“How many developers run for the post of leader of the Debian GNU/Linux project and cite as part of their platform a desire to make Debian sexy again? None that I know of – except Sam Hocevar who won the recent election for leader of the project. One among eight who put forward their cases to the 1043-odd developers who are eligible to vote, Hocevar modestly puts his election down to ‘luck’. He says it is a vote for change.”
Anthony Towns failed to impress me, but this guy seems to have a good head on his shoulders and communicates well. He wants to steer the ship rather than stay the course or replace the rudder.
I think the key to restoring the glory days of Debian is collaboration. Debian needs to be a central hub of the free software community, if not the hub. It should have a mandate to develop tools that help the community collaborate. It should serve as a model for how to develop and deliver high-quality free software. It should set the bar as an unfocused distribution and let others elaborate in their respective areas.
Red Hat and Novell will continue to do their own thing, and there will always be various niche distribution projects going at it on their own. But for the most part, Debian is the platform on which the rest of the community is based. It’s the supermarket for distributions, the bio-ooze out of which innovative new distributions grow. Debian isn’t a blast from the past, it’s the way forward. Debian is sexy because it brought us Ubuntu, perhaps the sexiest distribution project we’ve seen in a long time. Debian will continue to be the basis of new derivatives more than any other mother distribution.
Debian should have an important role in the proliferation of Linux-based devices, including PCs. Why would an OEM like Dell go with a Red Hat-based solution when Debian has proven itself to be the most flexible platform for creating targeted Linux products? Wherever a differentiating software platform is a requirement, Debian should be atop the list of solutions.
Debian’s success is important to everybody that uses free software, regardless of your distro of choice. I wish Sam the best of luck in his new position and hope to see great things from the project going forward.
Almost any distribution based on Debian is a winner.
Lets hope this guy can dress Debian in something sexy for all to drool over.
Enough said.
Maybe I am too sceptical, but for some reason I don’t buy your “all hail Debian” message.
“I think the key to restoring the glory days of Debian is collaboration. Debian needs to be a central hub of the free software community, if not the hub.”
Really? Why should it a be a central hub for the free software community? What makes you think it’s even possible for the free software community (Who is the free software community for that matter? Does this include Redhat or Novell?) to have a hub?
“It should have a mandate to develop tools that help the community collaborate. It should serve as a model for how to develop and deliver high-quality free software.”
How are you going to implement this mandate? And why should Debian be in charge? You might think they are the holy grail of free software, but not everyone (thankfully) agrees with you. If Debian cannot manage itself, why should it be a model for free software? Yo might think that Debian is a model for the perfect free software organization. Many other people don’t. For instance, I don’t like their attitude towards the Firefox trademark (I mean wtf, don’t they have anything better to do? The whole incident didn’t benefit anyone at all), I don’t like their lack of leadership, I don’t like their lack of pragmatism.
“It should set the bar as an unfocused distribution and let others elaborate in their respective areas. ”
If it sets this bar, Debian will automatically become ‘the one distribution to rule them all’. Once that happens I’ll take your ramblings about a model for free software more seriously.
“Red Hat and Novell will continue to do their own thing, and there will always be various niche distribution projects going at it on their own. But for the most part, Debian is the platform on which the rest of the community is based. It’s the supermarket for distributions, the bio-ooze out of which innovative new distributions grow. Debian isn’t a blast from the past, it’s the way forward. Debian is sexy because it brought us Ubuntu, perhaps the sexiest distribution project we’ve seen in a long time. Debian will continue to be the basis of new derivatives more than any other mother distribution.”
I like how you discard Redhat and Novell as some kind of sideliners not really involved with the free software world. You obviously have little understanding of Redhat’s contribution to the free software world. You seem complacent about Debian’s position as the mother of all distributions. Okay, Ubuntu is based of Debian and Debian has the most derivative distributions. So what? What’s your point? How does that make Debian the Holy Grail of free software? All this means is that Debian is a good base for starting a distribution? What does this have to with success in the marketplace? I guess you could talk about Ubuntu, but with Ubuntu I have to say it’s not that special. As desktop system it’s really not that good. Until the latest release, I believe it couldn’t even repartition your NTFS volume during installation and don’t even get me started on how unpolished Kubuntu feels.
“Why would an OEM like Dell go with a Red Hat-based solution when Debian has proven itself to be the most flexible platform for creating targeted Linux products? Wherever a differentiating software platform is a requirement, Debian should be atop the list of solutions.”
LOL, because Dell is a business! They need to work with a business entity, you know someone who can be held accountable for a release slipping by several months. Do you even understand how a business works? Judging from Hocevar’s description Debian’s decision making process is all over the place. Dell is not going to waste time and money debating and trying convince every single team about the merits of a certain change that would be needed for their derivative distribution. They want to tell someone what they need and by when, without any BS and waste of time on things like Iceweasel vs Firefox (OMG Mozilla is corporation with trademarks! Evil Alert!). Hell Dell would probably prefer to have Firefox as opposed to Iceweasel as a browser.
“Debian’s success is important to everybody that uses free software, regardless of your distro of choice.”
No not really, I guess Debian is important in free software’s current ecosystem, but I wouldn’t be surprised if without Debian we would have more dynamic ecosystem and end user (end users as in real desktop users – the ones that don’t know or care about what an OS is) orientated free software projects.
IMHO the real promoters of free software are organizations like Mozilla and Redhat who have been able to crack the end user space (in their own respective markets). When I see my uncle or even a frat kid (examples of people I know who use Firefox) using Debian or Debian derivative OS I’ll agree that Debian is crucial for the success of free software.
P.S. I am free software advocate/user and at one point I was helping Firefox (FAQs/support on mozillazine/ themes and extensions that kind of stuff). I am not criticizing your visions of Debian and Debian itself just because I like some other distribution (I actually use windows and I hate it), I am doing this because I genuinely believe that free software developers have to take a new approach to crack mainstream markets (and without cracking mainstream markets we will never be able to use a completely open software stack – think of things like PC Suite for Nokia phones, you’ll never get a verion of PC Suite for Linux until we have real end users using Linux). And Debian and its developers’ attitudes are very far from what (IMHO) is needed to get mainstream success. I am out.
“LOL, because Dell is a business! They need to work with a business entity, you know someone who can be held accountable for a release slipping by several months. Do you even understand how a business works? Judging from Hocevar’s description Debian’s decision making process is all over the place.”
So what happened to Microsoft when the Vista release slipped by several months, if not years? Did they refund Dell?
You raise very good points, but I think that mine mostly flew over your head, and reading back, I really don’t blame you for that
Why should it a be a central hub for the free software community?
Because Debian was the first and remains the only top-level distribution project that is rooted in the principles of free software development at the system level. The basic idea of free software is that it allows to software to take on a life of its own, to be living thing that evolves to meet the needs of different users, diverse applications, and changing times. The GPL enforces this at the executable level, but Debian extends this principle to the entire system. It’s relatively easy to create a new system based on Debian that takes a new direction or focus while remaining a part of the Debian development ecosystem. The underlying platform technology used by Debian makes this happen. They don’t enforce it. It happens out of pragmatic considerations.
There are Red Hat derivatives, but it’s hard to make a Red Hat derivative that tracks with RH development while maintaining its own identity. You’ll notice that Red Hat-based distributions are either re-badged releases (CentOS, Unbreakable, etc.) or completely different systems on their own evolutionary course (SUSE, Mandriva, etc.). It seems that the Red Hat platform can only split further over time, while the Debian platform can grow its ecosystem through derivation. Much as the GPL creates an IP bubble that can only grow and never shrink, Debian creates a platform that binds its derivatives together through their common use of unstable.
Yo might think that Debian is a model for the perfect free software organization. Many other people don’t.
I’m not claiming Debian is perfect, far from it. If you look back through my previous comments on Debian, you’ll see that I have been highly critical of their project management in the recent past. I’ve suggested that they should quit making releases altogether to focus on developing unstable and development tools. I’ve argued that Debian is in danger of having its mindshare sucked away by Ubuntu, which could never fill its shoes on its own. I stand by these comments, and that is precisely why I believe that Debian must rediscover and embrace its role within the larger free software community. I never argued that Debian is those things I mentioned, I argued that it should be more like that.
For instance, I don’t like their attitude towards the Firefox trademark
Mozilla made a decision to protect their trademarks, and Debian is respecting that decision. The branding elements aren’t redistributable in derivative works, and so Debian will not distribute them. Branding is an important issue for the free software community to consider going forward. How do we reconcile the vendor’s need to use branding to identify their work with the community’s right to create derivative works? Some distributions have chosen to distribute their Firefox build without the branding elements and others have chosen to distribute the official binaries. Debian has chosen to create its own branding elements that are freely redistributable in derivative works. This is actually what Mozilla wants downstream distributors to do if they don’t distribute the official binaries. They want them to differentiate their work by creating their own branding, not by distributing a brand-less browser that seems to be Firefox.
I don’t like their lack of leadership, I don’t like their lack of pragmatism.
Hopefully the new DPL will change your mind. For all of his (arguable) failures, Anthony Towns enacted some seemingly pragmatic policy during his tenure. Dunc-Tank is the most prominent example. Etch is a great example of Debian beginning to stick more closely to a release schedule and making a serious attempt at enhancing usability. Etch was not released on time, and it is not as user-friendly as one might like (compared to commercial and desktop-oriented distributions). But it was clearly a step in the right direction.
I like how you discard Redhat and Novell as some kind of sideliners not really involved with the free software world. You obviously have little understanding of Redhat’s contribution to the free software world.
I think that Red Hat is one of a key reasons why Linux become the most popular free software kernel. They could have chosen FreeBSD, which was arguable superior at the time. RPM was created as an incompatible alternative to the existing dpkg system already developed by Debian, which I believe was a crucial event that fragmented the fledgling Linux community. Red Hat and Novell have chosen to go at it alone with respect to packaging and quality assurance, whereas the Debian ecosystem shares in these responsibilities.
Dell is a business! They need to work with a business entity, you know someone who can be held accountable for a release slipping by several months.
The other poster made a good point about Vista being more than several months late. I wouldn’t have thought of that. But more generally, Realize that Dell would most likely work with a commercially-backed Debian derivative such as Ubuntu or Linspire.
IMHO the real promoters of free software are organizations like Mozilla and Redhat who have been able to crack the end user space
Those are real promoters of free software. But I find it hard to believe that Ubuntu won’t crack the end user space. Further, appealing to end users is not the whole ball game. Free software also needs to appeal to device vendors, service providers, and consultancies. These are areas where Debian might have more appeal than any other distribution project.
Debian is the computer equivalent of Anne Widdecombe – hardly what I would call sexy.
Doesn’t anyone at this site know what humour is?
Not at all, it is pretty damn unfortunate. Let’s start a new thread:
“Making OSAlert ‘fun’ again: The Readers Speak”
Not really a good comparison. No, Debian has never been sexy but it used to be the most Cosmopolitan distribution out there — a hub of new ideas and the mixing of various cultures into a messy but cohesive whole. These days it’s lost a lot of that Cosmopolitan shine to Ubuntu and settled for a more “boring world city that works because it always did with one cosmopolitan suburb called Ubuntu”.
Debian still has a chance to re-earn that Cosmopolitan but it needs to focus on learning a few things that Ubuntu did right, like having regular releases so 3rd party projects can co-ordinate properly (it’s just a matter of project management and having hard feature freezes and long stabilization periods) and, more importantly, customized packagings of Debian’s packages that aim at different target areas. Ian Murdock tried to create something like this with “Component Debian” (http://componentizedlinux.org/index.php/Main_Page), but Ubuntu delivered a “good enough” version of in practical terms with all it’s Ubuntu offshoots that are really just configurations of the common Ubuntu repository.
Edited 2007-04-30 14:56
“Debian is the computer equivalent of Anne Widdecombe – hardly what I would call sexy.”
Maybe it was in the past, but since the release of Etch Debian is more like Edwina Currie
it freaking works and works well! i don’t have to worry about applying a gazillion patches every month and when i do I’m fairly confident that its not going to break something. That is what is important to enterprise level computing. All you people who are obsessed with sexiness keep on your desktop, not my server.
Debian was ever sexy?
yeah, yeah, mod me down…..
Well if Justin Timerlake can bring sexy back why can’t Debian? Anyway I wish people would stop comparing Ubuntu to Debian. Ubuntu has some stuff going for it but I don’t trust it in the least. With Debian I know a stable system means stable! I’ve been screwed a few times with Ubuntu and updates breaking X or some other applications, which I find unacceptable on a “stable” system.
Edited 2007-04-30 15:13
plib, just because Ubuntu had(?) some growing pains doesn’t mean that it has nothing to teach Debian. After all, Debian does 95% of the work but Ubuntu gets 70% of the exposure. Why is that?
It’s not all Timerlake-fluff, since it it were, Ubuntu would die after a few releases since all hype eventually gets deflated once people wise up and they’d all move to Debian. As mentioned above, the only thing Ubuntu really contributes is that missing 5% and while you might think that this 5% is unimportant, it’s not. It’s the whole reason for Ubuntu’s success and it’s something that can easily be adopted by Debian without changing the core of what Debian is. All that’s required is the will to do it.
Let’s put aside the regular predictable release date comment (personally I think Debian stable should be have *longer* release cycles for enterprises but have more predictable release dates, but that’s just my preference). One area where Debian can improve is to create metapackages several of the common Debian installs with experts in each area choosing the best of the breed apps in each area. Create a “web server Debian”, a “router Debian”, an “LSB Debian”, “a firewall Debian”, a “Terminal Server Debian”, an “Enterprise GNOME Desktop Debian”, “Enterprise KDE Desktop Debian”, “Enterprise XFCE Desktop Debian”, “Recovery Disk Debian”, etc. Then create ISOs for each of these custom configurations along with the existing “allow me to pick exactly which packages I want Debian” and place them on a common high profile web page on Debian.org that is easy to find. There are hundreds of areas that would benefit from these “already configured” Debian subprojects. Debian would quickly become *the* metadistribution to go to when you have a need and want to satisfy a need and you either don’t have the experience to make informed choices yet or you don’t have the time to carefully configure Debian the way you need it configured.
One area where Debian can improve is to create metapackages several of the common Debian installs with experts in each area choosing the best of the breed apps in each area.
There’s already something like this in Debian, only it’s called “tasks” (which are better than metapackages). You just download the netinst ISO image and burn it onto a CD. Then, after installing the base Debian system, you start aptitude and go to “tasks” to choose what kind of system you want.
And if the Summer of Code project for Debian Live turns out successful, users can in the future build easily custom Debian live-cd’s with the “tasks” of their choice. Debian Live should become considerably more interesting when the live-cd gets a hd-installer.
http://code.google.com/soc/debian/appinfo.html?csaid=AE1F86C9D1EAF7…
http://lists.debian.org/debian-desktop/2007/04/msg00018.html
To those of us who use Debian and have used it for years, it is sexy–in a geek sort of way. It really does work the way you expect it to work, no guessing. If you want box with a specific task, even something as modest as a print server, you can do it with Debian without worrying about what other stuff the vendor has put in their distribution.
I know the average home user wants their desktop OS to be filled with pretty translucent colours and intricately designed animations programmed by fury creatures from nether-nether-land, but I personally couldn’t give a rats arse.
Granted nobody wants to navigate X using a GUI modelled on Windows 3.1, but at the same time does Linux /need/ to be sexy? After all, it’s not an either/or. Plus, imo, KDE and Gnome have already reached the point where they can compete with Windows XP in the GUI pageants.
I personally would rather see people focus on fixing some of those annoying little bugs that most of us know about but which often scares the newbies of – or developing innovative new concepts / interfaces before the likes of Apple and Microsoft patent them.
I don’t doubt that Debian does focus a lot of attention on those examples as well, but the whole argument about making Linux sexy strikes me as a little like ‘Pimp My OS’.
[edit]
maybe I’m just getting old and don’t look at other OSs like I used to when I was a young GUI slut
Edited 2007-04-30 16:31
I don’t think making Debian sexy has anything to do with dancing icons that fart angel dust when you do a mouse-over.
I think sexy _to me_ means being a leader at the forefront of innovation.
#1 – For end users, Debian could be a leader in promoting/creating tools that every-day users want:
* – simplified hardware configuration.
* – simplified backup tools.
#2 – For system admins:
* – simplified meta-packages, e.g., LAMP, Mono
* – simplified repository tools
* – simplified network-wide management tools
#3 – For developers:
* – simplified, unified tools for RAD. Includes auomatic package creation. Create tools for eveyday businesses to use. Development isn’t just for C/C++ dudes.
I realize that Debian is a distribution, thus it doesn’t actually “create” most of the software, it builds the system. However, it has the awesome power to focus and promote tools. It is the integration of all of these tools (most of whom already exist) into a useable form that is essential. Debian doesn’t just have to be about combining thousands of individual packages into a final working compilation. It can also be about focusing attention on a cohesive, useable, working system. This may involve the attrition of some of the thousands of choices.
I think the revolution is on creating a very useable system for everybody, not just a select few. To me, Debian should be on the forefront of this revolution.
Or maybe it is about creating pretty pictures, AKA, “Debian Does Icons”.
On a seperate note, people should check out this mentor program debian now has. It is a good way to get your packages into debian repos.
http://mentors.debian.net/cgi-bin/welcome
While I may not be a linux expert, I did learn to love linux on Kanotix (Debian for all intents and purposes). I am not afraid to edit sources.list or install/remove packages from CLI; I am comfortable with Debian. As a laptop user, it is hard to get everything working on Sarge. So, I patiently waited for Etch. From time to time, I would download a development snapshot and give it a try. To my surprise, the network manager under Gnome does work with my wireless card. I have gotten it to work under various distros without the need of non-GPL software. In fact, I can use the wireless NIC for the netinstall, but it won’t work under Gnome (and it will work under KDE, but I don’t like KDE). I was very, very disappointed in Etch. So it is back to FC6, but I really miss Debian.
PS – the fact that you can’t remove Evolution without removing Gnome is also not so sexy.
Wireless has nothing to do with gnome or kde. You must however add you user to “netdev” to use nm-applet (NetworkManager).
You can remove Evolution without removing gnome but you will not be able to keep the gnome package. How sexy would it be if it said “GNOME” (DEBIAN GNOME) was installed and it was not? Ofcourse you should not be able to have the main package installed if the dependencies are not met.
Love it or hate it, Debian is consistent with it’s stability. Even if it means annoying us by not including software that it doesn’t consider “stable enough.” It’s “sexy” by knowing what works!
A sexy OS to me is simple, clean and effective.
Sort of like the ivory girls. They were all pretty yet still had some dirt under their nails.
apt-get is not really a gui tool but is so easy to use, and many debian versions of packages aren’t the latest but they are very effective – it’s this stability why I stuck it out with debian.
It’s a free country man, use what you want.
“and updates breaking X or some other applications, which I find unacceptable on a “stable” system”
The infamous “Breaking X in Dapper Drake”=basically due to proprietary nVidia drivers. Not their fault, really. Anyways, they include a dapper-proprosed repo now, instead of feeding updates right into the LTS.
Either way,he was right,people should stop compareing unbuntu to Debian.
Anybuntu is NOT debian.