At: DebConf7
http://penta.debconf.org/dc7_schedule/events/83.en.html
https://debconf7.debconf.org/

Suggesting Packages based on Popularity Contest

Debian is easy to install as never before, Debian has also reached a new level of complexity with 15.000 Packages. The sheer mass makes it difficult for average and even advanced users to find programs they're looking for. To improve search a system to generate customized suggestions based on popularity-contest data is outlined.

This talk plans to give an introduction to data mining, especially market basket analysis. A prototype implementation can be found at http://dps.parkautomat.net.

Room: Upper Talk Room
Scheduled start: 2007-06-17 14:00:00

At: DebConf7
http://penta.debconf.org/dc7_schedule/events/181.en.html
https://debconf7.debconf.org/

Continuing what we started

Room: Lower BoF Room
Scheduled start: 2007-06-18 20:30:00

At: DebConf7
http://penta.debconf.org/dc7_schedule/events/56.en.html
https://debconf7.debconf.org/

What has changed since Sarge; what is going to change for Lenny

Debian Installer was first released with Sarge. This talk will give an overview of how it has performed for Sarge, of the major changes implemented since Sarge, and of the planned changes and challenges for Lenny.

The Debian Installer as released with Sarge was a completely new system for installing Debian from the boot-floppies system used with previous releases. Now that Etch has been released, this seems like a good time to look back and see how it has performed during the lifetime of Sarge.
We will also give an overview of the major changes that have been implemented since its first release for Etch and the challenges faced in that period.
Finally we will discuss the changes planned for Lenny and challenges we face to maintain Debian's official installation system. Some of these issues will be discussed in depth during BoF sessions later during DebConf7.

Room: Upper Talk Room
Scheduled start: 2007-06-17 15:15:00

At: DebConf7
http://penta.debconf.org/dc7_schedule/events/21.en.html
https://debconf7.debconf.org/

Reordering the boot sequence for correctness and speed

There are several boot options available in Debian, and the default
SysV-derived one can be tuned and corrected. The talk will present
the available options, and go into details on some of the pitfalls and
advantages of the default boot system. It will present some results
based on the Google summer of code project working on the boot system.

Room: Upper Talk Room
Scheduled start: 2007-06-17 16:30:00

At: DebConf7
http://penta.debconf.org/dc7_schedule/events/45.en.html
https://debconf7.debconf.org/

Revamping Debian's network configuration system

This talk is about netconf, a possible replacement for ifupdown, which is more suitable for today's needs. The talk outlines the motivation behind and the design goals of netconf, give a first glimpse at the implementation, provides an outlook, and solicits feedback from attendants.

netconf is a tool designed to replace Debian's ifupdown in a backwards-compatible manner while aiming for maximum extensibility and flexibility. While it can read existing interfaces files and execute hooks in much the same way, it ties in closer with aspects of network configuration
that have become more and more important in the past decade: link and location autodetection, including a framework to configure printers, smarthosts, proxies, etc; zeroconf; wireless LAN configuration; VPN; firewalling; advanced routing and traffic control; etc. Different from NetworkManager, netconf will be designed from the bottom up, allowing for its use on servers, but also providing the necessary interfaces (control pipes, d-bus) for an improved end-user experience (with polices controlling access).

Room: Upper Talk Room
Scheduled start: 2007-06-18 11:00:00

At: DebConf7
http://penta.debconf.org/dc7_schedule/events/182.en.html
https://debconf7.debconf.org/

improvements to the current system, alternatives like upstart, how to deal with a possible migration

Room: Upper BoF Room
Scheduled start: 2007-06-18 18:00:00

At: DebConf7
http://penta.debconf.org/dc7_schedule/events/51.en.html
https://debconf7.debconf.org/

http://git.debian.org has been recently added to alioth, allowing the git revision control system to be used in conjunction with alioth for package maintainance. Even though git is currently used by very high profile projects in the linux world, including the linux kernel itself and X.org, extensive experience with this tool is generally lacking in the Debian community. This BOF will provide the opportunity for people to discuss and learn how to use git to maintain Debian packages. This can include discussions of the advantages and disadvantages of using git, its distributed model, and simple tips and tricks for using it effectively. If it is generally desired, there will be a basic tutorial on using git prepared in advance.

Room: Upper Talk Room
Scheduled start: 2007-06-18 19:00:00

At: DebConf7
http://penta.debconf.org/dc7_schedule/events/209.en.html
https://debconf7.debconf.org/

Essentially the entire lintian team is here, so it seems like a good opportunity to talk over the current state of lintian, future work, and possibly interaction with the new Policy team.

Room: Lower BoF Room
Scheduled start: 2007-06-23 14:00:00

At: DebConf7
http://penta.debconf.org/dc7_schedule/events/47.en.html
https://debconf7.debconf.org/

Versioning, usertags, other new features, and common gripes

debbugs, the bug tracking system that Debian uses on bugs.debian.org
has been slowly undergoing changes to enable developers, users, and
release managers to better track bugs that affect specific releases
and versions of packages, more quickly locate bugs that affect them,
personalize the BTS to a maintainer's workstyle, and see how the
distribution as a whole is keeping up with bugs.

debbugs, the bug tracking system that Debian uses on bugs.debian.org
has been slowly undergoing changes to enable developers, users, and
release managers to better track bugs that affect specific releases
and versions of packages. Unfortunatly these additional features also
bring an increased level of complexity: no longer are bugs merely open
or closed, they can also be fixed or unfixed in specific architectures
and distributions. The versioning system will be explained; both what
it can do and what it can't, along with the adaptations maintainers
need to make in their interaction with the BTS, and some new ways of
determining why a bug is in the state that the BTS says it is.

Beyond versioning, suggested uses of usertags will be presented along
with an introduction of a few of the newer features in the BTS, like
the soap backend, full text searching of the BTS, per-bug
subscriptions, and various subtle tweaks which have been made in the
past two years.

Finally, some of the new feature enhancements that have been suggested
for the BTS will be discussed including better statistics, better
remote debbugs support, and an open discussion about new directions
and features that Developers want to see in the BTS.

Room: Upper Talk Room
Scheduled start: 2007-06-18 09:45:00

At: DebConf7
http://penta.debconf.org/dc7_schedule/events/42.en.html
https://debconf7.debconf.org/

A light hearted look from a Debian perspective of World travel

When I travelled around the World, my Debian skills came in useful. In this informal talk I would like to share some of my stories of the kind hospitality of the Debian community all over the World.

A brief description of countries travelled around the world from a Debian perspective.

General Pros&Cons for each region are provided and details of work opportunities therein.

Room: Lower BoF Room
Scheduled start: 2007-06-23 10:45:00

At: DebConf7
http://penta.debconf.org/dc7_schedule/events/33.en.html
https://debconf7.debconf.org/

About the difference between being a boss and a leader

There is a big difference between an employer - employee
and a Free Software project leader - volunteer developer
relation. In the first case making an income is the driving
force in the second the common goal to get out a working
project. In the first case fun is a nice thing to have
in the second case it is essential.

In the workshop I try to summarise some experiences
I gained in the Debian-Med project. My personal goal
was to start the project and just taking over the initial
initiative to get something happen. I had to learn that
people tend to address me as "leader" of the project -
a position I did not really intend to take over. But
I just realised that people have more fun if there
is a person that starts spreading ideas and generating
tasks that might be interesting to take over.

So being a Free Software leader is more like taking the
slack end of the red thread and drawing it in the right
direction by making sure that enough people stay close
to this line than like administrating a company with a
certain amount of employees. I'm currently in the process
of realising these facts and want to discuss these
ideas with other people to enhance the quality of
leading a project, I'm far from claiming that I did
a good job in the Debian-Med case - so it is not as if
I would be able to give an example to copy. I just
want to take the chance of this workshop to share ideas
with others to make leading Free Software projects
more effectively.

Room: Lower BoF Room
Scheduled start: 2007-06-21 17:15:00

At: DebConf7
http://penta.debconf.org/dc7_schedule/events/31.en.html
https://debconf7.debconf.org/

Do we provide what users need? Do users get what they really want? A critical view on current limitations regarding user support.

Debian GNU/Linux is a Linux distribution that is heavily used as a
base for several derivatives - there is no other distribution that has
such a large number of derived distributions. This is one reason to
make us proud because we are considered solid and rock stable. But
the question remains: Why do these people not use plain Debian
instead of building their own distribution?

Debian wants to be "The Universal Operating System" - at least
the title of the homepage says so. What means universal?
One of WordNet answers is "adapted to various purposes, sizes,
forms, operations". If we agree to this definition we have to
experience: Debian is not really adapted to various purposes but
it is perfectly adaptable.

It is not really bad if Debian is adaptable to various purposes
and the fact that Debian and even its derivatives are used as
a base for adaptations makes us proud. It is s good sign that
Debian is an unspecialised base system which is not really
modified for a particular purpose or function which makes it
attractive for adaptation to any purpose because it is known
to be very robust and stable. So in short: We do a quite good
job but it does not really qualify for the attribute "universal".

The first question is: Do we really want to be universal?
If we answer this with yes, the next question is: Can we as
a group of more or less independent maintainers afford
this? My answer to this question is yes and the method to
do this are Custom Debian Distributions. Currently we leave
most of the work that could be perfectly done inside Debian
to outsiders. The fact that outsiders constantly are working
on enhancing Debian in a certain direction means: There are
things missing in Debian and there are things our users are
needing that we do not provide.

Over several years I was constantly seeking for reasons, why
we seem to fail the task of making adaptations and derivatives
superfluous. What makes people keen on changing Debian to
fit their needs. The answer to this question goes in two
directions:

1. The current structure of Debian does not allow the
   creation of an universal, multipurpose system.
2. It is impossible to fit all needs - so why not at least
   making it brain dead easy to adapt Debian.

In the talk I try to discuss issues that might help us to
change Debian from a geekish system to something people who
consider themselves "only simple user" (whatever this might be)
want to use instead something they are afraid of. Even
if we members of the conference know that there is no reason
to be afraid of Debian ...

Note: That is not really a technical talk but it concerns
considerations about organisational structure, ways to widen the view
from single maintainer - single package (1:1) or maintaining group -
single package (n:1) development model to a customisation group -
Custom Debian Distribution (n:m) development model to offer our users
who need m packages interacting together the optimal system.

Room: Basement Talk Room
Scheduled start: 2007-06-22 11:00:00

At: DebConf7
http://penta.debconf.org/dc7_schedule/events/74.en.html
https://debconf7.debconf.org/

There are currently different Security Frameworks available in the FOSS community. Among them there is SELinux, AppArmor, Systrace and RSBAC. They all offer different concepts and security policies. In this brainstorming BoF we want to discuss about the different Frameworks, their availability in debian, and how we can integrate them in a consistent way.

Room: Lower BoF Room
Scheduled start: 2007-06-22 11:00:00

At: DebConf7
http://penta.debconf.org/dc7_schedule/events/11.en.html
https://debconf7.debconf.org/

Collecting things that would be Nice To Have

There are a few ideas out there that are mainly regarded as "nice to have", but are still lacking a bigger picture, a clear transition plan or just the proper thinking through with all their consequences. The BoF is about collecting these ideas, discussing their merits (if present) and ideas on how to implement them (if possible).

This is planned to be mostly discussion around attendee-provided content, so not much is known yet, and probably won't be before the start. There are already some ideas in the "pool", though.

Room: Lower BoF Room
Scheduled start: 2007-06-17 11:30:00

At: DebConf7
http://penta.debconf.org/dc7_schedule/events/118.en.html
https://debconf7.debconf.org/

A set of analysis and statistics about the incoming Debian 4.0 release

Debian is possibly the largest free software distribution, with more than 10,000 source packages
in the release currently in preparation (Debian 4.0, codename Etch). However, we wish to know what
these numbers mean. In this paper, we use David A. Wheeler's SLOCCount
system to determine the number of physical source lines of code (SLOC)
of Debian 4.0. We conclude that
Debian Etch includes more than 280,000,000 physical SLOC (much more than other
contemporary GNU/Linux distributions such as Fedora Core 6),
showing that the Debian development model (based on the work
of a large group of voluntary developers spread around the
world) is at least as capable as other development methods
(like the more centralized one, based on the work of
employees, used by Red Hat or Microsoft) to
manage distributions of this size.

Room: Lower BoF Room
Scheduled start: 2007-06-17 17:30:00

At: DebConf7
http://penta.debconf.org/dc7_schedule/events/16.en.html
https://debconf7.debconf.org/

Introduction

OpenStreetMap is a project aimed at creating and providing free geographic data such as street maps to anyone who wants them. The project was started because most maps you think of as free actually have legal or technical restrictions on their use, holding back people from using them in creative, productive or unexpected ways.
The idea is to get people all over the world who wish to contribute have group walks to map out streets and areas using GPS.

Much of the same as the above, I have no more input than that at this point.

Room: Basement Talk Room
Scheduled start: 2007-06-18 18:00:00

At: DebConf7
http://penta.debconf.org/dc7_schedule/events/44.en.html
https://debconf7.debconf.org/

Researching into the volunteer nature of Debian

This presentation is about ongoing research on innovation diffusion in the Debian project. The goal is to determine the conditions under which volunteers adopt new approaches to everyday challenges. The talk illustrates the concepts behind innovation diffusion research, introduces the assessment framework used, gives a short overview of the research approach, and offers some of the results to date.

The Debian project is a very complex social system, and its volunteer nature and associated inertia can have quite a dampening effect on the adoption of innovative techniques aiming to improve cooperation and increase efficiency. However, Debian is not the only project confronted by such issues; studies of how to diffuse novel methods into existing, complex social systems are
numerous and diverse. The study looks at past diffusions within Debian and other projects in an attempt to provide a prescriptive framework to guide future diffusions and increase their success.

Room: Upper Talk Room
Scheduled start: 2007-06-21 11:00:00

At: DebConf7
http://penta.debconf.org/dc7_schedule/events/23.en.html
https://debconf7.debconf.org/

How to add Debtags to your everyday Debian work.

Debtags is mature for general Debian use and this talk marks the launch of the official Debtags reference for the entire project. The talk will cover, among other things, how to use the tag data, how to tag packages, and some lessons learned during more than 3 years of Debtags development.

Room: Upper Talk Room
Scheduled start: 2007-06-21 15:00:00

At: DebConf7
http://penta.debconf.org/dc7_schedule/events/122.en.html
https://debconf7.debconf.org/

including: Conspiracy, dissent and FUD, Balloon Open Hardware Board & Licence and Why tracker is cooler than unsliced cheese.

This talk session will be split into five minute segements for lightning
talks on a variety of subjects from different speakers. Lightning talks have
long been a staple of Perl conferences and offer a way to get an idea,
technique, interesting piece of software, etc out to the audience without
all the overhead of a regular talk; now let's try them at DebConf.

The five minute time limit will be strictly enforced. Hopefully we will
manage to have 6 or 7 talks in this time. Anyone who would like to give a
talk should contact the organiser. We will have a rehersal session at some
point beforehand.

Notes to organisers: A fixed mic will be needed for the talks. The Organiser
will handle timekeeping reminders, etc. Use of laptops will be discouraged.
The track is "technical" just because most of them probably will be. It's
possible that we might conflict with speakers in some other track since all
the lightning talks won't be known in time; I'll try to have an alternate
lined up in case this happens.

Room: Upper Talk Room
Scheduled start: 2007-06-23 15:45:00

At: DebConf7
http://penta.debconf.org/dc7_schedule/events/12.en.html
https://debconf7.debconf.org/

Fundamentals of Software Defined Radio and Practical Applications of GNU Radio

Room: Upper BoF Room
Scheduled start: 2007-06-23 14:00:00