Thursday Mar 24, 2011

Now blogging over at bärwolff.de/mblog

To all those who have subscribed to this blog -- I have now moved over to bärwolff.de/mblog. The first goodie is already up there, so don't be shy.

Thursday Oct 21, 2010

Just a quick note on how I resolved two vexing issues I encountered when trying to put a Linux on a Thinkpad A22m. The driver for the sound card needs some special kung fu as detailed here. Which in turn makes the sqeeze testing system a good choice (rather than going for the lenny backport option). Also, note that texlive is still on 2007 in lenny, even in the lenny backports.

The real big fish though was not the sound card, but the xorg.conf for the display. You'll have to add a certain magical modeline to the xorg.conf. In my case I also had to boot the system such that a root maintenance console interrupted the boot process, because once gdm was about to load, the machine went kind of dead. So what I did was enter the grub menu, typ "e", add vga=normal to the boot line, then enter the root maintenance console, insert a usb-stick with the proper xorg.conf, mkdir temp, mount /dev/sdb1 temp, cp temp/xorg.conf /etc/X11/, and CTRL-D -- done.

The xorg.conf Monitor and Screen sections read:

Section "Monitor"
Identifier "Configured Monitor"
#HorizSync 30.00 - 62.00
HorizSync 30-110
VertRefresh 60-110
# one of the following does the trick
Modeline "1024x768_60.00" 64.11 1024 1080 1184 1344 768 769 772 795 -HSync +Vsync
# 1400x1050 @ 60.00 Hz (GTF) hsync: 65.22 kHz; pclk: 122.61 MHz
Modeline "1400x1050" 113.89 1400 1432 1552 1752 1050 1053 1056 1084 -hsync -vsync
Modeline "1400x1050_60.00" 122.61 1400 1488 1640 1880 1050 1051 1054 1087 -HSync +Vsync
# 1400x1050 @ 75.00 Hz (GTF) hsync: 82.20 kHz; pclk: 155.85 MHz
Modeline "1400x1050_75.00" 155.85 1400 1496 1648 1896 1050 1051 1054 1096 -HSync +Vsync
Option "PreferredMode" "1400x1050"
EndSection

Section "Screen"
Identifier "Default Screen"
Monitor "Configured Monitor"
SubSection "Display"
Depth 32
Modes " 1400x1050" "1024x768_60.00" "1400x1050_60.00" "1400x1050_75.00"
EndSubSection
EndSection

Monday Sep 27, 2010

Von Berlin nach Hiddensee

Man nehme: den IC 2351 Richtung Stralsund, buche das Ticket ab Bernau, bis dahin mit ABC-Ticket; Rückweg ditto. In Stralsund mit dem Bahnticket den Tram-Bus 6 zum Hafen. Von dort mit der Fähre rüber nach Hiddensee. Auf der Insel selbst bewegt man sich zu Fuß, mit Fähre oder Pferdekutsche. Fuhrmannshöfe Neubauer (Kloster) und Mach (Neuendorf) sind nett, die anderen sicher auch, kenn ich aber nicht persönlich.

Der Fisch unten am Hafen in Vitte ist top (wenn man vom Edeka aus auf den Hafen zukommt, links, am Fahrkartenschalter vorbei, und dann direkt rechts neben der Eisdiele, der kleine Imbiss, die haben jeden Tag fangfrischen Fisch). Sparkasse gibt's in Vitte, Cashgroup nirgends auf der Insel. Die meisten Restaurants sind mäßig, in Grieben gibt's ein sehr gutes (Zum Enddorn), wo allerdings die Inneneinrichtung mal wieder etwas Frische vertragen könnte.

Friday Jul 02, 2010

Aus der Abteilung "poetischer Spam"

Am besten find ich ja: "... dies haben könnte Sie Kosten, es ist Teil unserer Politik zu machen, um sicherzustellen, dass alles nach allen Regeln der Sicherheit im Rahmen dieses Finanzinstitut mit keinerlei Kosten. Wir hoffen auf eine bessere Zusammenarbeit von YOU."

Das gesamte Zitat:

"Dear Customer Value (s)

Dies ist eine Meldung aus der Abteilung Sicherheit von POST-BANK.DE; es ist
unsere Politik, Ihnen mitteilen zu können dass es Internet-Betrüger derzeit
auftretenden unserem Finanzinstitut.

Seit 31. März 2010 ein neues System / Datenbank installiert wurde, um die
fallenden betrügerische Aktivitäten in unserem Bankensektor zu beseitigen,
das neue System heißt "Für Ihre Sicherheit".

Es ist auch unsere Verpflichtung, dass Sie einen einzigen monatlichen
Newsletter mit Tipps, die Sie verwenden, um Opfer des Internet-Betrug zu
vermeiden empfangen kann. Mit diesem System halten wir Sie auf dem
Laufenden / Benachrichtigung über die Aktivitäten des Internet-Betrüger und
auch zum Schutz vor Skimming.

Wir bitten Sie, folgen Sie den untenstehenden Link, um Ihr Konto, indem
einige der erforderlichen Daten in Ihrer finanziellen online / Internet-
Konto, unter unserer Institution, um maximale Konto / Fonds / IP-Schutz "Für
Ihre Sicherheit" gewährleisten zu aktualisieren.

Verwenden Sie den unten stehenden Link

Wir entschuldigen uns für die Unannehmlichkeiten, dies haben könnte Sie
Kosten, es ist Teil unserer Politik zu machen, um sicherzustellen, dass alles
nach allen Regeln der Sicherheit im Rahmen dieses Finanzinstitut mit
keinerlei Kosten.

Wir hoffen auf eine bessere Zusammenarbeit von YOU.

Mit freundlichen Grüßen,

Frau Anna Vetter
Security Department
POST-BANK DE
Berlin, Deutschland."

Friday Dec 18, 2009

Mixing two PDFs -- reversing page order, interleaving pages from two files, all using bash and pdftk

Had the following problem: scanned loose two-sided pages with a big scanner who did manage two-sided scanning but got stuck too often, garbling my pages in the process of turning the pages around; so I turned to one-page scanning which obviously left me with two PDFs per scan -- one with the odd-numbered pages, and one with the even-numbered ones in reverse order.

Here's what I did after playing around a little bit. No doubt this can be made into a more robust, better documented, and more versatile script, but a one-liner did for me:

NUMBER=3984; DUMMY=pdftk Report\ \echo $NUMBER\\ $$even$$.PDF dump_data output | grep -i NumberOfPages; PDFPAGECOUNT=${DUMMY:15:3}; pdftk Report\ echo $NUMBER\ $$even$$.PDF cat end-1 output Report\ echo$NUMBER\ $$evenrev$$.PDF; pdftk A=Report\ echo $NUMBER\ $$odd$$.PDF B=Report\ echo$NUMBER\ $$evenrev$$.PDF cat for a in \seq $PDFPAGECOUNT\; do echo "A$a B$a"; done; output Report\ echo$NUMBER.pdf

This will work for sets of PDFs named "Report 3984 (even).PDF" and "Report 3984 (odd).PDF". For reports with other numbers simply change the NUMBER bash variable.

Basically, what the script does is (1) finding out the number of pages in the two pdf files, assuming them to be the same; (2) reversing the page order in the even-pages PDF so as to get the proper page order; and (3) merging the odd-pages PDF and the reverse-order even-pages PDF using pdftk and a little bash loop that feeds pdftk the proper sequence of pages alternating between the two input PDFs.

Coming to think of it, the second step could also be dropped by elaborating the latter mentioned bash loop so as to give, say, "A1 B7 A2 B6 A3 B5 ..." instead of "A1 A2 B1 B2 A3 B3 ...". But I'll leave this exercise to you.

Amend the script at your leisure.

Saturday Sep 12, 2009

Kommentar zu ZEIT-Artikel "Reguliert das Netz!"

Gestern habe ich einen etwas längeren Kommentar zu einem Artikel in der letzten ZEIT (2009/38) verfasst und dort auch in sechs Teilen online gestellt. Hier noch mal in ganzer Länge:

[Im Artikel] wird gesprochen von "gesellschaftlichen Konsequenzen" die nicht genügend bedacht werden [aha!], von "Kräften" "auf die gegenwärtig weder Bürger noch Staat direkten Einfluss nehmen" [scary!], von der Einschränkung "bürgerlicher Freiheiten" [welche nochmal? dass Arbeitgeber die Internetaktivitäten ihrer Mitarbeiter kontrollieren?], garniert mit der Pointe, dass "nur staatliche Regulierung solche Probleme lösen kann" [natürlich, der Staat, der macht das schon, genau wie damals die öffentlichen X.25-Netzwerk-Monopole, war ja auch eine gute Idee]. Das ist natürlich starker Tobak, und dem Autor kann ich keinen Vorwurf machen, ein komplexes Thema mit journalistischen Mitteln an eine breitere Öffentlichkeit zu vermitteln.

Wirklich durchleuchtet hat der Artikel das Thema -- Netzwerkneutralität -- hingegen nicht, und auch nicht die vielen Piraten in den Kommentaren. Dabei ist schon der Titel so provokant wie es nur eben geht: "Reguliert des Netz!". Der Staat soll sich also der Probleme des Netzes annehmen und diese kraft seiner legislativen Mittel und Wege lösen. Wie soll das konkret aussehen? Das bleibt, abgesehen von einem Verweis auf "Netzneutralität" und Amerika (... du hast es besser unklar im Text, und im übrigen auch im Wahlprogramm der Piraten. Dort steht nur: "Das Netz muss sich neutral gegenüber den transportierten Inhalten verhalten", S. 23) Die Piraten geben dort auch noch anderen Unsinn zum besten, etwa auf Seite 19: "Die Ausschaltung von Zwischenhändlern ermöglicht es, dass den Künstlern vom Erlös ihrer Werke ein größerer Teil verbleibt und direkter zufließt. Außerdem wird damit das Spektrum der Kulturszene deutlich erweitert." Super, verbietet alle Zwischenhändler im Cyberspache, ach was, warum nicht gleich Handel und Kommerz überhaupt! Aber ich schweife ab.

Der Punkt ist: Netzwerkneutralität klingt zwar gut, ist aber akademischer Unsinn. Weder war das Internet je "neutral" (es gibt kein Ding das wertneutral sein kein, auch wenn manche das für das Internet behaupten; glaubt mir, es stimmt nicht), noch wäre ein solcher Zustand sonderlich erstrebenswert, denn Menschen reagieren nun mal auf wirtschaftliche Anreize, ob es nun die Piraten sind, die gerne ihre Musik kostenlos hören, oder die Netzbetreiber die gerne ihre Mitarbeiter und Aktionäre bezahlen möchten -- so ist die Welt nun mal. Und wäre es wirklich so klug, das Internet so wie es heute ist für immer in Stein meißeln oder in staatliche Monopol-Obhut übergeben? Darüber müssen wir doch eigentlich nicht diskutieren.

Den Staat zum Hüter von "Netzneutralität" machen zu wollen ist also bestenfalls abenteuerlich, schlimmstenfalls katastrophal. Warum sollte der Staat den Internetnutzern (also auch den Botnetzbetreiber, ja?) alle Macht, und sich selbst und den Netzbetreibern keine geben? De Sola Pool hat das schon vor 25 Jahren diskutiert und geschlossen: "Regulation is a last recourse. In a free society, the burden of proof is for the least possible regulation of communication. If possible, treat a communications situation as free for all rather than as subject to property claims and a market. If resource constraints make this impossible, treat the situation as a free market rather than as a common carrier." Ich wüsste nicht, was an dieser Einschätzung heute falsch sein sollte.

Und noch etwas: Es ist bezeichnend, dass die IETF -- das heimliche Parlament derjenigen die das Internet betreiben und technisch weiterentwickeln -- explizit keine Stellung zu moralischen Fragen wie Telefonüberwachung oder eben auch Netzwerkneutralität einnimmt. Ted Hardie hat es kürzlich beim 75. Treffen der IETF in Stockholm wie folgt ausgedrückt: "We made a choice a while back about what our values were. And it wasn't for network neutrality, it was for liberty. We chose to try and design a network that enabled the maximum number of end-to-end flows that enabled people to connect, for communities to form, and to allow all of us not to be mere consumers but actual providers of content onto the network." Die Rechtsphilosophen haben ihre Ansichten, die Techniker die das Internet bauen die ihren. Vielleicht sollte man mehr auf letztere hören denn auf erstere. Dann würde vielleicht auch endlich die wesentlich wichtigere "Tussles in Cyberspace"-Debatte zu ihrem angemessenen Platz in der öffentlichen Wahrnehmung kommen, und nicht die ewig alte Leier von Netzwerkneutralität versus den bösen Netzbetreibern. Es wäre an der Zeit. Wir schreiben das Jahr 2009.

Und, weil es so schön ist, noch eine letzte Sache als Nachtrag. Nach Hayek ist menschengemachtes und kodifizieres Recht eine feine Sache, aber vage und informelle Regeln oft besser, denn meistens gehen bei der Kodifizierung und Durchsetzung die entscheidenden Aspekte einer Regel verloren. Von daher sind die Policy-Prinzipien der FCC aus dem Jahre 2005 vielleicht das beste was dem Internet passieren konnte. Denn ob die FCC das Internet überhaupt regulieren darf, ist ja unklar.

Saturday Aug 01, 2009

Application Layer End-to-End Arguments: From Ends to Means, and Beyond Network Neutrality

This Thursday there was a plenary session at this year's IETF meeting in Stockholm, Sweden on what net neutrality means for the IETF and the IAB. There was a generic summary of what network neutrality is about, but also a very interesting talk by Mark Handley and a couple of interesting responses from the audience. The audio of the whole session is up here.

Anyway, I was thus reminded of an unpublished paper of mine that has been sitting on my hard drive more or less idly for the last eight months. A version that I submitted to a conference at HU Berlin was rejected, because they had no clue what I was on about, and also lamented that it's a bit longish. A version with a couple of corrections from this May titled "Application Layer End-to-End Arguments: From Ends to Means, and Beyond Network Neutrality" is now up at bepress.com. Still not properly published, but available online, at least. And apparently on about very relevant points.

Friday Jun 26, 2009

Cleaning a Lenovo T60 Fan

The other day my T60 fan stopped working. I took the machine apart, following the very useful instructions from the Hardware Maintenance Manual. Turned out a little adhesive label had gotten stuck in the fan. Removed it, bought some new heat thingy paste for the CPU, and put it all back together -- and, voila, it works again.

Just for the record, here's some pictures of the inside and the processor, plus the corpus delicti:

and

Sunday Jun 21, 2009

Vertical and Horizontal End-to-End Arguments

A poster of mine at this week's IEEE ICC in Dresden (workshop on the future of the internet): here

Thursday Apr 16, 2009

Mary Anne Hobbs off the iplayer

BBC radio shows are always on for a week to listen online. There are a couple of nice ones such as Mary Anne Hobbs. Since I am not getting BBC's browser based iplayer to work on my box, I have now resorted to the following bash one-liner: mplayer curl http://www.bbc.co.uk/iplayer/aod/playlists/g3/wj/j0/0b/RadioBridge_intl_0100_bbc_radio_one.ram

Simple as that. The beauty of this approach is that it works every week anew, without having to fiddle with new URLs all the time. It also works from outside the UK.

Do accordingly for the other shows (Westwood and what have you), the URLs actually follow some logic, too; but are best grabbed from the BBC Radio website.

Sunday Mar 01, 2009

There is more to slums than meets the eye

A truly interesting article in today's Boston Globe, on the virtues of slums' informal regulation dynamics that in many respects outperform the dull urban life in "developed" cities -- they are pedestrian friendly, are much more community oriented, and sustainable: Learning from Slums. I for one, had never thought about this, much less seen it this way. But, 1 billion people live in such settings, that's a huge number, and hence the timeliness and relevance.

Wednesday Feb 04, 2009

Ubuntu on my T60

I moved from Gentoo over to Ubuntu 8.10 recently. It mostly works fine out of the box. The only thing that is flaky is suspend which I haven't quite figured out how to get working solidly.

Some tweaks though: I largely followed:
this to get some of the non-free stuff such as Google Earth, and this one in order to install a recent Eclipse version (Ganymede, 3.4).

Plus, I installed tuxcards and last not least xminds, a decent mind mapping tool that finally makes freemind redundant for me. Xmind also comes as a plugin to Eclipse which makes productive writing and note taking much more integrated. Nice one.

Thursday Oct 09, 2008

Ist vom Betrug ermudet und der Luge ...

Das ist ja schon fast Poesie, die Spammer werden romantisch!

Quote:

Ist vom Betrug ermudet und der Luge suche ich den ehrlichen und richtigen Mann

Hallo werde ich der schone Fremde, uber unsere Bekanntschaft mich froh sein rufen Marina.
Ich suche seine einzige und eigenartige Halfte. Ich suche nicht einfach den Mann ich ich suche den Freund, des vorliegenden Freundes auf ganzes Leben. Ich das warme, zartliche, zarte Madchen. Mir 27 Jahre.
Es ist ein wenig uber mich: ich liebe, die Gemutlichkeit zu schaffen, mir gefalle wenn zu Hause gut und ruhig, ich in das Ma? das ernste, verantwortliche, richtige Madchen. Au?er dem ich romantisch, empfindlich, verstehend und emotional. Ich verstehe mit dem Humor, auf die Probleme bezogen zu werden und lustig zu sein. Fur mich in der Einsamkeit verliert das Leben den Sinn ich ich verwelke wie das Blumchen. Betreffs der Einfachheit kann, dass ich offen wie das Buch sagen. Mich mu? man nur lesen. Wenn es Ihnen interessant ist, uber mich jenes dann zu erfahren schreiben Sie!
Ich warte mit der riesigen Ungeduld auf die Antwort.

Thursday Aug 14, 2008

Free software and economic value

It's been all over the blogs, the Judge White opinion in Jacobsen vs. Katzer, overturning the lower court ruling that hass followed the defendant's argument that there was no copyright violation, just mere contract violation with virtually no damages, since the plaintiff suffered no economic harm. This is good and sane news. A quote I liked very much on this subject at page 8 of the opinion:

"Traditionally, copyright owners sold their copyrighted material in exchange for money. The lack of money changing hands in open source licensing should not be presumed to mean that there is no economic consideration, however. There are substantial benefits, including economic benefits, to the creation and distribution of copyrighted works under public licenses that range far beyond traditional license royalties."

Tuesday Jul 15, 2008

OpenOffice on speed

I have always hated the slow start algorithm of openoffice. By coincidence, though, I have finally stumbled across a blog entry describing how to speed it up by changing some of the option avaiable in openoffice.

Monday Jul 14, 2008

I've rewritten the heise4rss script, as it stopped working once more due to changes in their markup. Now it uses their rudimentary rss feed, goes through it it fetches the content of any item not yet fetched previously. Works fine, and should be more stable for now.

Tuesday Jul 08, 2008

Bildschirmhintergrund

Wednesday Jul 02, 2008

Chicago Style auf Deutsch

Für alle, die dass Chicago Manual of Style mögen, hier die bibtex style files für entsprechende Anwendung auf deutsche Texte. Also und statt and, Hrsg. statt Ed., keine Kommata vor den unds, etc.

Big up Frank Pallas and me for bringing Chicago style to Germany, then

Friday Jun 06, 2008

Ich habe nochmal ein bisschen rumgebastelt am PHP-Filter für liferea, der aus heise news einen gültigen RSS-Feed zaubert. Es nennt sich immer noch heise4rss. Viel Spaß damit.

Wednesday Jun 04, 2008

Speed up emerge by compiling in RAM

Of the many useful tips at gentoo-wiki.com this one has become my favorite: Speeding up portage with tmpfs. It really speeds up compiling big time and leaves your hard disk alone by moving the compile action to RAM.

Thursday May 29, 2008

A quick and not so dirty CV in LaTeX

(updated, now featuring a little png with the signature neatly put at the bottom, and a way more elegant way of pushing things down to the bottom)

I wanted to typeset a CV and realised I had no templates whatsoever on my machine. The old versions had simply gone. So I had to do it all anew, reluctantly at first, but then I found some neat things that made it a rather feasible exercise.

Basically two things helped me most. First, I managed to get some nice control over the description lists by using the enumitem package. Second, I devised a way to get the date at the bottom of the page by simply using \vfill.

And, while I was at it, I looked for fonts and found a comprehensive catalogue of all freely available LaTeX fonts. I chose a modern nice looking one. Also, note that I redefined the font to be used by the url command. You don't want a typewriter font in print anymore.

\documentclass[a4paper,12pt]{article}

\usepackage[utf8]{inputenc} % input encoding
\defaulthyphenchar=127 % Trennungszeichen für T1-Fonts
\lccode\==\- % Definition des Hyphenation=-Bindestrichs
\usepackage[T1]{fontenc} % output encoding
\usepackage{graphicx} % wir wollen Grafiken
\usepackage{textcomp} % Sonderzeichen z.B. Euro
\usepackage{url} % URLs
\def\UrlFont{\rm} % the fefault is \tt but I don't like the look of it in print, \it neither
\usepackage{setspace}
\usepackage{enumitem} % we want nicer description lists
\usepackage{lmodern} % going for the neat stuff!
\usepackage{microtype} % this is said to be cutting edge, see http://en.wikipedia.org/wiki/Microtypography

\hypersetup{
pdfauthor = {Matthias Bärwolff},
pdftitle = {Matthias Bärwolff (2008) CV},
pdfcreator = {LaTeX},
pdfproducer = {pdflatex}}

\pdfadjustspacing=1 %%% force LaTeX-like character spacing
\onehalfspacing
\addtolength{\textheight}{4mm} % this is to allow for the signature png to go a little over the bottom limit

\begin{document}

\pagestyle{empty}

\section*{Curriculum Vitae\\of Matthias Bärwolff}
\vspace{6mm}

\subsection*{Research Experience}

\begin{description}[font=\itshape, labelindent=3mm,leftmargin=15mm, topsep=0mm]
\item[Open Source --] Extensive and ongoing work for the Open Source Jahrbuch series (Open Source Annual, \url{www.opensourcejahrbuch.de}); plus various articles and book chapters (\url{ig.cs.tu-berlin.de/ma/mb/ap})
\item[Internet Design Principles --] Extensive research over the last two years with primary focus on end-to-end, network neutrality, and the history of the internet
\end{description}

\subsection*{Education}

\begin{description}[font=\itshape, labelindent=3mm,leftmargin=15mm, topsep=0mm]
\item[2000 -- 2002] MA International Business Finance at Bournemouth University, UK (with distinction); master's thesis Linux and Windows -- A Case of Market Failure?'' published by Palladian Law Publishing (\url{www.cippm.org.uk/publications/index.html})
\item[1997 -- 2000] Work training as Mathematical-technical Computer Scientist'' at Humboldt-Universität zu Berlin %; web based e-learning platform as final project
\end{description}

\subsection*{Work Experience}

\begin{description}[font=\itshape, labelindent=3mm,leftmargin=15mm, topsep=0mm]
\item[since 2003] Research and teaching assistant at Technische Universität Berlin, Chair of Computers and Society (\url{ig.cs.tu-berlin.de})
\item[2000] Head of front-end department at meOme AG
\end{description}

\vspace{6mm}
\noindent
\textbf{References available upon request}

% pushing the date down to the bottom of the page, signature a little lower for optical reasons
\vfill
\noindent
\today \hfill \raisebox{-4mm}{\includegraphics[width=0.45\textwidth]{mbaer-signature-2008}}
%May 29, 2008

\end{document}

Thursday May 22, 2008

End-to-End vs. Net Neutrality

Commentators on both sides of the net neutrality debate tend to equate end-to-end with net neutrality (see e.g. here).

In sum, the equation is false and misleading no matter which end you apply it to. The two principles are different in content, and, more importantly, different in their normative interpretation.

Net neutrality is about neutrality in a very abstract and often vague sense (due to the problem that neutrality always implies some specific purpose, for there is no absolute neutrality) beyond common carriage. Now, some have applied the same level of abstraction to end-to-end, saying it is about having no functions in the network and all functions at the end points. Having simplified end-to-end this way, plus claiming the internet has (or had) always been end-to-end (after all the principle was "invented", so popular parlance goes, by the very people who have created the internet) leads to the neat conclusion that adhering to the end-to-end principle will also achieve the purposes of the new net neutrality principle.

But, in actual fact, end-to-end has never been about having all functions at the ends and non in the network. The normative purpose of the principle was a technical one, not one of control and power. And, it was most definitely not about "innovation", "democracy", and "free speech". It was about error correction. Saltzer, Reed, and Clark opened the floodgates of interpretative scope in 1981; but as for the Arpanet and TCP/IP which were developed mostly in the 1970s, end-to-end considerations may at most be construed to have informed (and have been derived from) the splitting of TCP into IP and TCP in order to make IP "thinner" and more suitable for applications that needed no error correction. Yet, and inevitably, there are a host of functions relevant to the end points within the network. The question, of course, is which functions. but most don't get to that question in the first place.

The proper distinction between end-to-end and net neutrality is not merely of academic concern, it is crucial to understanding and appreciating the elusiveness of bending the end-to-end principle as a computer science practice of the 1970s to today's concerns about "innovation" and sovereignty of end users. Again, the often quoted link between end-to-end and net neutrality does not logically flow from the contents and purposes of the two. They are not the same.

We should thus distinguish very carefully between discrimination on technical grounds and discrimination on economical and political grounds.

Thursday May 15, 2008

Microsoft's Backdoors

I haven't been using Windows for ages, and despite the latest Debian GAU (größter anzunehmender Unfall) after having read that little piece on Microsoft's privacy infringements in the name of security I think I better not come back.

Saturday May 10, 2008

JMRI vs Katzer

I've been following this, and just wanted to link to the latest hearing in the appeals court essentially discussing whether copyright could protect open source software when there is no conventional economic interest from which damages may flow. There's an mp3 of the hearing.

Thursday May 08, 2008

A New Way to Look at Networking - Transcript of bits

I took some notes while watching Van Jacobson at Google Tech Talk talking about the inappropriateness of our TCP/IP conversation paradigm. If the bulk of traffic is ultimately about data, not conversation, then we might be way better off with a data centric paradigm that structures our design principles.

What follows is raw unedited notes and mostly verbatim quotes fro the talk with some time stamps in between:

Copernican revolution in networking
(7.20)
generation 1 (around 1900): the phone system (focus on wires);
generation 2: the internet (focus on endpoints);
generation 3: dissemination (focus on the data)

Thus, it may well be that the end-to-end TCP/IP model becomes a lower layer that will become less important in the face of a more timely content-centric/data centric networking overlay that works on any networking layer as long as it moves bits. Data may become the key paradigm, and if it is authenticated, and there is sufficient competition in moving bits, then we fears of ISPs messing with our connections/data will be obsolete/futile.!!!

Examples for such "ad-hoc dissemination overlays" (presentation slide, minute 1:14:12) include "Akamai, BitTorrent, Sonos mesh, Apple Rendezvous".

!!! It is not about replacing TCP/IP, it is about doing something new with it (minute 1:15:27)

!!!!BitTorrent goes in the right direction, but as it lacks information about locality, topology, etc. it has to do these things via statistical means, it doesn't have a sufficient control of that level of the infrastructure.!!! (1:18:00)

!! proble with IP multicast: "you can't scale it up, you can't make it global." (1:18:59)
But this is actually thinking too far down, because you want things local anyway. And if they're not there than they shall get it for you. We want to architect the system so that we "only need local multicast."

multicast is very efficient. E.g. "resloving DNS names by a multicast is very efficient". (1:20:11)

• problem with phone system: reliability decreases exponentially with the number of elements in a path. Probability of failure is P to the N. (P hoch N). Thus the individual pieces had to be extremely reliable in order to deal with the structural unreliability of the scaled system.
• plus: huge setup costs (overhead of establishing the connection vs. actual data transmission)
• (aside: ISDN only 64KB per second, we don't want that, we want more)
• generation 2: packet switching, the focus is no more on the path, but the data packets carry their destionation address with them and the routing decision is made instantaneously by the network. -> breaking with the telephone patadigm. Of course, they also used the phone system, it was there, it was giving them paths to be used, albeit in a different way then the telephone system used to.
• the routing really has wonderful properties, because even when you do it statistically and from thereon try to improve ir, you don't need to engineer anything as a start, you can just put things together and it'll work! (think Kleinrock random routing)
• (Small error in the presentation at around minute 22, Licklider was in 1964 o more head of the IPTO)
• NICE QUOTE at min 22:40: "The ideas were so simple they couldn't fail." (referring to Baran)
• In the very beginning the packet network arpanet was a "pure overlay" of the telephone system. (23:45)
• !!! Back in the telephone system the topology was all that mattered, with packet switched networks nothing could matter less. The ends were just interested in the address of the other end, they didn't care about topology. The topology was something the packet saw when it was going trough the network. (around 26:00)
• Vint Cerf -> CATENET, interconnecting all the different networks that had emerged everywhere in the US by 1973: Vint said: "We need a common encapsulation format and a common address instruction. It only has to be common where two networks touch."
• Specs for TCP/IP in 1974, first multinetworks public demonstration 1977: Catenet (from concatenate), the military liked TCP/IP better so they chose that for a name
• "[With TCP/IP] you never set up anything, you just launch a packet. If the network cannot deliver it, it throws it away. Okay, I don't care, I'll launch another one." (28:00)
• 29x "We're not scheduling anything." !!! "The end notes will fix it up."
• 30:15 "[With TCP/IP and packet switching] the addresses are explicit in the data, but not implicit in the particular pattern of wires."
• from ca. 33:00 problems with TCP/IP today. The internet was build when there were few machines, few people working on those machines, little data. Today things are different and the net was not built for that.
• "Connecting to the internet requires a globally known, unique IP address that's topologically stable on routing time scales (minutes to hours)". Thus "The net doesn't like things that move, plus "The transport protocols hate broadcast." - "The net hates things that broadcast, the protocols weren't designed for that." They were designed for point-to-point communication.
• ca. 37:00 protocols based on the idea of conversation are highly inefficient when we actually have dissemination, or broadcast!
• "The success of the protocols change the world, and when you get a new world, the protocols may not fit the new situation" (!!! 37:30)
• 38:48 "Conversations are the central architectural element of TCP/IP, as opposed to the phone system where they are what happens by accident after you build a path."!!!
• 39:00 Today, however, "people don't want to have conversations. What they want is their webpage." "The high 90% level of traffic is people trying to get some named chunk of data. They hand in the URL, and they want something back. That is not a conversation." !!!
• We have "conversational network" and use them for dissemination. Obviously, the two don't quite fit in three fundamental ways:
• First, "we can't do a security model at the network level, because it's blind to the data."
• Second, broadcast is hugely inefficient, again, because the network doesn't know about the data.
• Quote, 41:46 "Up at the user level, we've got all of these apps that are dealing in dissemination style data that want to deal in URLs, mail messages, sensor readings, individual voice packets. And our network model is to write this app, I have to figure out what server to open a TCP conversation to. It has nothing whatsoever to do with the problem that you're trying to solve. And so both the app writers and the users have to do this horrible plumbing operation in their head and in the network. If you've ever set up a VPN to get at the internal websites or get your email when you're offsite, you're doing this nasty ugly plumbing to solve a problem that the network should be able to solve for you. But it doesn't know your intent, and it can't phrase what you want, in terms of conversations.
• 42:41 What we really want is "dissemination networking" (the resulting paradigm may be called "Content-centric networking", this term is also on wikipedia)
• 42:50 "Data is named" (not end points!!!): "When you want some chunk of data, you present its name to the network, and it doesn't matter what kind of network technology you've got. You use them all."
• 49:18: "The top-level design philosophy is "data has got a name, not a location"."!!!
• This is not just caching, "the effect is very much the same, but yout point of view is completely different."
• "Wherever I get it, I can authenticate it. I know whether or not I can trust it. ... I don't care where it is."
• "Integrity and trust are derived from the data, not the channel it arrives on."!!!
• 51:10 "I want properties of the data where I don't have to trust remote agents, that the data itself lets me figure out what it means."
• 53:37 "If I can say the intent at the top level, then I can do things lower down to concept the right semantics."
• 54:37: I would like to be able to say to my service provider: send my stuff first [rather than my son's gaming traffic]. But there's no protocol for that. And the people that want to do QoS are the service providers. They want to make it end-to-end because they can extort more money out of Google if they say, oh no, pay us for end-to-end quality of service."
• next 55:11 "You get an architectural focus in sort of the wrong kind of QoS. It won't make things better."
• 57:20: With content centric networking "there is much less nees for global topologies, global naming."
• As for security!!! 58:55 "By its design, a conversational network wants to be completely agnostic to its data. It wants to be transparent. ... When the protocol [TCP/IP] was designed, it was important to make it completely agnostic. But if it is completely agnostic, then you cannot do a trust model. Anything can be in that pipe, and you don't control the other end."!!!
• On stuff to be solved: Certificate hierachies are bad, because you need to trust the one root, which you may not want to do. But "if you use PGP models, you can get some pretty good trust in things." (PGP web-of-trust model)
• Plus, naming is gonna be a major issue, sequence and time information, version number etc. will have to be part of the name, and tie that into the verification and authentication machinery. (1:05:30)
• Recap, as for the advance of packet switching by adding address to the data so that the network always knows what to do with it: (1:08:23) "Making implicit information explicit suddenly enables a whole bunch of new functions." Later: "You do the analogous thing in dissemination, where you take information that is implicit, like time information, or version information [and make it explicit]."
• The harder stuff (around 1:11:00): "Incentive structure (flow and congestion control, sharing and redistribution incentives); miscreant and freeloader detection and interaction with anonymity; redistribution (content routing, storage replacement strategies, liability issues); corruptet, invalid or expired content (reputation, gossip, epidemics, ...)"

Sunday Apr 20, 2008

Wild style: Skydiving with wing suits

Check these pictures.

Friday Apr 18, 2008

On the origin of the Arpanet

Check out Computer Networks - The Heralds of Resource Sharing at Google Video, a documentary from 1972 when the Arpanet was essentially up and running and the basics of the lower layers had evolved to a point where the network could be used for some applications such as file transfer and remote login.

It's got some very nice footage of the fathers of the Arpanet, precursor of today's internet, including Licklider (from MIT; he had also been at IPTO in the early 60s), Lawrence Roberts (who made it all happen), Robert Kahn and Frank Heart (of BBN, who build the first IMPs and protocols to link those and the hosts attached). Donald Davies (out of the UK) who had been inspired by Paul Baran and is understood to have come up with the term "packet switching" is also interviewed briefly. Davies actually saw quite clearly the everyday impact of a packet switched network, even though he never got anywhere with his plans, the UK government were simply not interested in the thing.

Now for a quote of Licklider about books becoming obsolete:

"If everyone had a display console in his home and in his office, he could be reading from electronically stored information instead of from a book. And the difference is, he could have access to anything he wanted to read instead of just what was within reach.

It turns out to be surprisingly inexpensive if you get wideband transmission facilities to send the stuff right when it is asked [not sure here with the transcript] to be read instead of sending it to a local bookstore or a local library in the hope that it might be read." (at 24 minutes into the film)

I also like the bit at minute 22 where they show a teller machine and there's "Cool-O-Mat" written on top.

Thursday Apr 10, 2008

Scary Stuff

this.

Wednesday Apr 09, 2008

Texlipse is just awesome

While I'm at it, I'd just like to repeat, this time in English: TEXLIPSE IS AWESOME! Spread the word!

What is Blogging, anyway?

It occurred to me yesterday that blogging is "the institutionalisation of change and interaction as a paradigm on the web". Now, this will hardly be news, but it made the whole blog notion seem more approachable to me. Mind you, we just had a conference in Berlin devoted entirely to "blogging".

First, there were static websites. Websites had to be written in an editor and uploaded to a web server. Clearly, change was possible here, too; but the paradigm was that of a "page" as in page in a book.

CMS systems made change easier. But still, the paradigm of the web was pages - now more easily changed via the web itself - on webserver, to be read by human visitors.

A blog, on the other hand, gives technical shape to a new paradigm, that very "change and interaction". It is not a page, but rather a pile of pages which grows and may be annotated by visitors. Also, being a pile rather than a page, it lends itself to being conceived as a pile, or a queue, or whatever - in any case something that may then fit to other established paradigms such as email or news, which are often displayed as a list. And so may blogs when viewed as RSS feeds, not websites anymore.

Thursday Mar 13, 2008

Texlipse

Bin gerade über TeXlipse gestolpert und schwer begeistert. Noch nie so eine IDE für Latex unter Linux gesehen, die Parser sind wirklich gut. Werde meine Arbeit jetzt damit weiterschreiben.

Monday Mar 10, 2008

Article on Open Source

I've written an article on incomplete contracts and informal credits as the "social glue in Open source". It's up here in a proof version. Due out this spring.

Wednesday Mar 05, 2008

Zwei neue Papers von mir

Hier finden sich zwei neue Artikel von mir. Der eine über Monopoleigenschaften von Freier Software, der andere eine kurze Notiz zur aktuellen Network-Neutrality-Debatte (in englisch).

Friday Feb 29, 2008

Are we autonomous anymore?

Norbert Elias has argues some time ago that we are no more, for we are social beings and as such depend on others. And civilisation is precisely about making people conform -- use fork and knife, don't spit, don't burp, you name it.

Enter technology and things get worse (not better) -- so worse, in fact, that people have gone mad or outright criminal. It is for a reason that by now 1 percent of all Americans is imprisoned, that is kept fo the majority of people who manage to remain normal and cope with things. That is the world's highest rate. Still, reading "Industrial Society and Its Future" by Theodore Kaczynski, one of those who didn't like what he saw and now serves a life long prison sentence for murder and a host of other offenses committed in pursuit of "making people understand how the system works", is highly inspiring. Here is the book. I've just read it from back to front, couldn't stop.

Just three quotes to get a taste:

"Today people live more by virtue of what the system does for them or to them than by virtue of what they do for themselves. And what they do for themselves is done more and more along channels laid down by the system. Opportunities tend to be those that the system provides, the opportunities must be exploited in accord with the rules and regulations, and techniques prescribed by experts must be followed if there is to be a chance of success."

"Modern society is in certain respects extremely permissive. In matters that are irrelevant to the functioning of the system we can generally do what we please. We can believe in any religion we like (as long as it does not encourage behavior that is dangerous to the system). We can go to bed with anyone we like (as long as we practice "safe sex"). We can do anything we like as long as it is unimportant. But in all important matters the system tends increasingly to regulate our behavior."

"Many modern people [...] are disturbed by the prospect of death, as is shown by the amount of effort they expend trying to maintain their physical condition, appearance and health. We argue that this is due to unfulfillment resulting from the fact that they have never put their physical powers to any use, have never gone through the power process using their bodies in a serious way. It is not the primitive man, who has used his body daily for practical purposes, who fears the deterioration of age, but the modern man, who has never had a practical use for his body beyond walking from his car to his house. It is the man whose need for the power process has been satisfied during his life who is best prepared to accept the end of that life."

And, last, as for technology:

"[H]uman race might easily permit itself to drift into a position of such dependence on the machines that it would have no practical choice but to accept all of the machines' decisions. As society and the problems that face it become more and more complex and machines become more and more intelligent, people will let machines make more of their decisions for them, simply because machine-made decisions will bring better result than man-made ones. Eventually a stage may be reached at which the decisions necessary to keep the system running will be so complex that human beings will be incapable of making them intelligently. At that stage the machines will be in effective control. People won't be able to just turn the machines off, because they will be so dependent on them that turning them off would amount to suicide."

Wednesday Feb 27, 2008

ccTLD "us" stands for United Status

So if your status is united, too, consider registering a .us domain (found at the .us entry at IANA database of top level domains)

Update: They've fixed it now, wayback for the previous mistake.

Thursday Feb 07, 2008

Most of us do have a name they share with others, and so do I. Mix that with google and interesting things happen. Basically, you're in for a fight for the top spot and as many hits as possible on the first serarch results page. Now, mix that with people search engines like pipl.com and even more interesting things happen. While on google my "opponent" only manages to fetch two entries on the first results page, on pipl.com (which also shows up on the very first google results page) we simply get merged into one person with a mix of keywords that does little justice to either of us. Well, at least apart from his being active in a far left wing party he seems to be a reasonably decent chap.

Friday Feb 01, 2008

IPod Touch 1.1.3 Issues

Just three brief hints about obstacles I encountered upon upgrading the thing. Loads of different websites helped.

First, make sure you don't run out of space on the first partition, which is particularly nasty when upgrading to 1.1.3 and the process gets stuck halfway through. Do the following with Term-vt100:

cd /
cp -pr Applications /var/root
mv Applications Applications.old
ln -s private/var/root/Applications /Applications

Then reboot (pressing the power button on the top until the slider for powering appears, slide and its powered down properly). You may want to remove the backup on the first partition. Done. New apps will all be installed via symlink on the second partition.

Second,when you're done with upgrading to 1.1.3 the Term-App will now ask you for a password which you don't have. Install SUID Lib Fix from trejan.com (adding that to the sources of the installer first). Then the terminal will be sane again, password is alpine. As of now, you can't change that though I haven't tried, hmmm. Not funny, actually, you may not want to run an ssh server when in public.

Third, Google Maps (of applerepo.com) stopped working. Do the following:

cd /System/Library/Frameworks/GMM.Framework/
mv gmm GMM

That's it.

Monday Nov 19, 2007

Linux is evil

Linux is evil and only used by European anarchists, Marxists and terrorists, of course. This blog entry finally reveals the truth:

"And guess what software Osama Bin Laden uses on his laptop? If you guessed it was Linux you would be 100% right. Osama uses Linux because he knows [it's] designed to counterfit DVDs, curcumventing the Digital Millenium Copyright Act, and defraud companies like Disney. Next time somebody asks you how Al Queda agents pay for their rifles and rocket launchers, you can tell them that foreign hackers make software called Linux which helps them steal from Americans."

Thanks, Tristan for helping us sinners see the truth!

Wednesday Oct 31, 2007

FOAF Profile

The social web is upon us, thus I have ventured to finally add a FOAF profile to my website. Here it is. The foaf browsers I have just tested cannot cope with umlauts, so an umlaut free variant is this.

Thursday Oct 11, 2007

I just downloaded the new Radiohead album. What to say, I think it's way better than the last one. Not remotely as dark, much more balanced and less "over the top experimental" by any measures.

Friday Sep 28, 2007

"Es hat sich ein hohes halluzinatorisches Potential entwickelt" - Wohl wahr.

Eine wohltuende Abwechslung zu den überwiegend spekulativen Streitschriften (à la Benkler, Frischmann, Lessig), in denen Open Source als Vorform neuer, am liebsten doppelter Freiheiten, geadelt wird - aus einer Vorabveröffentlichung des Textes "Der Unwert des Unwissens - Verkürzte 'Wertkritik' als Legitimationsideologie
eines digitalen Neo-Kleinbürgertums" von Robert Kurz
(erscheint in Exit! Nr. 5)):

Wenn man "unter sich" auf das Copyright verzichtet, und zwar um dennoch "sekundäre Geldeinkommen" zu erzielen, dann ist das vielfach mehr der Inhaltsnot geschuldet als der emanzipatorischen Tugend. [...]

Es handelt [beim von Meretz propagierten "Remixen"] sich um die Fortsetzung der bekannten Verfahrensweise "copy-and-paste". Wie sich Leute, die selber nicht einmal einen Aufsatz auf Mittelstufen-Niveau zustande bringen, ganze Diplom- und Doktorarbeiten erschleichen, indem sie aus dem Internet "heruntergeladene" Textbausteine zusammenkleben und dies notdürftig durch Umformulierungen tarnen, so wird nun im Kontext der digitalen Alternativ-Ideologie das schlichte "Abschreiben" als emanzipatorischer Akt propagiert. So stellen sich denkfaule, aber hochanspruchsvolle Pennäler die "Befreiung" vor. Der Gipfel der Dreistigkeit ist es, wenn solches Vorgehen in Verbindung gebracht wird mit dem Kampf gegen die Ausbeutung etwa von Pflanzenwissen der bäuerlichen und "indigenen" Bevölkerung in der Peripherie durch westliche Pharma-Konzerne, die sich dieses geklaute Wissen dann patentieren lassen. In Wirklichkeit übernehmen hier die "Remix"-Ideologen selber den ausbeuterischen Standpunkt, indem sie sich die Anstrengung des Begriffs, der Recherche und der berühmten "Kreativität" weitgehend sparen, um die individuellen Produkte anderer einzusaugen, sich mit fremden Federn zu schmücken und das auch noch als besondere Art der "Kreativität" auszugeben. Der schlichte Ideenklau, das Plagiat, das Nicht-Ausweisen von Referenzen und Zitaten und überhaupt die systematische intellektuelle Unredlichkeit werden zum emanzipatorischen "Kampf" gegen das "geistige Eigentum" geadelt. Ist doch alles freies "Menschheitswissen" hier, ungefähr wie das Ohmsche Gesetz und der Satz des Pythagoras. (S. 43-44)

Thursday Sep 20, 2007

Getting rid of the mounted volume icons on the gnome desktop (nautilus)

This had bothered me for sometime. The solution was found at http://yoten.blogspot.com/ and reads: Start gconf-editor, in terminal or with ALT+F2 run application window. Find /apps/nautilus/desktop and check out volumes_visible!

Wednesday Sep 05, 2007

Listening to BBC Radio with Firefox on Gentoo

Having Totem installed plus the nsplugin use flag set will make firefox want to play real audio streams such as those delivered by bbc.co.uk/radio with totem, not realplayer. The following entry with /etc/portage/package.use will make firefox go sane again:

media-video/realplayer nsplugin
media-video/totem -nsplugin

Friday Aug 17, 2007

The Ultimate Victory of Linux

I wonder whether Linux has now finally become mainstream and its users are becoming as computer illiterate as the ordinary victim of silly ads purporting to be some sort of system message telling you to click in order to update or improve your PC. The ad I've just stumbled upon is this:

The translation of reads roughly: "We have ascertained that yours is a Linux box. Your PC is now ready for a performance improvement. Configuration will only take two minutes. To start press Okay." Really, the funniest web ad I've seen in a while.

Thursday Aug 09, 2007

Sometimes the Past Seems just Plain Cool

Did you know that in 1924 people seriously attempted to intercept radio signals from Mars with their shiny new radios, for "[c]ommunication with these beings [from Mars] promised great scientific revelations"? I didn't, here's the source.

Bug or Feature?

Now I'm getting all Google conspiracy theory paranoid No seriously, I just did a search for a random text bit from a piece of Eli Noam, unlike Google not exactly a fan of net neutrality. (The text fragment was "By making communications ubiquitous and essential") It struck me as a little surprise that Google returned but a blog comment from cisco.com instead of the obvious original. Pressed a little harder by clicking the rarely ever noticed line reading "In order to show you the most relevant results, we have omitted some entries ..." I got to the article, eventually. Granted, the paper is 13 years old, but why omit it?

Mind you, live.com, Microsoft's rival search sensibly returned the article first, and then the blog entry. But no surprises from Google anymore since my yesterday Makruk encounter.

Wednesday Aug 08, 2007

Finally, The Joy of Personal Search

I'd never have stumbled upon Makruk had I not searched for Abhandlung. Thanks, Google!

Tuesday Aug 07, 2007

Ever searched for a search engine on Google?

A search for "search engine" (without the quotation marks) yields a host of search engines other than, emmmh ... Google. Yahoo, Microsoft, ask.com, all make it above google, the first of which to appear is www.google.co.uk on page two of the results. I mean, they must have put it on, or why should Google's ranking mechanism fail at their very own site? It would be interesting to know how many such terms have been rearranged in their search just in order to look "not evil".

Thursday Jul 12, 2007

Alternative Compensation Systems - In Fiduciary Duty we Trust the Government?

I am not too sure about whether alternative compensation systems such as those proposed by Lessig will work. The proxy sample to compute a true account of music usage would have to be way to large given the heterogeneity of people, distribution channels, listening habits, and devices. The complexity of a due system would almost inevitably invite the dark side to game it, others to lobby for exceptions making the system even more complicated and possibly unjust (what is justice, anyway? it's in the eye of the beholder). Tomorrow I'll be at SERCI congress in Berlin and meet some people that have been working on the issue. I am curious whether something has changed over the last two years.

Wednesday Jul 11, 2007

Keep 'em News Coming

If only every blog would put out news as rarely as Marcus Comstedt, creator of amiwm. His newsflash reads: "Newsflash: amiwm0.20pl48 now out! (1998-03-15)". Feels like yesterday, doesn't it.

Friday Jul 06, 2007

When Inefficiency Strikes at Regulating the Ends

An interesting news at heise: An internet provider has been ordered by a Belgian court to filter for copyrighted content. Now that makes for a powerful exception from any net neutrality attempts, in effect rendering completely futile any net neutrality efforts. On second inspection, this violation of the hypothetical neutrality of the access layer is part of a broader class of reasonable exceptions from any normative call for neutrality. As long as there are untrustworthy ends there will be innovations that reduce value from other ends or conflict with positive law. Be it DDOS attacks, spam, or, well, copyright violations. End if it takes unreasonable costs to keep the originating ends from inflicting such harm it sounds entirely reasonable to focus on the network to prevent such incidents.

Kile and Gnome

Now that I am getting serious on writing my thesis, I have finally spend a minute on customising kile to look good with Gnome. First I had to emerge kcontrol, 50 mb and one hour of compiling just for setting the look of the only KDE app on my system - but hey, I just don't want to put up any more with a wack looking kile. Having emerged kcontrol I further emerged klearlook, a theme that resembles Gnome's clearlook theme. Finally just choose the klearlook theme under "Appearence and Themes", submenu "Style", and you're done. Now I am curious if I could just remove kcontrol with the changes remaining on my system. But I am gonna try that another day

Thursday Jul 05, 2007

What is the (economic) value of the internet?

After having read a number of pro and contra articles about net neutrality I was always kept wondering what, in fact, were the application level innovations the proponents of net neutrality repeatedly claim to be in grave danger from ever imminent "potential" discrimination efforts of incumbent network operators. Email? Way too old. WWW? Hardly an innovation on the "egde", rather one that garnered significant institutional support before taking off. P2P? File transfer, more sophisticated, granted, but an innovation? Maybe, but certainly not one that is unequivocally positive value only, by any measure, not only that of the RIAA.

The fact is that there is simply no empirical measure of internet value, let alone that stemming from "application level innovation". Metaphorical statements abound but cannot solve this problem satisfactorily. Consider this one by Brett Frischman, for example:

"End-users [...] engage in innovation and creation; they speak about anything and everything; they maintain family connections and friendships; they debate, comment, and engage in political and nonpolitical discourse; they meet new people; they search, research, learn, and educate; and they build and sustain communities. [...] These are the types of productive activities that generate substantial social value, value that too easily evades observation or consideration within conventional economic transactions. When engaged in these activities, end-users are not passively consuming content delivered to them, nor are they producing content solely for controlled distribution on a pay-to-consume basis. Instead, end-users interact with each other to build, develop, produce, and distribute public and nonmarket goods."

Not even a trace of remotely scientific measure of value here. Maybe we should have a little more regard for the intermediaries that make it all happen. Consider this statement by Stephen Smoliar, found here:

"VALUE resides not in the ability to create content nor even in the facility at transmitting that content but in INTERPERSONAL MATCHMAKING that enables bringing the content and its transmitting channel to where it is needed."

Now, matchmaking has traditionally been the domain of intermediaries, due to the simple reason that people are social beings who specialise their work efforts and commonly benefit from the optimised total. If you look at it this way, there are intermediaries all over the place. Making the underlying markets in the first is an intermediary activity, too, just as more specific ones like search, or blog hosting, or whatever. Maybe we should look more closely at the ways intermediaries create value, how they generate revenues, and how much surplus they leave with customers. Maybe this way will we get more reliable estimates of how valuable the internet really is. Plus, wherever there is a positive effect, there may just be a corresponding negative one nearby. Let's not forget those, either.

Tuesday Jul 03, 2007

I just had to post this

If that ain't the funniest patent ever then god knows which is. No wonder all these kids' accessoires are so expensive ...

Thursday Jun 21, 2007

Shaping the Matrix as Long as We Can

I have just read an inspiring article at Illuminata Perspectives about the pitiful current state of virtual worlds. People construe of, expect of, and behave in them the same as in the real world. Buildings look the same, people look the same, and do very much the same things than they do in reality. Still, one day those virtual realities may well become more important, entirely new ways of behaving and communicating may emerge. It's been that way with virtually any new media before. Television at first stuck to the old way of doing things the "radio" way, there were speakers who basically just sat in front of the camera saying things. That soon changed as we know today, "the medium is the message" as Marshall McLuhan famously put it, and changing the medium changes just about everything. Enter climate change and growth of world population, and the idea of settling with virtual worlds becomes less far fetched. The question that springs to mind, then, is who gets to control these worlds, and what will be the rules, if any. The law is becoming more humble thanks to Lawrence Lessig and the insight that code, market, and norms regulate as well, if not better. We should start thinking and educating about this issue as long as we still can.