www.nettime.org
Nettime mailing list archives

<nettime> Hive networks: meshing in the future
Armin Medosch on Sat, 25 Feb 2006 11:27:26 +0100 (CET)


[Date Prev] [Date Next] [Thread Prev] [Thread Next] [Date Index] [Thread Index]

<nettime> Hive networks: meshing in the future


[This text has been written for Media Mutandis - 
the Node.London Reader, edited by Maria Vishmidt 
et al. The reader is available online and as a 
Print on Demand publication: 
http://publication.nodel.org/Publication. The 
research about and writing of text has been 
supported by media art laboratory Graz which also 
runs http://theoriebild.ung.at/ (theory under 
construction) where you can find more of my 
writing in English and German.] 


Meshing in the Future - The free configuration of 
everything and everyone with Hive Networks
Text by: Armin Medosch, London/Vienna July - Dec 
2005

Table of Content
1. Introduction	
2. From OpenEmbedded to HiveWares	
3. Hivewares: The Self-Managed People's Net	
4. Conclusions	
References	
Glossary	



[Note: Words or acronyms marked with an * are 
explained in a glossary appended to this text.]

1.Introduction

One day in spring 2005 I popped over to my friend 
Adam's. In his tiny living room, which also 
serves as the headquarters for free2air 
(www.free2air.org), I found another friend, 
Alexei, hunched over a small technical device. 
Its case had been removed and the circuitry of 
the board and the chips could be seen. It had a 
small hard drive strapped to the back of the main 
board. Alexei and Adam were trying to make the 
thing boot from the hard drive. They were so 
focussed that I barely managed to get noticed 
when I said hello. Slightly daunting technical 
buzzwords such as 'cross compilation', and 
'zeroconf' where flying through the room. Not all 
of this meant something to me at the time but 
what I could figure out was that they were on to 
something special. This little thing on the table 
represented the seed of an idea much larger than 
its petite techno-crab like self. 

Over the last two decades free/libre and open 
source software (FLOSS)* has  provided accessible 
means for people to write their own software, 
encompassing creative, educational and 
professional uses. In the meantime, IP 
(Intellectual Property)* regulations have become 
a battleground. While the proprietary monopolies 
marshal an army of lawyers and policymakers to 
fight their 'battle', the FLOSS communities have 
responded by creating realities on the ground. In 
a quite distinct manner, ever more areas of 
computing have become 'free'. The FLOSS universe 
is an expansionary one. 

It has been very interesting to watch how the 
free software and open source software 
communities have creatively made their 
'investment'. While some FLOSS projects are 
supported by companies and universities, many 
projects remain outside such institutional 
context. There is no formal structure to decide 
which projects are taken on. Many free software 
projects exist only because developers make a 
personal commitment to them. Thus, the investment 
is a highly personal one. The communities have 
nevertheless been able to make wise decisions 
expanding on existing building blocks. First, the 
tools to build tools were released from the 
corporate lockers. Then, the PC was liberated by 
being given a range of free operating systems. On 
top of that the internet boom of the nineties has 
been built, with most of the services and 
applications such as email and web servers 
powered by FLOSS. What is happening now is that 
the same versatility of the universally 
programmable machine is needed in smaller 
devices, in those digital technologies which 
pervade our environments anyway, from the car to 
the mobile phone, palm computer or home 
entertainment system. In industry terms this is 
called ubiquitous or pervasive computing.

Salesmen-gurus like Nicolas Negroponte have been 
telling us about the merits of 'being digital' 
(Negroponte 1995) for two decades now, which 
implies that computer chips need to be 'embedded' 
in the environment. What usually usually gets 
left out of the marketing is that the world of 
embedded computing* is also the world of embedded 
capitalism, where everything is done by large 
corporations whose systems are by default more 
secretive than the mafia. The chips which are 
used in embedded systems are different from the 
chips used in PCs. For developers to be able to 
make use of those chips it was necessary to buy 
very expensive licences or to engage in time 
consuming and difficult reverse engineering. This 
provided huge obstacles for 'free' developments 
in this area. Assuming there was once a well 
meaning vision by computer engineers of  
'augmenting' reality with smart devices, this 
dream may long have been highjacked by corporate 
ambitions to sell more hardware. At best  it 
promised 'intelligent' homes where fridges and 
toasters would all communicate with each other; 
at worst clouds of smart dust* would conduct the 
remote controlled warfare of the 21st century. 
Technology, in those 'visions', is meant to strap 
people into a regime of consumption and control.

Embedded Capitalism also means that those 
developments are driven by an industrial logic of 
higher volumes of cheaper goods sold at lower 
margins. In the digital world this is often 
euphemistically referred to as Moore's Law*. The 
mundane economic aspect behind the spectacular 
growth of processor speed and memory capacity is 
the need of producing and selling chips in very 
large numbers to make a profit, because the 
initial investment costs are very high. This 
logic has - maybe oddly enough - benefited the 
FLOSS community by giving it cheaper toys to play 
with. What is happening now is that embedded 
computing gets 'liberated'. The name of the game 
is to replace the firmware* of small devices - 
from wireless routers to palms to practically 
anything that beeps - with trimmed down Linux* 
distributions*. Once the operating systems of 
those devices have been replaced with free ones, 
their functionality can be rewritten to perform 
other types of services. Embedded computing 
becomes transparent and may, eventually, reflect 
the needs of ordinary people instead of 
shareholders. 

Enter the HIVE Networks (www.hivenetworks.net/) 
project. Devised by Raylab (www.raylab.com) and 
affiliates, I had witnessed the development of 
this project in its very early days during that 
first magic afternoon. Hive Networks was 
initiated by Alexei Blinov, Vladimir Grafov and 
Ciron Edwards of Raylab, supported by other 
developers such as Bruce Simpson, Adam Burns and 
a growing network of non-techie supporters such 
as Ilze Black, James Stevens, myself and others. 
Raylab brings a particular experience to the 
project. Blinov and Grafov, both originally from 
Russia, have been working as artist/engineers (or 
engineer/artists) for many years now, often 
technically supporting the work of artists such 
as Eric Hobijn and Atau Tanaka. After his move to 
London Blinov worked with the group Audiorom.  
Their interactive sound art works won the BAFTA 
Interactive Award, and  Blinov's electronics 
skills played no small part. More recently, with 
Take2030, Blinov helped to create the Lunchboxes. 
Those boxes, whose cases consist of typical 
Japanese Bento boxes, contain a fully functional 
miniature computer running Meshlinux as an 
operating system. They are capable of connecting 
to each other and other computers via WLAN* on 
the fly by using the OSLR* (www.olsr.org/) 
dynamic routing protocol. In many ways the 
Take2030 lunchboxes have been an important step 
towards Hive Networks. 

Finding solutions for those art projects usually 
involved a lot of tinkering and risky actions 
with the soldering iron. Another important 
background influence was the involvement with 
London's nascent free network* movement which - 
under the banner of Consume (http://consume.net) -
 in the early 2000s highlighted the possibility 
that people can create their own networking 
infrastructure by using WLAN technology of the 
802.11 standard family and in the 2.4 GHz 
spectrum (cf. Medosch 2004). In those early days, 
old computers were often fitted with Linux, then 
used and repurposed as wireless access points 
(AP)* and routers*. But an old computer is still 
an old computer, which implies that it has many 
ways of breaking down. As Blinov pointed out in 
conversation, with Hive Networks the days of the 
soldering iron and of crappy old computers are 
over.

For Raylab and affiliates the goal is now to work 
with state of the art hardware which is produced 
in industrial quantities and whose design follows 
widely established industry standards. Usually 
these devices use solid state computing, that 
have no moving parts which could mechanically 
break and fail. Liberating or repurposing such 
devices signals nothing less than a paradigm 
change in creative computing. This time it is not 
the artists asking technicians for a creative 
solution, it is the engineer/artists who are 
proposing a framework for which artists and other 
media practitioners are asked to come up with 
project ideas. Hive Networks transcends the 
boundaries between engineering and art. It is a 
work of art as well as a platform for other 
artists to create works. Most importantly, it 
combines the element of content with the element 
of networking. 

Each Hive device is capable of gathering content 
(through webcams, microphones, sensors) and 
disseminating it (web server, audio/video live 
streams, bluetooth*, WLAN). At the same time each 
Hive device also acts as a node in the network, 
which means that it is capable of storing and 
forwarding data. The conjunction of those two 
elements means that the perception of the network 
as such changes. The network is no longer only a 
connectivity structure through which access to 
the global internet is facilitated, but it 
becomes a content structure, a hiving network of 
desires and cultural creations. An additional 
motivation is the urgency to open up the world of 
embedded computing and make it available to the 
highest possible number of people. So much for 
the concept, now to the realization.

2. From OpenEmbedded to HiveWares

FLOSS developers have found ways of replacing 
company firmware with custom Linux firmware on a 
number of devices now, specifically product 
families by Linksys, Netgear, Asus and others. 
The meta-tool Open Embedded and distributions 
like OpenWrt make it easier to open those gadgets 
and install applications customized to individual 
needs. With Hivewares Raylab adds a particular 
flavour to the orchestra of voices. What is now 
only possible for serious geeks should become 
part of everyone's lifeworld. 

One of the first items to draw the attention of 
the community was the Linksys WRT 54G, a 
broadband router and wireless access point. 
Harald Welte is a Linux kernel developer from 
Berlin who is deeply involved with the 
Iptables/Netfilter project which adds security 
features. Welte had discovered that a number of 
companies who sold WLAN equipment had based their 
firmware on Linux. As Linux is protected by the 
GPL*, the terms of this licence make it mandatory 
to release the source code* of any software based 
on it. Companies such as Linksys, Sitecom and 
Fujitsu-Siemens who sold their Linux-based WLAN 
devices had for one reason or another 'forgotten' 
to make the source code accessible. The Free 
Software Foundation (FSF), who is actually 
safeguarding the GPL, had traditionally been 
reluctant to take violators to court. But Welte 
sought the help of lawyers and started  GPL-
violations.org, a project which sent legally 
backed warning letters to GPL violators. Welte's 
initiative succeeded also in court, in a landmark 
case in Germany against the company Sitecom. 
Subsequently it became clear that the GPL was 
more than a well meaning declaration of intent 
and that it really was a legally binding licence 
agreement. Industry giants such as Fujitsu-
Siemens settled out of court  and complied with 
their obligation to release the source code. 
Linksys, confronted with similar allegations, 
slowly and reluctantly released the source code 
of the WRT 54G. This opened the floodgate for a 
range of firmware hacking projects. 

Replacing the firmware of a device such as the 
WRT 54G with Linux-based firmware is of great 
advantage. Not only does the way of working of 
the device become transparent, it also unleashes 
the full spectrum of its  capabilities. Usually 
manufacturers restrict the functionality of 
devices to what they think that consumers need. 
And specifically in the low cost or 'consumer' 
market there seems to be an assumption that 
people would not want to or should not have the 
ability to tinker. By replacing the firmware a 
device which was meant to be a relatively stupid 
AP only could become a web-server or a hub for 
internet telephony (Voice over IP* or VOIP) - in 
other words, anything that anyone might possibly 
imagine it to become within the limits of 
existing technological development. 

The legal hacking of the WRT 54G brought  the 
OpenWrt project (http://openwrt.org/) to life and 
aimed at facilitating the making of custom 
firmware. OpenWrt is a Linux distribution for a 
range of wireless routers. It provides only a 
minimal firmware - just what is necessary to boot 
the device and provide its most basic 
functionalities. Its key feature is that it 
allows users to add and manage packages*. Users 
can custom tune their AP, they can remove 
unwanted packages and add packages they like. 
Developers don't have to deal with the 
intricacies of the hardware to create a whole 
firmware of their own but can focus on developing 
useful packages instead. 

Highly skilled developers from the free network 
community have put OpenWrt to good use. For most 
ordinary humans OpenWrt is still quite a scary 
bit of software which can only be controlled via 
the command line interface. Sven Ola Tuecke from 
the c-base and Freifunk (www.freifunk.de) 
community in Berlin has put together the Freifunk 
Firmware. It is based on OpenWrt but offers a web-
like interface for customization and 
administration so that less skilled users can 
also make a proper free network node. Elektra, 
another Berlin based network wizard, has worked 
on improvements of OLSR and its inclusion in the 
Freifunk Firmware. Now dozens of nodes and hubs 
on the roofs of Berlin create an elegant mesh 
network which largely maintains itself and 
shovels around bits and bytes outside the 
networks of corporate greed and state 
surveillance. 

Naturally, the WRT 54G did not stay the only 
liberated hardware device. Under the banner of 
OpenEmbedded (http://oe.handhelds.org/) there is 
a development under way to make it easier to 
"bake" custom Linux kernels for potentially a 
very large number of devices. A hairy issue on 
any PC under Linux is the compilation of source 
code to make it work with a specific hardware. 
With embedded devices the added difficulty is 
that the source code needs to be compiled on 
another platform first and then installed on the 
device. This is called cross-compilation and is 
one of the most difficult areas in contemporary 
computing. OpenEmbedded has created a tool named 
BitBake to make cross-compilation work. The 
project is in its early stages and follows an 
almost utopian meta-level strategy, but some 
branches already show signs of success. Out of 
the original OpenEmbedded effort came the 
OpenSlug (www.nslu2 
linux.org/wiki/OpenSlug/HomePage) development 
which tries to make a truly open source custom 
kernel (kernel 2.6) for the NSLU2 (Netgear 
Network Storage Link Usb 2). The NSLU2 is 
particularly interesting because it works with an 
external HD and it can be made to run on 
batteries. You can have a web-server on a 
wireless battery driven device. People could make 
mesh mobile networks and do VOIP - internet 
telephony - completely for free on their own 
community network.  

3. Hivewares: The Self-Managed People's Net 

Blinov and Grafov watched those developments 
carefully and decided to work with another 
product family, the WL series by Asus. Custom 
firmware development for those devices is 
supported by a lively community called the WL500g 
Forum (http://wl500g.info/) which basically 
thrives around "Oleg's firmware". 

Oleg is a Russian guy who rewrote Asus firmware 
for the WL-series of products (WL500g, WL300G, WL-
HDD) and added lots of useful stuff to it, 
including the possibility to use the root 
filesystem from an external drive (either USB 
flash or IDE, in case of WL-HDD). (Grafov 2005)

Blinov and Grafov have put Oleg's Firmware on the 
WL-HDD2.5. This little box which I had seen first 
during that magic afternoon is now available for 
around 50GBP. Like the WRT 54G it supports both 
WLAN  and ethernet connections on top of which it 
also offers an IDE connection and USB 1.1. Both 
IDE  and USB allow the connection of an external 
HD which is crucial for expanding the capacity 
and adding features. Raylab spent quite a few 
afternoons making the WL-HDD boot from an 
external drive and adding a few other essential 
functions. 

What we did is that we used his [Oleg's] firmware 
with its built-in possibility of adding packages 
as basis and added some features that make it 
possible to run Hivewares. Hivewares are self-
contained "product personalities" that make sense 
to a non-geek person. Without Hivewares, a non-
techie could probably still get the same 
functionality from his/her box, but only after a 
lot of painful seeking through many different 
sources of information and forum postings. 
(Grafov 2005)

After an initial project presentation at the 
media art lab in Graz, at WSFII prepconf 05 
Raylab were able to show such Hivewares in 
action, by presenting a WL HDD and a number of 
different pre-packaged configurations on Compact 
Flash drives. By replacing the Compact Flash card 
the primary function of the device is changed, it 
could either be a web server or a web cam, a net 
radio player/receiver or a wireless media 
jukebox. They have also been conducting 
experiments with the WL 500Gx which is very 
similar to WL-HDD but even better equipped with 
plugs connecting it to the outside world. With 
Hiveware the little Asus boxes become freely 
configurable devices. A number of Hivewares are 
already downloadable from the Wiki. 

The Hiveware developers put particular attention 
to a concept known as Zeroconf, called Bonjour in 
the Apple world. 

Addition of Bonjour and linking of Hivewares 
personality to service advertisement supported by 
it, made it possible to have hassle-free 
discovery of Hive devices in the neighbourhood of 
supported clients (Windows and Mac running 
Zeroconf client software). (Grafov 2005)

By including Zeroconf/Howl, Raylab hope to 
overcome the carrier/content dichotomy. The 
network becomes more than just a carrier medium, 
it also identifies and advertises 'services' in 
the vicinity or network-neighbourhood of a node. 
People are no longer getting access to an 
anonymous world wide web but connect to content 
and services which reflect their (local) 
interests. Last not least Raylab are 
experimenting with further interfaces such as 
bluetooth, FM radio and a break-out box, which 
has analogue-digital switches, so that sensors, 
for instance,  can be connected to a box. 
Participants in the Hive Network could 
potentially have their own meteorological 
environmental station.

In summary, what Raylab have been trying to do is 
to make the process they were going through last 
year over a period of several months as hassle-
free as possible for other users. Alexei Blinov 
wants to make "information processing truly 
accessible without usurping human space." "Just 
like bees and ants and other social insects," 
Blinov says, "those devices are living in 
symbiosis with people rather than presenting 
problems that demand a lot of dedication to find 
solutions." (Blinov 2005)

Ideally they would like to offer the 
customization of devices on a web platform. Users 
first need to buy the hardware, a common device 
available through many stores such as the WL-HDD. 
Then they come to the web-site, where they can 
choose how to configure their Hive device by 
clicking radio buttons on a web form. Once 
finished with this, a specific version of the 
software is compiled. Users download the compiled 
software and install it and are ready to fire up 
their Hive device and join the network. For 
accomplished Linux users this is already 
possible. 

In the interests of minimising the obstacles for 
users at every level of expertise, Bruce Simpson, 
BSD developer and friend of Raylab, has 
experimented with OpenEmbedded and BitBaking. As 
OpenEmbedded is still in an experimental stage, 
there is some way to go. Currently it is only 
advisable for people with some knowledge of 
Linux/Unix to get hands-on involved. For those a 
Hiveware compilation is envisioned which consists 
of a Linux image with a built-in packaging system 
(ipkg), Zeroconf (Rendevouz/Bonjour) service 
advertisement and discovery protocol, the 
standard Linux command line toolkit (Busybox) and 
a PHP-based web interface. Thus, more 
accomplished users who know some PHP and 
Javascript are able of developing application 
interfaces without having to go into hardware 
hacking. As an example, Blinov recently strung 
together a nice interface which turns a WL-HDD 
into a net radio receiver, but any sort of other 
web application development is possible.  

Because, after all, the chipsets inside the WL-
HDD are not that powerful, what Raylab have in 
mind is that each device can do one thing very 
well, but one only. So for instance a WL-HDD can 
be turned either into a video streaming server, 
or an Internet radio tuner, or a music jukebox 
and Internet radio tuner in one, or an audio 
streaming server which converts audio input 
(line/mic in) to a live-stream on the net. It can 
not perform all those tasks at the same time but 
it can do it each at a time. Because the 
individual devices are quite cheap, large numbers 
could be spread out over the cityscape to work 
together. What makes the Hive really buzz  is not 
just the price but also the added network 
capacity. Raylab intend to make each device 
capable of joining ad-hoc networks*. Each device 
creates a wireless cloud of potential network 
connectivity around itself and seeks to link up 
automatically with other devices. The point is to 
make this really work  automatically. If 
successful, a sort of Trojan Horse strategy could 
be played out. If a technophobe - an aged parent, 
for instance - can be persuaded to use a Hive 
device, which is as easy to use as a radio 
receiver or CD player, it will also potentially 
become part of a free network. If adoption of 
such devices is widespread, local free networks 
can connect together and large scale community 
owned wireless free networks finally become 
reality. What remains to be resolved is how 
exactly this is going to be made to work with 
Hivewares. As mentioned above, free network 
developers in Berlin and elsewhere have 
experimented very actively with ad-hoc mesh 
networking* protocols such as OLSR. Those have 
been tried and tested now with 90 clients and 
more forming a mobile mesh network. It looks like 
Raylab is aiming at something similar and will 
include OLSR into its Hivewares. But the 
scalability of mesh networking up to areas of 
1000 nodes and more remains to be proven.

There is a host of other potential points of 
criticism, and not just technical ones. The 
development of the free network community has 
shown that those projects make only slow progress 
in areas which are covered by affordable ADSL 
broadband offers from commercial Internet Service 
Providers (ISPs). The finer points of the 
political difference between commercial centrally 
controlled networks and community networks just 
do not seem to matter for the majority of people. 
The thrill of becoming a content provider on the 
community network is felt most strongly by the 
younger and more net savvy ones. The free network 
community has also focussed so far mainly on 
making the networks work and cared little for the 
content. There remains a pronounced gender gap in 
the demography of such groups. Those issues are 
known to be difficult to overcome. Even if Hive 
developers solve all the technical problems we 
will have to wait and see if Hive devices will be 
adopted by large numbers and a diverse range of 
people. 

 4. Conclusions
There remains the potential criticism that Hive 
devices add only to the flood of digital gadgets 
which already threaten to become an environmental 
hazard, as SF author Bruce Sterling pointed out 
at his Siggraph key note speech in 2004. This 
could be countered by the claim that Hive devices 
will be the last gadget that anyone will ever 
need because one and the same piece of hardware 
can serve different purposes. Ideally, new 
functions can easily be downloaded and installed 
with a one-click process. But isn't this the same 
sort of techno-utopianism which is a generic part 
of the marketing blurb of the ICT industry? Is 
there really a connection between the intrinsic 
properties of this or that technology and 
desirable forms of social change? Those are big 
questions which cannot be answered within the 
limits of this text. They are also real questions 
in that sense that they do not offer themselves 
to be answered by simple or reductive statements. 
However, it is significant that the Hive Networks 
project poses those questions in a new and 
intriguing way. Hive Networks may well fail as a 
techno-utopian project if it formulates its 
objectives on a generic and universal level. It 
has a much better chance to make any impact if 
the technological development gets embedded into 
the community and gets driven by the situated 
knowledge of people to whose needs the project 
responds.

In ubiquitous computing it is usually the devices 
which get smarter and the people who remain 
stupid. So far the concept of 'pervasive' 
computing sounds like a threat to ordinary 
people: another layer of technology which remains 
unseen, little understood but potentially 
influences and controls the life of many. By 
merging the concepts of FLOSS, DIY and embedded 
computing, Raylab threaten to turn that trend 
around. 

There can be no real conclusions with regard of 
Hive Networks at this point. The project has made 
some achievements but is still in its early 
stages. After initial good responses from 
different sides - artists, developers, 
institutions - it appears that the developer 
community needs to grow to take it to the next 
level. It would be good to see some exemplary 
projects get off the ground to illustrate the 
concept. To this end, Hive developers are about 
to launch a number of collaborations with artists 
and media art institutions in Britain and abroad. 
The public needs to see what happens if swarms of 
Hive devices are set free. Otherwise the concept 
remains too abstract for most people. 
 
Acknowledgements: This text benefited 
substantially from comments by Alexei Blinov, 
Vladimir Grafov, Adam Burns and Elektra. 

----------------------

References

Blinov, Alexei, 2005. Interview with Armin 
Medosch. Private notes.

Grafov, Vladimir, 2005. Email conversation with 
Armin Medosch. Private Notes. 

Medosch, Armin, 2004. Society in ad-hoc mode. In: 
ECONOMISING CULTURE: ON 'THE (DIGITAL) CULTURE 
INDUSTRY'. eds. Cox et al. Plymouth and New York: 
Autonomedia (DATA browser 01). Available online 
from: www.ejhae.elia-artschools.org/Issue2/2a-
medosch.htm   (last accessed January 2006) 

Negroponte, Nicholas, 1995. Being Digital. New 
York: Alfred A. Knopf. 

Sterling, Bruce, 2004. When Blobjects Rule the 
Earth [online]. Conference Speech, Siggraph, 
August 2004 Available from: 
www.boingboing.net/images/blobjects.htm  (last 
accessed January 2006)
 
-------------------

Glossary

Access point (AP) - a device which allows a WLAN 
client - for instance a notebook with a WLAN 
interface card - to connect to the AP and the 
internet. Such a set-up is also called a hotspot. 

Ad-hoc network - a network which uses mesh 
network technology (see mesh networks). WLAN 
technology uses also a so called ad-hoc mode, 
which is a specific way of configuring an AP or 
wireless network card.  

Black box - a device whose inner working stays 
hidden to the user.

Bluetooth  - is a network technology which works 
at very close range.

Boot - a process by which a computer is 
'bootstrapping' itself, starting up the system 
and checking its main system devices. 

BSD - Berkeley System Distribution, first 
released in 1977 by Bill Joy. At Berkeley 
University and other campuses students and post-
graduates worked on improvements of the AT&T 
operating system Unix. Most significant 
contributions were the inclusion of the Internet 
Protocols (IP) in Unix and the BSD licence. 
Besides Linux, BSD is another stream of how a 
version of Unix became 'free'. Since the closure 
of the research group at Berkeley, University of 
California, BSD lives on through the three follow-
up projects NetBSD, FreeBSD and OpenBSD. 
 
Embedded computing - describes a type of device 
where hardware and software form a very close 
unity. It is used primarily in large industrial 
systems such as traffic systems or power plants, 
but also in consumer communication devices which 
present themselves as a black box.

Firmware - is the software which comes pre-
installed with consumer devices; it contains a 
software which is specifically written for the 
type of hardware it runs on. 

FLOSS - stands for Free Libre Open Source 
Software. The inclusion of 'Libre' signals that 
the word free is used as in 'freedom' and not as 
in gratis. 

Free Network - a computer network which is 
neither owned by the state nor by a commercial 
company but by the people who create, maintain 
and use it. 

GNU - stands for GNU is not Unix, the sort of 
joke programmers like to make who are used to 
recursive structures. The GNU tools and libraries 
have been developed by Richard Stallman and the 
Free Software Foundation (FSF) since the early 
1980s. GNU made possible the development of 
Linux. The licence which protects GNU software, 
the GNU General Public Licence, has since been 
widely adopted and is the pillar of FLOSS 
development.

GPL - General Public Licence. The 'free' in free 
software is safeguarded by a specific software 
licence, the General Public Licence (GPL) which 
is maintained by the Free Software Foundation 
(cf. FSF 2006). The legal and normative basis of 
FLOSS are enshrined as four freedoms in the GPL. 
These are: freedom to use a work, freedom to 
change it, freedom to distribute exact copies of 
it and freedom to distribute adapted copies. 
These freedoms are made practicable through the 
obligation to provide the necessary resources - 
for software, this is the human-readable source 
code.

IP as in Internet Protocol - short form for a 
family of internet protocols at the core of which 
is TCP/IP, the protocols on the network layer 
which facilitate the receiving and sending of 
'packets' of information. Other internet 
protocols are for instance  SMTP (for email) and 
HTTP (for web). The technical details of IPs are 
documented in Requests For Comments (RFCs) which 
are stored publicly on the net (www.rfc-
editor.org/). 

IP as in Intellectual Property - is corporate 
language to describe intangible goods. The term 
is controversial because it implies that all 
fruits of intellectual and creative labour are 
commodities. 

LAN - Local Area Network, a cluster of computers 
connected locally. One of the most widely used 
LAN technologies is Ethernet, invented by Bob 
Metcalfe. 

LINUX - is an operating system which is very 
similar to Unix. It has been created by Linus 
Torvalds using the GNU libraries and tools which 
is why some insist it always should be called 
GNU/Linux. 

Linux distribution -  a specific version of the 
basic Linux operating system plus additional 
packages. Distributions are compiled for a number 
of reasons and often to make particular tasks 
easier or to make Linux run on specific hardware.

Meshlinux -  is a variation of the free operating 
system GNU/Linux which supports mesh networking. 

Mesh networks  - are highly distributed networks 
which use special routing technology. In standard 
routing technology as used to send and receive 
information via the internet the 'routes' which 
data packets take are fixed. In mesh networks the 
software decides 'dynamically' or 'ad-hoc' which 
route data packets take. Sometimes 'mesh 
networking' and 'ad-hoc networking' are used as 
synonyms.  In wireless and mobile networks mesh 
networking has the obvious advantage that the 
software adapts dynamically to changes in the 
structure or 'topology' of the network. There are 
a number of routing protocols which support mesh 
networking amongst which OLSR is one of the most 
advanced and most widely used ones. 

Moore's Law -  a prediction by a former IBM 
director that the speed of computer chips of the 
same price would double every 18 months. There is 
no 'law' behind this formula in any scientific 
sense but so far the prediction has held or been 
surpassed. 

Node - a computer which is fully integrated in 
two way communication on the internet and is not 
just an end-point or 'leave'. In free network 
terminology a node usually combines the 
functionality of a router and an AP. 

OLSR - Optimized Link State Protocol, a routing 
protocol for mobile mesh networks.

Packages - are programmes in Linux-speak, for 
instance services or applications. 

Packet - to send and receive information on the 
internet, it is split up in so called packets, 
whereby a single packet is also called a frame. 

RFC - Requests For Comments (RFCs), a set of 
technical and organisational notes on the 
Internet Protocols stored publicly on the net 
(www.rfc-editor.org/). 

Router - a computer which transfers packets of 
data between networks (routing). The decision 
where to send packets is based on entries in 
routing tables which reflect knowledge of the 
structure of the networks involved. In mesh 
networks the routing tables are not fixed but 
updated frequently by an automatic process. 

PHP - scripting language which is widely used for 
interactive web applications. 

Radio buttons - buttons on a web form which can 
be clicked either on or off. 

Router - a computer on the net which send and 
receives packets of data on the net. 

Smart dust - is an experimental technology 
developed by contractors of the Pentagon where 
very small devices gather information and 
communicate.

Source Code - is the human readable form of 
computer code rather than just the machine-
readable binaries, consisting of nothing than one 
and zeros. Source code needs to be compiled in 
order to run on a machine. Vendors of proprietary 
software do not give out the source code so that 
the functions of a programme can neither be 
checked nor changed. 

Trojan Horse - appropriation of the ancient Greek 
saga to the computer world; usually means a 
software which hides its true purpose and is 
installed by users without knowing. Sometimes 
used for malicious reasons such as spreading 
computer viruses or forwarding personal 
information through a 'back door' in a computer. 

WLAN - is the acronym for Wireless Local Area 
Network and is called WiFi in marketing language. 
It is based on a family of standards by the IEEE 
which all start with the numbers 802.11 (a, b, c, 
etc.). The technology operates in a band of the 
electromagnetic spectrum which according to 
international conventions has been made licence 
exempt, which means that everybody can use it 
without having to ask for permission first.  One 
of the licence exempt spectrum bands is at and 
above 2.4 GHz. 






#  distributed via <nettime>: no commercial use without permission
#  <nettime> is a moderated mailing list for net criticism,
#  collaborative text filtering and cultural politics of the nets
#  more info: majordomo {AT} bbs.thing.net and "info nettime-l" in the msg body
#  archive: http://www.nettime.org contact: nettime {AT} bbs.thing.net