with them. One of these rebelling German guys was the main developer of K/Ubuntu's new live-cd.
Part of the problem seemed to be that these Kubuntu developers were not paid employees. There was one paid employee in the lead of the Kubuntu project and this employee did his best trying to convince people that there was no conflict, although obviously there was. Hiding problems and denying conflicts seems to be the official policy of Mark Shuttleworth's pet project and this carefully built image of easy success that they want to project to the public makes Shuttleworth now think that he can advise other projects about their goals.
"Let he who is without sin throw the first stone," but Mark Shuttleworth is not as innocent as he'd like to appear. He has faced conflicts in managing his own project and I'm not at all sure that he's the right man to tell other projects how to avoid conflicts. And advising Debian to concentrate its efforts on improving Sid is definitely a bad advice, although such decision would certainly suit Ubuntu that is built upon snapshots of Sid.
What's wrong with a little tension? It's all about how you handle the tension.
It really seems to me that Ubuntu really is more of a community than a just a distro. Packages appear to be carefully selected based on what the community is asking for, and the effect of the community forums for support make it a great distro for the newbie (be nice to the newbie! when properly nurtured, they can grow into gurus!).
I think that if Debian total fell apart for any reason, Ubuntu would continue to move forward. If the community found another distro to serve as a base, I think they would just use that base.
It's all about community, the people. Without that, software really doesn't matter much.
"The large print giveth, and the small print taketh away" -- "Step Right Up", Tom Waits
Right, when will that be? Aside from Office & AutoCAD there are still a LOT of half-assed custom installers full of boiler-plate text that no-one has bothered to customize. (Look for "Your Company Name here" in the "summary" info on a random Windows installer some time. Nullsoft installer really isn't helping matters, and so it Microsoft's lack of good tools for making clean MSIs. Well, and the fact that the MSI format does not make any intuitive sense, it's more or less a binary SQL dump of a bizzare data model for describing parts of a program's installed state, peice by explicit peice. WiX helps, but it still exposes you to the SQL-style silliness, like having to explicitly list every file with it's attributes at least twice in a verbose XML syntax (Once to define it, once to include it in a component that can be installed).
.deb's and there's no comparison, .debs are a LOT less work, are a lot easier to understand what's going on, and actual has a useful way to define prerequisites. That, and even though MSI is only supported on NT-based system Wix requires defining 8.3 format filenames.
I agree Windows would suck a LOT less to maintain if MSI were used, but even Microsoft seems to be moving away from it (not that they ever went whole-hog for it), with the Vista installer. They arn't installing their components from their own managable and patchable package format, the description of the new Vista installer seem to be that it takes a snapshot that somebody at MS duct-taped into working (hopefully) then burned to CD.
Not even Microsoft distributes things as MSIs or MSPs though, they distribute "I 0wnz0r your system!?1111!!111" EXEs. At least MS doesn't gimp their MSIs to refure to install without the setup.exe wrapper, like some companies do.
I've made a few MSIs as well as
I hate Windows because I have to make it do useful thing.