FreeBSD Foundation Newsletter, December 26, 2009
In this Edition:
In 2009, the FreeBSD project had the misfortune of losing two long
time contributors: John Birrell and Jean-Marc Zucconi. I chatted with
John recently, during this year's BSDCAN, so his death was all the more
shocking. It forced me to recognize my own mortality and to consider what
contributions from our lives remain after we pass away. Reviewing the
heritage of FreeBSD it becomes clear that our work on this project takes
on a life of its own. John and Jean-Marc's efforts live on in FreeBSD.
The value of the FreeBSD legacy has become even more apparent to me with
my return, after an almost 10 year break, to working on FreeBSD during my
day job. The kernel has the familiarity of an old friend, even though
it has been improved in countless ways by dozens of new contributors.
Although the principles and best practices engendered by the FreeBSD
project have changed little in the last decade, they are no less relevant
today. The faces and challenges may have changed, but the qualities of
FreeBSD - solid design, high performance, stability - have not.
Trying to communicate this real value of FreeBSD can be difficult.
Valuations typically take current features into account, but neglect to
consider the community that created and will support those features.
And what value can you place on future, community driven, improvements
to FreeBSD? There is no guarantee that every feature you need will be
added to the system in the time frame you need it, but the collaborative
environment created by the FreeBSD project makes it very likely.
The greatest asset of FreeBSD is our community. If we continue to
invest in the FreeBSD community, I have every confidence that the
FreeBSD legacy will endure.
The FreeBSD Foundation is doing its part to continue FreeBSD's
legacy. Through conservative management and careful planning we
have improved our financial position even when faced with the
most severe recession in 50 years. Our growing strength and
capabilities, all made possible by your generous donations, are
proof that the FreeBSD Foundation will be a faithful supporter
of FreeBSD for many years to come. Together, building on the
legacy left by John, Jean-Marc, and the countless contributors
before them, we will ensure the future of FreeBSD.
Justin T. Gibbs
President and Founder
The FreeBSD Foundation
We need just over $50,000 to meet our fundraising goal of
$300,000! We have been overwhelmed by the generosity of the FreeBSD
community. We have had 774 donors out of a goal of 1000, so far
this year. Wouldn't it be great to break our record of 1100 back
The number of donors is important for two main reasons. One, it
shows there are many users who are passionate about FreeBSD and
want to show their support. Second, it helps us keep our Public
Charity Status. It is great getting large donations from companies.
But, donations of $18,000 and under helps keep us a public charity
in the eyes of the IRS.
With the weakened economy we have been very conservative with our
spending this year. But, like each previous year we have increased
the amount we have spent on the FreeBSD Project and community. We
were blown away with the number of project proposals we received
this year. We were able to fund 7 projects this year. Unfortunately
we didn't have the budget to fund all the proposals we received.
This coming year we want to double the amount we spend on project
development. In order to accomplish this, we need to meet our
How have we spent the money this year?
Sponsored FreeBSD related conferences like BSDCan, EuroBSDCon,
AsiaBSDCon, KyivBSD, and DCBSDCon.
Provided 15 travel grants and funding to individuals to attend these
and other conferences this year.
Provided grants for projects that improve FreeBSD, like wireless
mesh support, FreeBSD TCP stack improvements, new console driver,
safe removal of disk devices, flattened device tree, and high
available storage projects.
Provided equipment for developers working to improve FreeBSD and
projects like the NetPerf cluster. We purchased servers, USB
analyzers, power controllers, and parts for computer repairs for
the Project. We also paid for shipping of equipment to various
Provided legal support for the project on issues like understanding
ownership and permission, and other legal issues that come up.
Please consider making a donation to the foundation. You can do
this by going to http://www.freebsdfoundation.org/donate/.
Project development is an integral part of our mission. Twice this
year we solicited for proposals. We were pleased to receive so many
interesting proposals. It's not always easy deciding which ones
we should accept. Besides spending the time to review each proposal,
we have to determine how this will impact the OS. Is this an
improvement or added support that we have received many requests
for? Is it a larger project that will use up more of our budget?
What percentage of users will benefit? And, sometimes we have to
bring in experts to help us understand the importance of the project.
We have four active projects that we are currently funding. We asked
the project developers to write short summaries of their projects
to help us publicize what they are doing. Sit back and take a moment
to catch up on our latest projects that are helping to improve
Even though most FreeBSD users log in to their favourite operating
system using a network login service, like OpenSSH, or use a graphical
terminal emulator, such as Xterm, almost all users are familiar with the
FreeBSD console driver called Syscons. Syscons provides lots of
functionality and works on many of the platforms supported by the
Project, but has some clear disadvantages.
One of the biggest disadvantages is the ease of understanding how
Syscons works, which has clearly caused a lack of maintainership over
the last couple of years. This has caused numerous bugs, but also a lack
of drivers for modern graphics chipsets. Therefore it only supports VGA
and VESA on i386/amd64.
The Newcons Project aims to replace the Syscons driver by a new driver
called VT. The VT driver uses Unicode internally, which means there is
no need for the user to switch between different locale settings, font
maps and keyboard maps to be able to use multiple foreign languages. It
also uses an improved algorithm to reduce graphics I/O, giving
acceptable performance even when using the slow VGA graphics mode.
Right now the core driver of VT is finished. It includes two graphics
drivers (VGA and Microsoft Xbox). The next milestone of the project is
to write a graphics driver that interacts with the kernel's DRM
subsystem to render graphics using modern graphics hardware.
contributed by Ed Schouten
Flattened Device Tree is a mechanism for describing computer hardware
resources, which cannot be probed or self enumerated, in a uniform
and portable way. The primary consumer of this technology are
embedded systems, where a lot of designs are based on similar chips,
but have different assignment of pins, memory layout, addresses
bindings, interrupts routing and other resources.
Configuration data, which cannot be self discovered in run-time,
has to be supplied from an external source. The concept of flattened
device trees is a platform and architecture independent approach
to resolve such problems. The idea is inherited from Open Firmware
IEEE 1275 device-tree notion, and has been successfully adopted by
the embedded industry (for example, it is a basis for Power.org
embedded platform reference specification, ePAPR).
The project goal is to bring FDT support to FreeBSD, so it can
benefit from a uniform and extensible way of representing hardware
resources on embedded platforms (ARM, MIPS, PowerPC etc.), instead
of multiple (duplicate) and incompatible approaches, which we have
right now. Among other advantages brought by this technology, is
the ability to use multi-platform kernels for a family of similar
boards, where a single common kernel image is supplied with just a
particular configuration data set for the given platform at boot
The project is expected to finish in February 2010. More details can
be found here:
contributed by Rafal Jaworowski
After some delays related to personal travel commitments and the
FreeBSD 8.0 release process, the project is in the final stage of
FreeBSD 8.0-RELEASE shipped with TCP Appropriate Byte Counting
(RFC3465) support and necessary changes to allow the remainder of
the project's work to be back-ported to the 8-STABLE branch. At
this stage, it is unlikely that the ABC patch will be back-ported
to earlier branches due to ABI considerations.
Candidate patches for TCP reassembly queue auto-tuning and integrating
the Statistical Information For TCP Research (SIFTR) tool into the
base system are currently under review. Public testing will be
publicly solicited in the coming weeks before the patches are
committed to FreeBSD 9-CURRENT. A decision on when to back-port the
patches will be made after some settling time has passed.
A number of additional areas for improvement have been identified
as a result of undertaking this project, which will provide an
interesting set of future work for myself and the community to
contributed by Lawrence Stewart
The HAST project goal is to implement highly available block device
(GEOM provider, using FreeBSD's nomenclature). This allows the
ability to share data between nodes in a cluster in a file system
and application independent way. The data will be synchronously and
transparently replicated over TCP/IP network between the nodes.
When a UFS file system or ZFS pool reside on such HAST block device
and primary node fails, secondary node can run fsck on the UFS file
system or import ZFS pool and start at the point where primary node
finished. No data will be lost. Once the former primary node is
fixed, modified data will be quickly synchronized.
contributed by Pawel Jakub Dawidek
70 FreeBSD developers and guests met in September, 2009 at the
University of Cambridge Computer Laboratory for two days of intensive
presentations, discussions, kernel hacking, and socializing. This
is the second time a developer summit has taken place in Cambridge,
and the largest European developer event in the history of the
project. EuroBSDCon 2009 immediately followed the summit, taking
place at nearby Robinson College, Cambridge. The event was
co-sponsored by the FreeBSD Foundation and UKUUG, the UK UNIX User's
Group. The Foundation was pleased to be able support travel expenses
for several FreeBSD developers, as well as EuroBSDCon itself.
Topics of discussion included: the 802.11 stack including new
wireless mesh networking support, jails and virtual network stacks,
a new BSD-licensed debugger, bug tracking, benchmarking, the NetFPGA
research platform, 32-bit EFI boot loader support, Apple's Grand
Central Dispatch framework, a new implementation of ATA using CAM,
FreeBSD 9.0 brainstorming, the new 8.0 USB framework, plans for new
security models, increased use of cryptography in FreeBSD, NTP and
time management, the FreeBSD.org cluster management team, a kernel
capability model, and the successful completion of the Subversion
More information on the developer summit, including slides for many of the
sessions, may be found on the FreeBSD project wiki:
The University of Cambridge looks forward to welcoming the FreeBSD
developer and user communities back to Cambridge in the future.
The FreeBSD Foundation was pleased sponsor EuroBSDCon 2009, which
took place at Robinson College, Cambridge along side a FreeBSD
developer summit at the University of Cambridge's Computer Laboratory.
The conference, 18-20 September, 2009, included one day of tutorials
and two days of technical programme, as well as a social event at
Clare College, Cambridge. The tutorial day featured the following
- Kirk McKusick's FreeBSD Overview, and a focus on FileSystems and VM
- SCTP Introduction and Workshop
Talks in the main technical programme covered a variety of topics
spanning the BSDs, including the following FreeBSD-related talks:
- Compiling a v4l DVB driver for FreeBSD
- Network event kit (NEK) - Marten Vijn
- PC-BSD - Making FreeBSD on the Desktop a mainstream reality
- FreeBSD kernel protection measures against SMTP DDoS attacks
- FreeBSD in a Complex Environment
- ISC and *BSD at the core of the Internet
- Towards a fully correct SMB/SMB2 stack for freebsd
- 100km@36Mb/s, or Long Distance Wireless
- Wireless mesh networks under FreeBSD
- Profiling and Debugging the FreeBSD* Kernel
- pfSense 2.0 and beyond
- 20,000 Ports Under The Hood
- FreeBSD and Isilon
- Complexity Attacks and Ipfw
- experiences of Google Summer of Code
- porting HPC tools to freebsd
- SuperPages in FreeBSD 8.0
- How FreeBSD finds Oil
- gemdisk: a GEOM class to emulate disk drives
- Embedded FreeBSD for industrial applications: a case study
Rui Paulo's mesh wireless networking talk presented his FreeBSD
Foundation- sponsored project.
We were especially pleased to meet many conference attendees at the
FreeBSD Foundation desk, signing up over fifty donors (as well as
equipping them with Foundation hats and flashlights).
More information about the conference, including slides and audio
recordings from many talks at:
BSD conferences are a vital venue for the BSD user and developer
communities, and the FreeBSD Foundation is pleased to be able to
support EuroBSDCon and similar conferences around the world.
It's no secret that in this post-Soviet state, the FreeBSD operating
system is well known, popular (relatively) and is in demand. Ukraine
is not last in the list of countries that have developers and users,
both existing and potential. Unfortunately, for various reasons,
companies using FreeBSD in its infrastructure, are not interested
in holding any type of FreeBSD activities. For a long time, it has
hindered the organization of such events as KyivBSD '09. But,
fortunately our foreign friends came to help! Many thanks to the
FreeBSD Foundation and iXsystems for fully sponsoring this event.
The goals of KyivBSD were: raising public awareness on the use of
free software in all areas; personal acquaintance and the strengthening
of relations between developers and users, live communication,
reports, and BSDA certification. More than 80 people signed up for
the conference, with even more wanting to sign up after registration
closed. We ended up with 60 people attending the conference, which
we still believe is a good number.
The most successful presentation was given by Alexander Drach, with
his talk on the use of FreeBSD in Cherkasy city council, it is quite
a successful start to draw attention to the opensource software in
the political sphere.
Another successful presentation was given by Konstantin Bilousov.
His presentation was on Release Engineering. This was very interesting
to the end-users who usually don't get to see the development cycle.
Almost all the visitors was very interested by Konstantin Bilousov's
report, about Release Engineering work, or rather on the part of
the work that does not see the end-users - the development cycle
All presentations were very informative, and indeed, the audience
was very attentive and engaged. Thanks to the sponsorship, we were
able to rent a room with sufficient space (which is a problem in
Kiev), print each visitor a PC-BSD disc, and provide each attendee
with some cool mementos for our first conference.
All this of course was absolutely impossible, without the sponsors
and our information partners. In our world, there is a vicious
cycle: if you don't have money, then you can not hold a decent
conference. And, if your conference is not massive, then the local
companies are not interested in sponsorship. With this sponsorship,
hopefully we have broken this cycle! We had a positive experience,
we have a story, and we now have published articles. I believe that
the next conference will be a lot easier to organize. This is how
I see the most valuable and meaningful sponsors contribution!
contributed by Alexander Yerenkow
Every year we sponsor FreeBSD related conferences, projects, and
developer travel. We believe that BSD-centered and FreeBSD-specific
conferences play critical roles in expanding the FreeBSD user
community and supporting collaborative development. Our grants may
be for something as little as performance software to large projects
like Java 1.6 on FreeBSD 7.0.
To find out how to apply for a travel grant, please visit
To get information on how to apply for a grant, please visit
Here is a list of projects, developers, and conferences we have
sponsored for 2009.
2009 Grant Recipients:
- DCBSDCon 2008 Conference
- AsiaBSDCon 2009 Conference
- BSDCan 2009 Conference
- EuroBSDCon 2009 Conference
- KyivBSDCon 2009 Conference
- Edward Tomasz Napierala - Safe Removal of Active Disk Devices
- Lawrence Stewart and Swinburne University of Technology -
Improvements to the FreeBSD TCP Stack
- Rui Paulo - Wireless Mesh Support
- Pawel Jakub Dawidek - High Available Storage Project
- Ed Schouten - New Console Driver
- Rafal Jaworowski - Flattened Device Tree Project
- Arnar Mar Sig - AVR32 Support
Travel Grant recipients for this year were:
- BSDCan - Attilio Rao, Ed Schouten, Erwin Lansing, Renato Botelho
do Couto, Thomas Abthorpe, Bjoern A Zeeb, Daichi Goto
- Ion-Mihai Tetcu - International Free Software Forum
(Fórum Internacional de Software Livre - FISL 10,)Porto Alegre,
- EuroBSDCon - Ana Kukec, Gerard van Essen, Daniel Gerzo, Ivor Prebeg,
Rui Paulo, Daichi Goto, and Lars Engels.
iXsystems is a leading provider of server class hardware designed
to run FreeBSD. An integral part of our business strategy is to
fund FreeBSD developers to continually improve and refine FreeBSD.
Our internal infrastructure and services are based on FreeBSD,
allowing us to easily identify areas in FreeBSD that need improvement.
We believe strongly that what is good for FreeBSD is good for our
customers, and what is good for our customers is good for us. The
BSD License allows us to both use and contribute code freely, and
allows our customers and the consumers of FreeBSD to do the same.
FreeBSD has benefited from years of open sharing and collaboration
which have resulted in a stable, mature, high performance operating
-Josh Paetzel, Director of Information Technology, iXsystems,
2009 Q1-Q3 Profit/Loss
2009 Q1-Q3 Balance Sheet
are posted on our website.