Money for the Crowd
Tuesday April 29th 2008, 23:06h
Filed under: Development Cooperation, ETH Zürich, Research

Today Evelyn, one of my ETH students, presented her master thesis on the role and potential of microfinance at Credit Suisse. While I knew the principles of microcredits beforehand, I wasn’t aware of the great scope of the highly growing industry. Just some number: It’s estimated that already today USD 25 billion are invested in such loans to the poors!

In the evening I was invited to a book vernissage in Zürich. The finance expert Naoko Felder-Kuzu presented her book “Kleiner Einsatz, grosse Wirkung” which was substantially supported by Credit Suisse. (thus I’ll keep my CS account despite their last weeks losses ;)

On my train ride home I read half of the book and may already say the purchase was a successful investment! While providing interesting footnotes and important references, the author accomplishes to fascinate the reader with her enthusiasm for microfinance. Among others Naoko writes a short history of microfinance and compares it with European lending models of the 19th century. She illustrates the fundamental differerence between dependency-creating donations vs. sustainable microfinance (while pointing out not all problems can be solved with microfinance). And she also explains disputed issues such as why the borrowers have to pay high interest rates or how finance institutes are finding the right balance between social assignement and profit maximization etc. I can warmly recommend this new book which provides a brief 100 page introduction into “doing good while doing well!”



GSoC 2008 started - great example of firm involvement in FLOSS
Tuesday April 22nd 2008, 23:30h
Filed under: ETH Zürich, FLOSS, Research, Ubuntu

While we’re trying to get open source projects on the Hackontest platform, I just read Leslie’s post in the Google Open Source Blog. I believe it’s really amazing: 1125 student proposals for 175 different open source projects got accepted for Google Summer of Code 2008 and will be financially supported by Google with USD 5000 each. That’s really a substantial amount of money! Of course it’s PR, of course it benefits their recruiting strategy, but it’s not just cheap CSR, it’s really substantial - and relevant and beneficial for the OSS community as a whole. It’s definitively an interesting and successful example of firm involvement in the FLOSS world - the focus of my PhD.

Great to see GNOME got 30 students (I’m desperately hoping John Ross Stowers will succeed with his Evo syncing proposal!), Scribus 3, Inkscape 5, OpenMoko 10 and OpenStreetMap 4 - congratulations Sebi! A pitty TYPO3 and Ubuntu, two of my favorite projects, aren’t on the list…



Trying the Slashdot experience
Friday April 18th 2008, 9:45h
Filed under: FLOSS

Please help us bringing the Hackontest announcement on Slashdot by voting for it on

Vote for the Hackontest announcement

Update 2008-04-19: We’ve done it (and our server didn’t crash), thank you for your support!

Hackontest on Slashdot

Update 2008-04-21: As expected the feedback on the announcement on Slashdot is amazing. In 36h hours almost 1GB of traffic, 110′000 hits and 4000 unique visitors on hackontest.org - resulting in many new registered open source projects and users! However, the comments on Slashdot are quite harsh thus I had to provide my point of view as well.

Update 2008-04-22: Reactions on our announcement on Slashdot are huge. Although our server didn’t break, there are some funny reactions resulting in a conspiracy theory - everything because of a wrong link… So don’t be afraid, there won’t be any dead hackers despite 24h coding and Hackontest doesn’t take place within the SARCOPHAGUS, it’s just a regular etoy.TANK ;)



Software Reuse: A Solution for one of Gartner’s «IT Grand Challenges»
Tuesday April 15th 2008, 15:00h
Filed under: ETH Zürich, FLOSS, Research

It seems like our academic reseach on software reuse (Spaeth et al., 2007; Haefliger et al., 2008) bears a lot of practical implications with future potential, at least for the next 20 years of IT management! Gartner found out:

While the exploration and development of tools to enhance productivity continues to capture attention, it would appear that effectively and efficiently exploiting reusable code is one of the most encouraging rays of hope to yield more output per programmer. But many challenges exist there as well. Minimizing the time required to find the perfect software module and avoiding the need to modify reusable software are among the many challenges.

Gartner further concludes that the monolithic architecture of Windows seems to be a dead end. Thus, Microsoft guys, stop hacking your own stuff but start collaborating within the highly modular Linux operating system!



Back to research
Thursday April 10th 2008, 1:12h
Filed under: ETH Zürich, Linux

Yesterday’s research seminar with Prof. Gareth Morgan inspired my a lot and revived my research vein. Although Gareth’s message was basically “There is no such thing as objectivity” and “Do not confuse your data with reality”, his comments about science, subject and object, reality and its representations motivated me in my own research work.

Gareth Morgan in his seminar at University of St. Gallen

Thus this morning I started reading the short but fascinating article about the 2.6.xx Linux kernel development, published by The Linux Foundation. It contains a lot of hard facts about the past 3 years of kernel development, e.g. who and which companies contributed how much code. Then I continued writing on the theory part of our new paper on innovation and control trying to come up with literature for our hypothesis. Then in the afternoon I had a great 1.5h phone interview with Mickey Lauer of OpenMoko talking about the challenges making a completely open mobile phone. The conclusion is basically: Although you might need to use inferior hardware (in the case of OpenMoko it’s the GPU of SMedia) since the major manufacturers won’t open their drivers, in the end the community will be more helpful and loyal contributing thousands of brains for your innovation project if you stick to your openism paradigm. E.g. in the case of optimizing the power management, voluntary community members even found ways on how to improve the use of the highly complex power management unit. This annecdote shows how powerful the support of the crowd can be if you open source your software, communicate transparently and accept external contributions.

My notes in Gareth's seminar

Also method-wise I’ve learned a lot. However, while I’m now a PhD student for over 2 years, it’s time to diffuse the experience on how to write scientific work. Thus I convinced Peter, Zeynep and Jan to offer a 2-day course for MTEC students about the basics of scientific work. For those of you who can’t wait till Fall, this is our preliminary syllabus:

What is science? How do we know what we know?
Introduction into philosophy of science
Strengths and weaknesses of different research approaches
Differences between solving practitioner’s problems and contributing to theory

Why theory is important
How the scientific publishing process works (about journals, editors, reviews etc.)
Where to find previous research about my area of interest
How to read a scientific paper
How to reference literature

Formulating the research question
Characteristics of good research questions
Deriving a question from theory: Finding contradicting theories, unexplainable phenomena, combining different research streams (Y-approach)
The question’s relevance to theory

Introduction to methodology
Inductive (out of data) vs. deductive (hypotheses testing) research
Qualitative and quantitative methods – Designing Interviews and Questionnaires
Overview of different methods (single/multiple case study, survey, grounded theory building…)

Data gathering and analysis
Sources of data
Quality of data
Instruments for data analysis (SPSS, MaxQDA etc.)

Power of conclusions
Summing up the findings and interpreting results
Generalizability of results
Implications for research and management
Limitations of own research, future research questions

So how do we start?
Writing good proposals
Planning the thesis
How the thesis will be graded



Hackontest takes off - slowly
Friday April 04th 2008, 22:58h
Filed under: FLOSS

This week we finally released the Hackontest platform which is used to select the most attractive open source projects and teams till August 1st, 2008. As in many of my software projects before the effort developing the website was much larger than initially expected… Luckily there was my old friend Emanuel Indermühle who helped bearing the burden - thanks to you for the many sleepless nights!! And designer Stefan Sicher was kind enough to invest a huge amount of spare time to make the nice templates of the site, thus thanks to you as well!

Now let’s see if the experiment takes off. Thanks to Google’s Leslie Hawthorn we were able to get a great list of ten prominent jury members for the event:

Jeremy Allison, Samba
Jono Bacon, Ubuntu
Brian W. Fitzpatrick, Subversion
Martin F. Krafft, Debian
Alexander Limi, Plone
Federico Mena-Quintero, GNOME
Bram Moolenaar, vim
Bruce Perens, OSI founder
Lukas K. Smith, PHP
Harald Welte, gpl-violations.org

So thanks to you as well giving your name to a yet non-existant project! Now let’s look forward and hope that more projects such as phpMyAdmin join Hackontest and file their planned features on the platform. Thus I hopefully may pass on ‘my’ projects soon to real maintainers…



Promising times for Ubuntu
Friday April 04th 2008, 11:43h
Filed under: Education, FLOSS, Ubuntu

This week the Geneva government announced 9000 Ubuntu installations on its school computers. That’s definitively a case we have to bring at the next teacher’s conference on open source software! But also in the Swiss German part of Switzerland we’re not inactive in terms of Ubuntu adoption: By the beginning of May our association /ch/open will release the new Ubuntu Swiss Remix Hardy Heron on DVD thanks to the sponsoring of leanux.ch. And while Dell Switzerland is still not offering pre-installed Ubuntu (probably because of the lack of local Canonical support) I ordered a new XPS M1530 this week hoping its hardware is well-supported by Hardy Heron.