Slashdot Mirror
Interview: Ask the Debian Project Leader
Wichert Akkerman, our interview guest this week, describes himself: "I'm a simple MSc Computer Science student who likes to work on Linux projects. I have been the Debian Project leader since February and that has taken most of my time. Debian is a project to produce a completely free operating system and as such we currently have a GNU/Linux distribution and are working on GNU/HURD as well." So ask away. One question per post, please. Moderators & assorted others will select the 10 - 15 questions we forward to Wichert Tuesday afternoon. His answers will appear Friday.
How do you feel about the emerging of distros basing themselves around Debian - eg. Stormix and Corel.. Chris. --
To what extent are you seeing Corel feeding back these changes to the Debian community? How good for Debian has Corel been, so far?
I'd like to know the status is on porting Debian to the FreeBSD kernel. A combination of the daemon and debian - how delicious.
Any news on when Debian will start accepting new maintainer applications again?
Is Debian looking to concentrate on a small set of well-established architectures, or are we going to see Debian for every conceivable environment that Linux can run under?
It's a small world and it smells funny; I'd buy another if it wasn't for the money; Take back what I paid (SoM)
What's the best way to get involved with the Debian project? Do you have a list of tasks that need to be done along with the required skills?
I ask because that seems to be one barrier keeping more people from helping out various free software projects -- they don't know where to start. If we could point to a list and say "The boot disks need testing; we expect these error messages:" or "The foo package has these ugly functions that need to be rewritten:" it would give us more concrete goals to reach.
--
how to invest, a novice's guide
With the things that have been learned from those attempts, is there likely to be some sort of dselect-ng ?
If you're not part of the solution, you're part of the precipitate.
Now that Qt 2 is free software, under the QPL, will Debian include KDE 2 when it is released, based on Qt 2?
Also, do you feel it is better to keep Linux entirely DSFG free software only, or to include software in some way restricted, such as Pine, Qt 1.x and Netscape?
Since you are working on both Linux (established) and the HURD (experimental), could you please tell what the advantages of using the HURD over Linux would be, once the HURD would near completion?
What are your feelings on RPM vs. dpkg? Would it be better for Debian to add any missing functionality to RPM, and then switch to that? In what way might Debian users benefit from sticking with dpkg over a modified RPM with equivalent functionality? From personal experience, the thing that really stood out in Debian was dselect, but that could sit on top of RPM just as well as it does on dpkg. Presumably the same applies to apt (although I haven't looked at Debian recently enough to know about apt).
"The invisible and the non-existent look very much alike." -- Delos B. McKown
Debian totes that it is a completely free distribution of Linux, and likes to look down on those that package things like Netscape and StarOffice in their distributions because they are not totally free. Meanwhile, Debian packages these things as well, but puts them in a section declared to be "non-free." How is that you can still maintain your air of superiority for a "free" distribution if you're distributing the same non-free programs as everyone else, yet still look down on distros such as RedHat solely because they don't keep a seperate non-free section? Would this not fall under the category of hypocrisy?
Interested in open source engine management for your Subaru?
Could you compare GNU/HURD to GNU/Linux, from a user's perspective. What are the advantages of each? What are the target groups for each? Why would I choose one over the other?
1. Eric Raymond has stated (during a speech, NYC :) that he believes that the Debian project is making it difficult for Linux to adopt a standard packaging system. What do you have to say in response?
2. I'm rather disturbed by the GNU Hurd's Debian-like appearance now. I was hoping for a radically different OS but I was disappointed with seeing a different underlying OS that still looks much like Debian. Does Debian have a legitimate reason for doing this?
3. Now from the GNU Hurd to Debianizing FreeBSD. My only question is: Why? Does this help improve the state of the world or simply help glorify Debian itself? I don't see a practical reason for doing this.
I am trying to understand your motives. Not ridicule them. Thank you for your time.
To my mind, the main problem that Debian has to sort out is its release cycle. It's one thing to have a well-tested distribution by the time it's released, but it's going too far to have packages a year or more out of date still in the current release. What steps are being taken to address this? Or is there an expectation that everyone is happy to use unstable?
11.0010010000111111011010101000100010000101101000
I was looking over the info on the attempt to integrate FreeBSD's kernel, and was shocked to find that the people doing it were using BSD libc! Since glibc was designed with a certian amount of portability in mind, why not port glibc to FreeBSD's kernel? This would seem to be to make the overall port MUCH easier, as the rest of the debian code should be far simpler to port to a different kernel platform, but the same libc....
Hi!
First of all, thanks a lot (to you and to the rest of the Debian team) for an oustanding job!
Several quick questions:
(A) What is exactly the BSD/Debian project? What are its goals? Is it officially supported by the Debian project, or is it just some sort of personal projects for some Debian programmers?
(B) When can we expect a release of the GNU/HURD? Do you think the HURD is the next great thing and that we'll all have to leave Linux/*BSD to install it on our machines?
(C) On a more personal note... how do you manage all this work? Jolt, Coca-Cola, designer drugs or just good old plain caffeine? Do you receive extra credit for working on Debian during your studies? =)
Thanks again -- keep up the good work!
The right to offend is far more important than the right not to be offended. (Rowan Atkinson)
How do you, and the Debian team feel about Corel only allowing adults (over 18) to download their distrobution of Linux?
Other major distributions have their derivative distros - eg RedHat and Mandrake etc. There are also some good jobs done of porting RedHat-written code to Debian (eg Linuxconf).
Given the problems experienced with Corel Linux, which is known to be "based on Debian", what are your feelings on having other distributions derived from Debian - in particular, how 'far' away from the original should they go and is there any preference in direction that they take?
~Tim
--
Rushing on down to the circle of the turn
It seems to me that Debian is stuck in a situation where it's minor releases are actual MAJOR releases.
At what point will potato be released, and what revision number will it have? It seems to me it would have been better to have released potato a long time ago, and release 'sub-releases' often to keep up to date, instead of waiting for that never-to-be-seen 100% perfect release.
'Release Often' is a term often used to describe Open Source projects, but it doesn't seem to apply to Debian. Just sticking in 'beta' forever.
-- I'm the root of all that's evil, but you can call me cookie..
The current stable debian distribution "slink", is still based on the 2.0.x kernel, while the other big players couldn't adopt 2.2.x fast enough just because of the version number. With the 2.4 kernel just around the corner, how will Debian be able to compete with i.e. Red Hat, as they even might launch the next version with the 2.4 kernel as the Debian maintainers are still working on getting potato out the door?
What are the plans to do something about it? Hon the lot of geeks that are eager to help out join in, or what are the criteria or needed skills? Does one have to be an über-geek to get in?
Okay... I'll do the stupid things first, then you shy people follow.
Okay... I'll do the stupid things first, then you shy people follow.
[Zappa]
Debian seems to be getting too big to be managed. (by the user and wakkerman)
There are many packages and they are getting more and more. ("What? There is a new window manager? - Package it!") - I don't think this is the responsibility of a distribution.
A distribution should be the base system to run linux. Every more advanced system should be installed by the unix administrator.
To make it even more worse, packages like netstd get split up in many others and packages which should be split, don't. (Look at tetex-bin. You only need xlib6g because there is xdvi in it. - If you drop xdvi in it's own package you don't have to install xlib6g and xbase on your servers)
Recently, Debian adopted "logrotate" as their standard log rotation tool. I thought this was interesting because it was developed by Red Hat, not so much to increase the pool of Generally Good software out there as to meet a specific need they ran up against in the process of building a distribution. If this distinction makes any sense (and I think it certainly applies to, say install tools), what can you borrow from other distributions to bring Debian forward?
--
Xenu loves you!
As an enthusiastic user of Debian, one of its greatest weaknesses to me has been the apparent "orphaning" of stable releases once they are released. All further development (security updates aside) appears to be done for the new, unstable development version. Even new versions of existing, well defined and stable products are rarely backported into the stable tree, resulting in the stable version of Debian quickly consisting mostly of outdated software.
Is there any provision being made to allow for and support a more aggressive backporting of newly released software in current, stable releases, such as newer versions of xmms, netscape, mozilla, and so forth?
I understand and recognize that some software (e.g. gcc, glibc, X, perl) may affect too many other dependent packages to be supported in both stable and development trees, but other software such as xmms and enlightenment don't fall into that category at all, yet debian packages often are never created to support the current stable release.
The Future of Human Evolution: Autonomy
I've been using Debian for a long time now, and I'd like to contribute back to the project. However, I've been put off by what looks to me like excessive bureaucracy and some infighting among Debian developers. Are there any plans to streamline the process to become a developer/maintainer, and the developer contribution process itself? What about fostering a more civil peer review process?
I understatand that every time the Debian unstable gets bigger and bigger and supports even more hadrware architectures. It take a while to make the distribution really stable and rock-solid before releasing it. However, what disappoints me is that, after release, Debian basically abandons the stable. The only updates that make it into stable are security (good job!) and fixes for some horrible bugs. So the current Debian stable still ships with Kernel 2.0.36 by default and XFree 3.3.2.x (this is rediculous!). Many people, and specially newbies get disappointed, they install it and then realize that their hardware is not supported and they have to rebuild kernel or hunt for "unofficial" official XFree 3.3.X debs for slink. Did Debian maintainers consider making more updates for stable? Specially, the updates that have to do with hardware support like X, Kernel and other related packages.
- akop
worlds as to what the `natural unit' of an operating system is. In
the Linux world it seems that the kernel is regarded as the operating
system, and the various different distributions are regarded as
flavours of Linux with their component parts being expected to be as
interchangeable as possible.
In the BSD world the whole distribution as deployed is held to be
the OS, and moving a piece of software from one distribution to
another has a bit of the flavour of a port. I think there are
advanatages to the BSD way of looking at things, since some properties
of OSs, especially security, make sense only when applied to the
system as a whole.
I am concerned, therefore, that the port of the Debian distribution
to the FreeBSD kernel might undermine this view. Is the Debian port
going to be positioned as an entirely new branch of BSD (say
DebianBSD), or is it going to be regarded more in the Linux way of
doing things? If the latter, how seriously do the Debian team take
these cultural differences between the BSD and Linux world? Who do
you see as your target user?
What do you make of all of this talk about appliances taking over the computer industry, and how sucessful do you suspect an OS like Debian will have if this really becomes mainstream?
Hi Wichert,
I've been following the development of the HURD for one or two years now, and I haven't seen much change in it. Do you expect the HURD ever going to be finished in the sense that it can be used in a serious way?
Cheers,
Martin van Boven
In light of the possibility of a Debian HURD distribution, and out of infamiliarity with Debian, I would like to ask what exactly Debian /is/...what comprises it. Is it a packaging/versioning/distribution mechanism? Is it a suite of bundled software? What exactly does Debian offer, or develop? How is it different from other distros? It seems to me that Debian isn't just a label and bundle of software slapped on to Linux, so what exactly is it?
It's 10 PM. Do you know if you're un-American?
I started using Linux at home with Debian 1.3, and stayed with it through 2.1. After that though, I switched to TurboLinux (though it might just as well have been RH or SuSE) for two reasons:
1) it seems Debian will forever be compiled i486,
and thus never benefit from the oft-huge speed increase of egcs/pgcc
2) the size and growth rate of Debian are, IMHO,
inexcusable. The main section no longer fits on one CD!
If it were just #1, I could probably live with it and just install a seperate compiler and library to compile and run the apps that I really need the speed from. But #2 is just nuts. The multi-CD method of install is very rough and difficult to figure out / use, and installing via ftp is simply not an option for those of us with 28.8 modems. Worse yet, this has caused the pace of Debian to slow to a crawl. "Stable" released versions contains libraries and apps of ancient (by the linux time scale) version, and the dependency structure of dpkg makes substituting self-compiled versions effectively impossible. In short, it's very difficult to have a Debian system that is at all current.
So, my questions are these:
1. Is the Debian project planning, at any point, to create a Pentium-optimized release?
2. Is the Debian project planning, at any point,
to create something like a Debian-lite, that includes only a core of packages such as commonly used libraries, X, popular user agents such as mutt, lftp, and lynx, essential and popular server daemons like sendmail, yp[stuff], nfs, and apache...? Basically, a distro of similar size to the more popular distros that fit easily onto one CD.
If Debian were to do those things, which I see as modernizing and streamlining respectively, I would switch back (or at least try it out in vmware =]).
MoNsTeR
After reading about 1/3 of the posts, the major complaint about Debian (it is free, who are we to complain?) seems to be that stable releases are abandoned. I have been using Debian throughout my Linux experience. I tried RedHat, but I was dissappointed with it. Debian does seem to be getting behind. I find myself downloading tar.gz files and upgrading a lot of components myself which is very time consuming, but probably good practice if nothing else. But it kind of defeats the point of having a distro. I am not presuming in anyway to be an expert or proclaim to have a better idea than the current one, but I had a suggestion on a new distro model that might make things a bit easier, especially with the fast moving world of open-source projects.
** A lot of open source projects make projections about when a new release will be made or one atleast knows the next version number obviously. Why not build a distribution in the following manner.
(1) Look at the open-source projects included in your distro or that you want to include. Look at your current distro set your new distro to include the next versions of open-source projects you currently include or plan to include. Make that your goal for the distro and when it is reached, up your version number and set a new set of goals.
(2) All the while, maintain your current stable and basically keep upgrading it with updates with the intent of meeting your goals for the new distro version. Once it is met, your current stable goes up one version number.
(3) Allow the users to set their version level. Maybe they don't want to touch packages above the goal of their current distro. They don't need the latest windowmaker, icewm, gnome, etc.
Basically, your current stable is a set of open-source project versions (kernel 2.0.36, GTK 1.2, etc.) that interoperate together. Then you have additional packages that fit into your new distro such as kernel 2.2.12, 2.3.18, etc.) A lot more testing would be required, but considering the larger developer base debian seems to be getting, it might make better use of the resource. This is in no way a totally developed thought. You may have already considered something like what I described, assuming what I described makes any sense at all.
What are the chances of getting Debian GNU/Linux released more frequently? Currently the releases happen about every 14 months or so. Is there anyway to speed it up so that it is say, every 6 months with minor bug/security releases to stay at it's usual speediness?
It seems to me that the whole HURD project [tho very promising in theory] has been in pre-beta hibernation for quite some time... so my question is:
What exactly would it take to ramp up work on the HURD so that *Debian/HURD* actually becomes reality... or maybe the question should be "is there no Linus Torvalds or Alan Cox for the lowly HURD??"
<---[singularity sig]
I've seen a number of comments already dealing with this subject but none have asked specifically what I want to know. As a new user to linux if anyone has a simple answer to my question, please answer
I've been wanting to try Debian for awhile now, the only thing holding me back is the tendency for Debian to fall behind by one current kernel release. With the upcoming release of pototo based on the 2.2 Kernel as well as the 2.4 kernel is Debian destined to fall behind once again? Are there any plans to adjust the release schedules or come up with a new release system to keep more up to date with the new software we're seeing more frequently?
IMO opinion most of the die hard Linux fans are faithful devotees of Debian yet they also want to be on the bleeding edge. It seems to me Debain would benefit greatly by keeping more up to date even if it was with sub-releases incorperating things such as KDE 2.0 and the 2.4 kernel. As I said before I'm releatively new to Linux and since I've had literally no experience I would not be surprised to find all of this is easily solved with dpkg or something similar. Please let me know if it can be.
LiNT
Do you think that Corel has taken the right path in making a user friendly distro of Debian? Aside from the hoopla of threatened lawsuits and other controversy do you think it is easy enough for a newbie and fleshed out enough for someone who wants to really use it?
"Share your knowledge. It's a way to achieve immortality." -- Dalai Lama
Er..you may want to know that gnome-apt development has stalled, and that console-apt is probably going to have to be rewritten from scratch to go anywhere. (gnome-apt may as well, but I've only examined the console-apt code in any detail) Luckily there's not much code, at least for console-apt :) :-\
I'm also working on a frontend (so my opinion is not objective!), but I don't want to say anything else until the next release as all its nice features, including downloading packages (rather important for a package manager) are currently vapor; I know pretty much how I'm going to implement them but final exams are happening
Daniel
Hurry up and jump on the individualist bandwagon!
I've been running linux for over 2 years(first distro was Slackware 3.2). After a while the libc5/glib2 thing was too much and I changed over to RedHat 5.2. Updated to RedHat 6 when that cam out and it was ok. Heard enough good stuff about Debian that I thought I've give it a try a couple months ago.
With Slink being SO out of date(based on a 2.0.x kernel for starters), that's not an option to install. I need stuff like XFree86 3.3.5 for my voodoo3, some programs need to be run on a 2.2.x kernel, etc. So I decided to just point to Potato and install from there. The Potato install scripts crashed on me 3 times in a row(yes I know it's "unstable") and I finally Just re-installed RedHat 6.1.
So my question. What do you say to someone that wants to use debian, but Slink is out of date, and Potato won't install? It possible to just install Slink and "apt-get dist update" and point to Potato after a successful Slink install? How does one do a freah install of Debian now, and have an up-to-date system(2.2.13, XFree86 3.3.5, etc.)?
and PPC? is it likely debian will ever come to the PPC platform in any real form?
there's a PPC section of the Potato part of the debian ftp site, but it doesn't have a lot of support, and last time i checked there was _no_ documentation. and potato's supposedly unstable anyway.
Please, please say this will eventually come to be a full distribution. i would _really_ rather have debian on this here mac than the quasi-redhat that is linuxppc. Not to mention that linuxppc's distribution, well, isn't perfect. there are a _lot_ of things broken right out of the box. It would be nice to have something resembling an alternative.
Irritable, left-wing and possibly humorous bumper stickers and t-shirts
This is an important question to ask, but it needs to be restated in a way that doesn't sound like a barely disguised attack from a Linux zealot. How about something like:
Since you are working on both Linux (established) and the HURD (experimental), what new mechanisms, facilities or areas do you see the HURD opening up in future years beyond the Unix space that is covered by Linux?
"The question of whether machines can think is no more interesting than [] whether submarines can swim" - Dijkstra
Potato, contrary to popular belief, is updated, but updates consist of bugfixes (many security-related). You won't have the latest and greatest stuff out there, but if you want a stable box, it's the way to fly.
Slink offers a rapid development cycle and plenty of opportunities to experience the bugs and incompatibilities which plague other distros -- well, sort of (Slink is usually fixable). Packaging for Slink usually trails application release by a few days to a weeks for more obscure stuff. Plenty quick for me.
So have at it -- stable and conservative, or bleeding edge. Take your pick.
What part of "gestalt" don't you understand?
What part of "gestalt" don't you understand?
...I thought I was confused, but now I just don't know....
What part of "gestalt" don't you understand?
Naturally, it is an Internet/Free Software concept, and that's as up-market as they get right now.
Naturally, the issuing share price will be free, and underwriting will be by volunteers.
What part of "gestalt" don't you understand?
Wasn't there something about one question per post?
You can use this to tell dpkg that certain dependancies are fulfilled.
This is probably the cleanest way. Grab the package source, type dpkg-source -x file.dsc to extract it, cd to the directory it made, hack away on Makefiles or whatever, type fakeroot debian/rules binary (leave off fakeroot if you're actually root) and voila - fresh compiled packages.
You might wish to put them on hold ("=" in dselect or 'echo packagename hold | dpkg --set-selections') so that the packaging system knows not to upgrade them.
Also a list of source dependancies is kept by the build daemon machines - I forget where it's at though.
Well, there's been a lot of discussion on this on -devel and a whole bunch of build dependency fields were recently added. Currently this is optional; since there are so many packages to convert and there's not yet an automated way to detect build dependencies I suspect this will take a while (possibly a release or two) to get straightened out. Then we can start the fun stuff: modifying build depends based on compile options selected by the user! :)
Daniel
Hurry up and jump on the individualist bandwagon!
There has been a lot of discussion on debian-devel lately (and for some time now) on the "package pool" system. How far have these discussions gone? I've already watched a few simply disappear with no concrete changes at all.
Congratulations on the great work!
rbp
Besides GNU/Linux and GNU/HURD, is it possible that someday Debian might attempt an OS that completely breaks with the UNIX tradition? Although I understand HURD implements UNIX in a different way, how about an OS that reflects and entirely new way of thinking in its API's and its visible structure?
-- John Truong
libapt-pkg takes care of the protocols, pulling data off the servers and so on, but I still have to call functions to start the download and do something sensible in terms of display while it's progressing (the display being the sticking point). I'm currently investigating just how far I can push libapt -- I have some extremely nifty interface ideas for the download screen, but they may not be possible with the current API.
Daniel
Hurry up and jump on the individualist bandwagon!
Just be careful not to relicense code you didn't write.
I thought I heard a long time ago that KDE was planning to relicense all the code they own to the Artistic license, does anyone know what became of this?
If there are no differences above kernel level, then there is little point to bothering with Hurd. Might as well just improve Linux a bit.
The point is twofold:
Things may not be there yet, but that's certainly the intent.
The result of this is much as you suggest, that there don't have to be a lot of differences visible in user space. Applications that run on Linux should also be able to run on Hurd.
The notion of filesystem translators, for instance, is something that Linux doesn't do.
As time goes by, if there is any merit to Hurd, the use of Hurd facilities such as translators should result in systems based on Hurd diverging from the way Linux looks.
Conclusion: Both Linux and Hurd offer many things that are similar, such as:
- Multiple users
- Multiple tasks
- Hierarchical filesystems
- GLIBC
...
which will result in them looking pretty similar in a lot of ways.The similarities at present comes from trying to get the stuff that works on Linux to work on Hurd.
Eventually, if Hurd "goes well," the differences will emerge...
If you're not part of the solution, you're part of the precipitate.