Slashdot Mirror
The Debian System Explained
An anonymous reader writes "XYZComputing has a great interview with Martin F. Krafft, the author of "The Debian System". From the article: 'Despite Debian GNU/Linux's important role in today's computing environment, it is largely misunderstood and oftentimes even discounted as being an operating system which is exclusively for professionals and elite users. In this book Krafft, explains his concept of Debian, which includes not only the operating system but also its underpinnings. Debian is not only a robust and scalable Linux distribution, but it has many other features which are worth looking into, like its open development cycle and rigorous quality control.'"
I'll use nothing other than Debian and Debian Based distro's. Ubuntu and Kubuntu are nice, as they are based off debian, have the massive package base available, but also are updated a bit more often.
Big ones, small ones, some as big as yer 'ead!
Give 'em a twist, a flick o' the wrist...
...RTFM !!
Ah, it is still good though.
Debian 3.1 is a dream. Easy to install, no more updating (except for security updates), and rock solid as my desktop OS. FreeBSD was similarly solid, but the package management and printer control for Debian is just so darned easy. Hats off to Debian!
The Death Penalty: Killing people to show others that killing people is wrong.
The organization is as interesting as the technology. Lots of people are willing to put in lots of volunteer time.
I wonder how long it will be before the business schools start to take notice of successful open source projects and learn a bit about management.
Interesting. I went through the same linux phases...
slackware, red hat, suse and now debian. Looked at Gentoo a bit but stuck with Debian.
I'm not an admin - outside of my own hacking at home. But, help me out here, is Debian more of an enterprise-admin friendly-scalable distro than, say, RedHat Enterprise?
From what I've seen between various distros(No Debian), there's their add-ons (desktop add-ins, installation software, etc...), and then there's just Linux, XFree86, and all of the GNU software stuff. Is Debian that much better whe it comes for day to day operations?
Don't tell everyone how AMAZING Debian is, please! I get this strange (probably very sick) psychological aversion to seeing things I feel are part of my ingroup assets become popular. As a Debian user of 7 or 8 years I get a little nervous at this. My choice of (vastly superior) operating system is what makes me feel different. Have a little mercy on a nerds elitist insecurities please! Im the guy who always discovered underground bands years ahead of everyone else, and when they finally became mainstream I wanted to disown them. My 'discovery' felt _violated_ by the hoards of unwashed sheep jumping on the wagon. 20 years as a 'geek' and now I hear that 'geek is chic'. Time to become a merchant banker. Stop following me around you horrible unoriginal soulless people!! Find something of your own. Debian is the best kept secret in the world to me right now, don't go around telling the oiks all about it or they'll hijack it, misrepresent it, and corrupt it by dragging it down to the lowest common denomiator like everything else they touch. Next thing I know some techno wannabe will be coming up to me in the street and saying "Hey, have you heard about this really awesome new operating system called Debian Linux!" ....Smack!!
Windows is the choice people! Windows is the best, trust me. Debian is rubbish!
This sort of thing reminds me of Joel Spolski's opinion on advertising:
The idea of advertising is to lie without getting caught. Most companies, when they run an advertising campaign, simply take the most unfortunate truth about their company, turn it upside down ("lie"), and drill that lie home
If Debian is so scalable, why does it take them so much longer than any other OS vendor to simply do a release? How comes the software even in the "unstable" version is so often out of date? If it's so robust, how comes that shortly after their last stable release it was revealed that their entire security infrastructure revolved around one man, and that when he went on holiday the flow of updates simply stopped?
It seems to me that if you wished to advertise Debian, scalability and robustness would be the last qualities you'd choose to highlight. Instead you might want to focus on its dedication to the ideals of free software (that doesn't entice many people to install it though ...) or the fact that it runs on so many CPU architectures (hmm ... ditto). Actually, I can't think of any compelling reasons for the majority to run Debian directly. I say this even though I use Debian on my own server .... the original reasoning for this was that I felt at least the community was big and stable so there would be a reliable supply of security updates. That was before I found out about the size of their security team and the bandwidth bottlenecks on their servers (eg, Xfree update). After that I found myself wishing I'd installed Red Hat instead.
I guess many people agree with me because these days I see very few people advertising Debian as the Wonder OS that it was promoted as when I first got into Linux. These days people tend to promote it by pointing to the (significantly more popular) operating systems built upon it, like Ubuntu or Knoppix. Of course, it's not exactly great PR to promote yourself as the base for what are effectively (policy and project-wise) forks, but marketing was never Debians strong point ...
I've been using only Debian for about 5 years. It's the best. I totally support the community, and the philosophy behind Debian. Debian Stable is great for some purposes, and Debain unstable is great for others.
I've been reading Martin's book (it cost me $30), and unless the second half has a lot more in it than the first half does, there's not much there that an experienced user of Debian doesn't already know. So if you're already an experienced Debian user, the news is good: you already probably understand a lot more than you think you do! If you're not already experienced with Debian, what are you waiting for?
Anything worth doing is worth doing badly -- G.K. Chesterton
I was born in 1979 and I became a beta-tester with 13. I don't think Microsoft actually ever cared about my age. I was contacted about beta testing because I completed all 6 MCPs required for the MCSE in a single day, or at least so the email said.
I still have the reports I sent to them as part of the beta testing, and mind you, even now they are everything but childish. But they clearly come from someone who was used to administering nets for 50-or-so users with Netware, and Netware wasn't the direction they wanted to take with NT 4.
For instance, I proposed to allow specification of ACLs per user, rather than per resource. Who here has used Netware and didn't come to love this feature?
-- martin
echo mailto: !#^."<*>"|tr "<*> mailto:" net@madduck
Kudos to Martin Krafft for writing his book. Many dream but few ever get it together ...
That said, I spent most of 2005 running Debian Unstable and Debian Testing on different systems and ended up finding both overrated and generally a disappointmennt. Debian was too demanding of time and needed seemingly endless fiddling around and careful management. It also took a lot of time to set up, though admittedly that is a one-off when an installation is still fresh. More important, the Debian developer community seemed shot through with an obsession with doing things the Debian way, with college-level debates (aka rows), with considerable disdain for new users and with frankly pretty obscure things of little interest to many in the everyday world. Overall, I began to wonder if some of these guys would recognize an end-user if they fell over one and my faith in the Debian way rapidly dwindled.
None of this should detract from Krafft's achievement, though. It's a heck of a good thing to have done. I do find it a little odd that he should recommend that new users try Ubuntu rather than Debian. One is tempted to ask: what's the problem whereby they can't use Debian, then?
For myself, I've now gone back to another distro. It's pretty nearly as capable as Debian, with the difference that its devs are technical experts who confine themselves to delivering what works. A distro that puts out for its users without striking tiresome poses or co-opting its users into politics of some kind is much the more preferable, for me at least.
Las qué passoun
tournoun pas maï
I have couple Sun Ultra 1 workstations and had no trouble installing Debian Linux (Sarge) on them. Just checked couple faqs and howtos to be sure my hardware was supported and how to change screen resolution and how to patch firmware to support 64-bit mode (though I think it was Solaris that required the firmware patch to even boot).
Either you did something wrong or just happened to have an SPARC that wasn't supported or tested.
- Raynet --> .
Debian installs fine on my Sparcs, but Linux kernel crashes on Sparc. Debian patch or vanilla tarball, you can get the kernel to shit within 30 minutes of running crashme. Try it.
OpenBSD is much better for these old Sparcs, hopefully the MP will come up to scratch for the MBus boxes.
POKE 36879,8
I have since installed gentoo, using the standard install process, and it's worked pretty much perfectly - only gripe I have is it won't automatically generate an initrd with the right scsi modules like it can on x86, but I can live with that. So much for no linux distro having supported it for half a decade - I don't think gentoo's even been around that long.
I am trolling
Debian 3.1 is really great. True, they have taken a long time to make this release but as I see it it is a great OS for the server. I am running it on my server and very happy with almost no fiddling I had to do with it. However, trying to run it on my workstation was a different story. Mainly because of all the wifi issues I ran into. But that is not solely debian's problem. Fedora had the same issue. So if you are looking for something rock solid for your server, go with debian. If you want something on your workstation supporting tons of new new hardware and offer opportunity for endless finddling, go with Fedora Core 4. Just my 2 cents.
I've never had anything against Debian itself. My problem, as with a lot of other people, was always the arrogance that just seemed to ooze from the average Debian user. If you don't know what I'm refering to, then you probably relatively new to the Linux Community. It seemed for the longest that every question posted on every forum yielded the answer "get Debian". Debian's problem was NEVER being misunderstood - it was being misrepresented by the zealots that actually think their pretentous attitude represents the Debian Community as a whole.
Linux debian 2.4.26 #1 Sat May 1 18:58:40 EDT 2004 sparc64 GNU/Linux
debian:~> cat
cpu : TI UltraSparc IIi (Sabre)
I'm not saying that your problems aren't real, but Debian certainly supports Sparc and in my experience, it does so very well. I've had nothing but great success with Debian and Sparc, including the installer. Perhaps you should consider filing a bug report with the Debian Installer team. The architectures which have fewer users, receive fewer bug reports. How can you expect them to fix a bug they may not know exists?
As an aside, there is a good chance Sparc will be cut from Etch, so you may not have to worry about Debian "pretending to support SPARC hardware" in the future.
2^5
Then, to look for something like... libvorbis, I would just have to do this:
And it returns this:
Easy as pie.
I've run into far too many Debian users who have contempt for anyone who knows less than them that I have to assume it's an integral part of Debian culture. I would never recommend Debian for anyone who isn't an expert. If you like Debian anyway, and want to gain the best parts of Debian, recommend that mere users use Ubuntu.
-russ
Don't piss off The Angry Economist
1.5 is in unstable already, which means it'll soon trickle into testing. It also means that the next Ubuntu release in April will have it.
echo mailto: !#^."<*>"|tr "<*> mailto:" net@madduck
Cross-platform compatibility is essential. If the upstream Apache maintainers say Apache can be stopped with apachectl stop, Debian should damn well support this interface. I don't care if they provide /etc/init.d/httpd stop in addition, but they should support the standard interface. This makes life infinitely simpler for people who deal with many different systems---they don't have to keep relearning things. It also makes things simpler for people offering support to Apache users.
The tremendous benefits of cross-platform compatibility come from a package's interface being exactly the same on every system. It is a relatively minor benefit for different packages to have similar interfaces. Breaking cross-platform compatibility, as Debian does, for the sake of cross-package similarity is a horrible idea.
I should point out that I'm picking on Debian here because they are especially bad about this, but almost every major Linux distribution is guilty of unncessarily violating cross-platform compatibility in some way.
Wish I was Debian-Guru enough to help you with all of your problems, but i'll say a few things.
Ubuntu is basically Debian, so I don't know if I would do a clean install to switch. If your having performance problems make sure your not using the default kernel that installs, go ahead and use APT to upgrade to the 686 or K7 kernel to take advantage of your CPU. It should be very easy to do.
What graphics card do you have that is giving you trouble? If it's ATI I can probably help.
Regarding Gentoo, it's a very interesting way to go. It requires a bit more time and tech-savvy though. Compiling everything from scratch will yeild better performance (if you know what your doing) and Gentoo is over all a very nice distro (if you can call it a Distro). Portage is very slick. I usually reserve Gentoo for hardware that I really want to squeeze every last bit of performance out of.
If its an ATI card your having trouble with, hit up my site (click my nick) and use the feedback form to send me an email.
Big ones, small ones, some as big as yer 'ead!
Give 'em a twist, a flick o' the wrist...
I'm curious as to why you said "Wait," and then proceeded to correct yourself within your own paragraph. I mean, they call it an "edit form" for a reason. You can edit what you've typed.
hello dear sirs my name is jamesh i are india (bihar) can u guide me install red had linux 9?
You are completely wrong. The reason why Debian includes so many "old versions" of all software is because of the principle that security fixes should never force you to upgrade to the latest version; rather the fixes should be backported to the old version of the software. Thus, you don't have to worry about an upgrade breaking the functionality of your application in an unexpected way. This approach it is much harder than just tracking the current release like other distributions do with the mindset that if the software is new, too new for there to be many meaningful bug reports, then the quality must be good!
Debian also has very stringent requirements on how many known bugs it can contain before a release can occur. Just because the software is older does not mean that it is automagically bug free. Enormous effort goes into squashing bugs, especially before a release. Don't forget that these bugs span eleven architectures too and fixing these obscure bugs which only appear on MIPS improve the software overall. I'd argue that Debian very likely does more to imporve the overall quality of Free Software than just about any other entity.
2^5
10-15 hours to install Debian? That's too long even for an 'expert' install. Instead of wasting 20-30 hours after the first install, why didn't you just go to lists.debian.org and troubleshoot? This sounds like a troll to me.
waiting for ad.doubleclick.net
Yes. A 2.6 installation kernel is available for Sarge but it isn't made very obvious. The initial root selection menu will give you a listing of the available boot images if you ask for it. I think it is INSTALL_26. I got stuck by this on my Sarge install when I ended up with a hardware incompatability under 2.4 and had to restart the whole process.
This is important if you want to compile your own 2.6 kernel later since there a lot of 2.4 specific packages (modutils, etc.) that you won't be able to use if you don't install with 2.6 from the start.
I am becoming gerund, destroyer of verbs.
I would like to congratulate Mr. Krafft for homing in on a critical issue; that is, how to maximize and focus the efforts of people working on all of these various distributions, especially the ones that are derived from Debian, to benefit a main project. As an editor I know that 90% of the work in publishing is in achieving the last 5% of perfection. To my mind, the open source movement needs to polish one beautiful gem and give it to the world. Do that and astonishing things will happen. Take Firefox for example. But it appears to my inexperienced eye that a lot of effort is being distributed across a very wide field.
I confess that am new to the Linux world, but an old hand at computing. I successfully installed Gentoo on an old PIII as my first Linux project. I am glad I began with that difficult manual installation as I learned a tremendous amount. I did a lot of stuff with the command line, but wanted to see the GUI. Of course I could have installed a GUI environment under Gentoo, but I was curious to try something new.
The next distribution I tried was Debian. I loved its automatic installation, especially appreciated after my experiences with Fastab, Grub, and the rest. Some irritations with the printing system aside, I was impressed by the stability and completeness and professional look this system displayed under Gnome or KDE. And the galaxy of software available is astonishing. It left me with no doubt that sooner or later open source software will leave the server farm and be the norm on the desktop, at least in some computing environments. It seems to me particularly suited to educational environments, because only open source allows students to legally take apart their tools and see how they work. But I digress.
When I saw how many different distributions there are while doing a bit of research looking for a distribution for another old computer http://distrowatch.com/ I became concerned. Put plainly it seems to me that there are too many chiefs and not enough indians.
Hopefully Mr. Krafft's work can harness all of this creative energy and focus some of it back into some center or other. From my limited experience the Debian distribution seems very well suited as a candidate to champion.
"No fear. No envy. No meanness." Liam Clancy
I've used Debian for 5+ years, not just on a desktop but also as the basis of the distributed system of a medium-sized research laboratory. I've been mostly happy, but I'm becomming restless. Yeah, apt-get is really cool, but what about...
/etc/init.d/ scripts: Debian doesn't have a nice way to turn these scripts on and off and monitor their status via a command-line tool. Red Hat's system here was very good.
user management: I use LDAP for user management; others use SAMBA and other stuff. But adduser isn't a shim that can interface to any of these back-end data-stores -- it can only do /etc/passwd.
ideology: Debian's ideological bent can be a real pain for those us using the distro for its technical merits. For example, Debian pulled SSL support from all the GPL network services that link to libssl in a fit of ideology that no other distro has had.
package management: Yeah, apt-get's dependency resolution logic is very cool. Other aspects of the system aren't so cool. Apt-get, aptitude, and other front-ends don't share the same back-end data-store, so if you mix and match these tools, you get inconsistent package data. And it's nearly impossible to force-remove a package (just delete all the damn files and forget about it!) if the associated removal script fails.
An Air Force evaluation of Multics, and Ken Thompson's famous Turing award lecture "Reflections on Trusting Trust," showed that compilers can be subverted to insert malicious Trojan horses into critical software, including themselves......This paper describes the technique, justifies it, describes how to overcome practical challenges, and demonstrates it.
This paper is full of it.
The technique is possible, but so impractical as to be completely useless.
Modern compliers aren't actually that advanced. their optimisation capabilites only go so far due to their poor ability to interpret the application. As such, I seriously doubt that compiler trojans are in any way a serious threat. The threat from actual trojans in binaries if far, far, far greater.
May the Maths Be with you!
So she can use Synaptic, which has pretty decent searching built in...