Interview with Debian's New Project Leader

With the recent news that Anthony Towns will be taking over as the Debian Project Leader, Linux.com took a few minutes to sit down and feel out the new DPL-elect. From the interview: "The immediate plan is to organize the various ideas I've had so that I can work out which ones are actually worth working on, and what order to do them in; and to make sure that all the people who volunteered to be DPL during the campaign, or offered their help don't go away without some good ideas about extra things they can do. "

Comments

ftpmaster

[th173]

I am not happy with an ftpmaster as new project leader. They were the cause why the last one gave up.

But on the other hand, there is finally a chance for some movement and some chances; both are needed by this brilliant and outstanding project which is completely stuck by politics now; just like "the real life"(tm) *sigh*.

Re:ftpmaster

[ISayWeOnlyToBePolite]

...ftpmaster as new project leader. They were the cause why the last one gave up.

Do you have any publicly available info to back that up, I've only seen refrences to an unnamed personal tragedy as to the reason of Brandens demise.

Re:ftpmaster

[wysiwia]

I don't know anything about Brandon Robinson but Martin Schulze (Debians Stable Release Manager) also resigned because of the ftpmaster (http://www.heise.de/newsticker/meldung/70616, german).

O. Wyss

Re:ftpmaster

[Anonymous Coward]

I am not happy with an ftpmaster as new project leader. They were the cause why the last one gave up.

It's an interesting paradox. A member of "the Debian cabal" was elected DPL. Personally I voted for none of the above since all of the candidates this year seemed resigned to the idea that the DPL has no power. The DPL does have power under the Debian constitution to appoint and replace delegates, but past DPLs have given in to the idea that there are no delegates...that the ftp-masters, DAM, etc are unaccountable to the DPL and by extension the Debian project as a whole.

In some ways it's a good thing that the DPL can't assert any real authority over other members of the project, but it also makes the DPL election a silly waste of time and energy and runs counter to a plain reading of the Debian constitution. If the DPL is a powerless and symbolic title we might as well make Ian Murdock, Bruce Perens or some other recognizable individual DPL for life and stop with the silly elections.
Anthony's platform was more active than the other potential DPL's, but it included plenty of caviats to indicate that ultimately he'll follow the tradition of doing very little with the position. For instance:

Another issue was that of "supporting delegates". As it turns out, that's perhaps an overly limited description, since a number of roles, including the security team and ftpmaster, might be better thought of as "infrastructure maintainers" instead, which implies a different relationship to the DPL.

And:
Some of the goals I hope to work towards in the coming year include getting updates accepted into the archive more frequently than once a day, having frequent beta releases of etch/testing that we can legitimately call a release (benefiting from the ongoing work of the installer and testing-security teams), and having reliably quick resolution of RC bugs in unstable. None of those require, or even necessarily benefit from magical DPL powers; but I think the project will benefit if whoever is elected DPL takes that idea on board, and sets a good example at making frequent and improvements to Debian.

For me, these simply aren't compelling reasons to elect a DPL. OTOH, Anthony's platform was better than what the other candidates offered.

Great interview

[Anonymous Coward]

unfortunately, it took Debian several years to release it because it had to be translated into 15 languages, including aramaic

If Debian is run by the DPL

[Beryllium Sphere(tm)]

then is Gentoo ruled by the GPL?

Re:If Debian is run by the DPL

[iabervon]

Yes, but you can change that with a USE flag.

Good luck

[upside]

Firstly, congrats and good luck to Anthony from an avid Debian user.

Having read the article and AT's campaign platform I got the sense that the project really needs not only direction, but also a leader who can steer the project while keeping people onboard and happy. This means leading the people as well as managing the project.

It seems that bickering and infighting are open source projects' achilles' heel due to strong personalities and oversensitive or overinflated egos. I hope Anthony does a good job at making the Debian team as strong as their product is already.

It's not that easy

[lanc]

One needs a stable _and_ fresh distro.
Debian is stable. period.
And bureaucratic.
and many others base their work on it and give nothing back.
and Debian moves as slowly as a rheumatic snake.

BUT

Debian is still the least bad. ( != best, that is nonexistent for years now)

Re:It's not that easy

[jZnat]

Debian doesn't really move slow at all; the only perceived slowness is in the stable distribution. If you keep up to date with unstable (which will literally always have something to update for you every day), you'd notice that they keep up to date with the majority of its software. For instance, KOffice 1.5 just came out, and it's available in Debian Sid (unstable) and thusly also available in Ubuntu Dapper (they keep their developmental releases in sync with Sid until a release-freeze starts every six months).

Re:It's not that easy

[petermgreen]

All secure-testing did is try to get security updates to those on the testing treadmill faster. you still have to either stay on the upgrade treadmill to make use of them (mixing bits from testing a few months apart can be a very bad idea).

Noone is going to provide security updates targeted at where testing was a couple of months ago.

the ONLY ways to get timely security support are
1: stay on the upgrade treadmill for testing or unstable (which is a lot of work and has fairly high risk of breaking something at any time)
2: use a stable release
3: make your own stable branch and backport fixes to it (which will be a massive ammount of work especially if you plan to do serious testing of every fix).
4: use the results of someone else who did 3 (e.g. use ubuntu)

First suggestion for the new chap:

[caluml]

You know what most pisses me off about Debian? I can't apt-get install mod_security because of their licencing issues.
I'm really glad they have principles. I really am. I admire them for sticking to their guns. But because of this, I have to jump through hoops, and use third party packages, or install the apache source packages and build against them.
It's all a lot of faffing around. Have 2 repositories. One for people that want only the most GPL'd, clean packages. And another one where they put the same packages, as well as the ones that people want.
Debian aren't going to change the world with this system, and they're just going to make it hard for people to have a complete system as they want.

Now, here come all the posts telling me "You just need to do this", or "Point your apt at this server", or x, y, z. Why not just have a setup flag or a config file - perhaps if /etc/allow-other-packages exists, it works.

Re:First suggestion for the new chap:

[jZnat]

You have included contrib and non-free in your sources.list, haven't you? Example:

deb ftp://ftp.us.debian.org/debian sid main contrib non-free

Re:First suggestion for the new chap:

[Jorgensen]

"I'm really glad they have principles."...

"Debian aren't going to change the world with this system, and they're just going to make it hard for people to have a complete system as they want."

Sorry, but you can't have it both ways. Either it's OK for Debian to have principles (and thus Debian is doing the "right thing") or Debian should forego the principles to make it easier for you to not abide them?

Logic Error. Parsing abandoned.

Re:First suggestion for the new chap:

[discord5]

I can't apt-get install mod_security because of their licencing issues.

Don't choose a distro that has that as one of their guidelines then. Debian is a great distribution if you're willing to abide by the principles on which it's built, unfortunatly sooner or later you'll find something that is missing because of it. In all honesty, the easiest thing to do is to build the package for yourself then, if you really want to use debian.

Again, this isn't a solution that works for everyone. There are time-issues, costs asociated with building those packages, and you have to keep them up to date yourself, but if you've got a couple of machines that need that package, building it once and running "dpkg -i" on several machines is a small price to pay

Now, here come all the posts telling me "You just need to do this", or "Point your apt at this server", or x, y, z. Why not just have a setup flag or a config file - perhaps if /etc/allow-other-packages exists, it works.

Really, what did you expect, this is slashdot after all. Nobody is forcing you to use debian, you know, and you can always get involved. But again, many people (like myself) simply don't have the time to get involved and build a package they can install themselves, apt it from someplace, or whatever seems like the best solution at the time.

If this is unacceptable, debian is not the distribution for you. You're better off with Redhat perhaps, or any other distro that doesn't make such an issue out of licensing. This isn't an elitist argument here, saying debian isn't for you in this case, it's simply pointing out that perhaps there are more time- and cost-effective solutions for you.

Re:First suggestion for the new chap:

[asuffield]

Obviously, it's very unlikely that he would sue Debian for distributing mod_security, and even if he did, I don't think he would be likely to win, considering it does seem like his intensions are for mod_security to be used with apache.

Since copyright violations are now a crime in many countries, Debian could be prosecuted by the government. That's either the DA or DHS in the US. The author would have little or no control over this; it would be a political decision.

That makes these things a lot trickier than they ought to be.

The "best" distro right now...

[MoxFulder]

In my opinion, the combination of Debian+Ubuntu is simply "the best" right now. I went from Debian to Ubuntu on my laptop about a year ago, and recently installed Ubuntu on my new AMD64 box.

Ubuntu is very stable, installed *almost* flawlessly (NVidia :-/) on my new AMD64 box, and best of all it's based on Debian. Also, they take a principled stand (IMHO) against closed-source software, but are more pragmatic in terms of offering closed-source packages while alternatives are developed.

Plus, Ubuntu and Debian devs interact a lot as far as I can tell, so Ubuntu is contributing to the improvement of Debian to a significant degree.

The way I see it:
* Debian is a super-stable FLOSS-only server OS
* Ubuntu is its almost-as-stable up-to-the-minute desktop OS

Neither of them is "the best" alone, but the combined strengths of the two are a knockout in my opinion.

Re:The "best" distro right now...

[lanc]

* Debian is a super-stable FLOSS-only server OS

Stable, no question about that. But always outdated.

marvin:~# for i in mysql-server tomcat4 tomcat5 postgresql ; do echo $i ; apt-cache show $i | grep ^Version ; done
mysql-server
Version: 4.0.24-10sarge1
Version: 4.0.24-10
tomcat4
Version: 4.1.31-3
tomcat5
W: Unable to locate package tomcat5
E: No packages found
postgresql
Version: 7.4.7-6sarge1
marvin:~#

yes, I am aware of apt-get source -b ... but the more machines I administer, the less magic I want to do on them. Don't get me wrong - I was/am/willbe a debian admin - just don't get overenthiusiastic.

Re:The "best" distro right now...

[Wolfrider]

IMHO, the *biggest* thing Ubuntu needs to work on, is BOOT SPEED. Bust out the parallel-INIT that we've been needing for YEARS.

On a 750MHz laptop with 384MB of Ram, it takes *several* minutes to boot into console (non-X) mode. XP takes 2 minutes on the same machine. Not a troll, I use Linux for 90% of my day-day work. But since XP boots faster, I end up booting that instead, if all I'm going to be doing is light work/Web browsing.

Re:The "best" distro right now...

[advocate_one]

the boot process is being worked on for the Dapper release. On my boxes, I can see a distinct improvement in the boot time in Dapper compared to that of Breezy.

But boot time has never really worried me as my boxes are on 24/7 and only get rebooted for a kernel upgrade. Even my daughter's laptop has got some 40+ days of uptime at the moment...

Re:That's a pretty fabricated example.

[kbmccarty]

Go download mod_security and look at the license, it is GPL.

That's actually the reason it was removed from Debian; from what I gather, it uses Apache headers that are licensed under the Apache License, which is apparently incompatible with the GPL. Here's the relevant bug: #313615

Disclaimer: I haven't done enough research to have an opinion on whether this removal was justified or not.

Not a job I'd want...

[Cumikaze]

That has to be one of the most difficult jobs in Debian. I couldn't imagine having to deal with over a thousand maintainers/developers all screaming for something different. So good luck to the new DPL!


A few things that would be good for this year:

1. Get AMD64 release into the main pool, enough already. Don't wait until December or whatever for Etch - just get it done!

2. Get security.d.o mirrored on a few more servers.

3. Try and trim the releases down to every 12 months (or less!) and drop the "when it's ready" attitude because that just drives people away.

4. (related to #3) If it's broken, don't include it, but don't hold up a release because of it - put it in "proposed updates" or something when it's fixed.

5. If it's ready for most archs, but not one (i.e. m68k) release anyway and m68k can just play catch up...

Re:Not a job I'd want...

[petermgreen]

1. Get AMD64 release into the main pool, enough already. Don't wait until December or whatever for Etch - just get it done!
afaict amd64 is almost totally built in the official sid archive now and should be making its way into etch gradually atm

the amd64 sarge is an unoffical rebuild and won't ever be part of the official archives.

3. Try and trim the releases down to every 12 months (or less!) and drop the "when it's ready" attitude because that just drives people away.
i think 12 months is a bit too fast given that the security team don't belive they have the rescources to support 3 versions of debian at once. 18 months or so is probablly about right allowing a year of security support for oldstable followed by a few months of secuirty support for testing in the run up to its release (when people will be doing test migrations etc).

Re:Not a job I'd want...

[mvdwege]

Try and trim the releases down to every 12 months (or less!) and drop the "when it's ready" attitude because that just drives people away.

I wouldn't know about that. I switched to Debian somewhere in 2002 (during the potato/woody crossover), and it definitely was a fringe distribution at that time. The big players were Red Hat, SuSE and Mandrake.

Since that day, I have seen both Debian itself and Debian derivatives like Linspire and Ubuntu making lots of headway. Debian is very popular as a base distro to start from because:

1. The main archive is fully redistributable Free Software.
2. The very strict packaging policy gives developers a stable base distro to build from.

I don't think the Debian project being very careful stops it from being popular. It might drive some people away, but I don't think that people that measure software quality by the amount of official release deadlines really are the kind of people the Debian project should care about.

If you look at those that advocate Debian the loudest, I see three distinct groups:

1. Server administrators who want a stable environment and have no requirements for the latest and greatest software.
2. Developers who require a basic stable distro to build their own software upon.
3. And a subset of 2: desktop users, both hobbyist and professional, who are willing to do or have done for them a bit of development/debugging to get a workstation install tailored to their personal needs (these tend to run unstable).

Debian succesfully meets the needs of those people, and that is all it should care about. The fanboys who want to brag about living on the bleeding edge can either try unstable, or they should move to Fedora Core or Gentoo, which specifically pander to their needs anyway.

Mart

Debian best for the enterprise

[Anonymous Coward]

Debian is the best disro for the enterprise, for it is stable (as in "doesn't crash AND doesn't change very often).

If Debian were to make major release more often than once in two (2) years then, I guess, we would have to be looking for something more stable. One release in three (3) years would probably be the best, from our point of view.