Dr Michlmayr

Having passed all requirements for the Doctor of Philosophy, I had my graduation today.

LinuxTag, Release Interviews

At LinuxTag, I conducted interviews on release management with developers from various projects, including Gentoo and OpenOffice.org. One question regarding time-based releases relates to the most appropriate release frequency. Should you do a release every 6 months, every 12 months or what? So far, Gentoo has done four releases a year. Unlike I initially thought, these weren't mere snapshots but real (i.e. tested) releases. They are currently moving away from four to two releases a year because they say that preparing a release takes significant effort and doesn't leave enough time for development when you try to make a release every three months. On the other hand, OpenOffice.org follows a (roughly) 18 month cycle. This is partly because Sun wants to sell a boxed version based on OpenOffice.org and customers wouldn't buy a new version if the cycle frequency was much higher. It also allows other commercial products to appear, such as books; they would be constantly out of date if new releases came out every few months. Some community members would like to see more frequent releases, however, and it will be interesting to see what model will be employed in the future.

Bugzilla best practices guide, Debian quality guide

Luis Villa, the quality maestro of GNOME, recently asked for help writing or reviewing a Bugzilla best practices guide. He thinks that putting the insights gained from GNOME into a guide will help other projects. I completely agree with him and I suggest that we put together a Debian quality guide as well. Not only will such a guide make it much easier for new people to get involved with Debian's QA efforts, but other projects will also benefit. They can which problems Debian is facing, how we are dealing with them, and maybe they have some good ideas for us too. Given that there are so many different processes employed by free software and open source projects, it would be good to have a collection of best practices. I'm always amazed when I look at other projects and see that they have really nice processes (Mozilla being one prime example).

Papers: final: 3, accepted: 1, submitted: 1

Last week has been fairly successful with regards to publications. I finished my paper describing quality problems and practices in free software projects as well as the long abstract detailing my PhD work and submitted camera-ready versions to the First International Conference on Open Source Systems. Gregorio Robles, Jesus Gonzalez-Barahona and I also completed the paper about the evolution of volunteer participation in Debian and submitted it to the same conference. Furthermore, I submitted a paper looking at process maturity and the success of free software projects to another conference, and I received notification that a paper looking at bugs in the Debian base system has been accepted as a chapter in a book. Papers and drafts are available on request. On Monday and Tuesday, I attended a workshop on research methodology organized by our institute.

22:22

I remember being fascinated as a child when I would stay up long enough to see those magic numbers on the clock: 22:22. Well, it's 2:22 (i.e. 02:22) now. A few hours ago I finished a paper I have been working on, and a few minutes ago another paper done in collaboration with other researchers was completed and submitted to a conference. I wish it would just be 22:22 now.

Research interests

Over the last few months, I've subscribed to several e-mail alerts by various publishing companies in order to follow new publications in areas of research I'm interested in. Yesterday, I subscribed to Blackwell's e-mail alerts and they allowed a very fine grained selection of topics. This gave me a good overview of what areas I'm actually interested in. I had to choose between book and journal alerts. I figured I'd subscribe to journals even if I only had a vague interest in the area, but for books I was more selective.

The following list shows my book selections, and can be seen as my core areas of interest:

The full list of journal areas I chose is pretty crazy and I'll probably understand less than half of those areas (or have time to actually read the papers). However, I intend to learn more about these areas in the next months so I might just as well start following those publications now.

Interviews on release management

Last night I returned from the Mataró conference where I spent the last four days as a researcher interviewing 20 free software developers from a variety of projects about release management in their projects. The aim of this study is to get a better picture of what processes, practices and tools are currently being employed by free software projects during release management and to identify problems. I'm very grateful to the people who agreed to participate in the interviews. I know that everyone was very busy with their work, and I'm glad they managed to free up some time for the interviews.

I was quite surprised how exhausting it is to perform interviews. You'd think it's no work since the other people are doing most of the talking, but I was really very tired at the end of the day. Maybe this is because I tried to summarize the interviews as I listened, which is quite a significant challenge. Anyway, I also recorded the interviews and over the next few weeks I'll listen to the interviews again and go through my summaries to make some sense out of this data. I definitely got some good input and ideas during the interviews.

Biology and Anthropology lectures

I've been interested in biology and anthropology for quite a while (in particular evolution) and for some reason this interest has grown much stronger over the last weeks. Sometimes I wonder whether I should have studied biology instead of psychology, but on the other hand I'm pretty happy I studied psychology because it has influenced the perspective with which I view the world in an interesting way. At the same time, there are many areas in biology I'm interested in and of which I only have a very basic understanding.

While I have a fairly good grasp of evolution, I really lack basic fundamentals of biology. I have therefore decided to take biology and anthropology lectures. Unfortunately, I'm not allowed to attend practicals but as a member of the university I can go to any lectures I want to. Matthew (who studied genetics) came over yesterday (with robster) and gave me some advice and I later printed the course information and read through it. I very much felt like a freshman again, going through the lecture list and creating a timetable. I've now decided to attend Biology of Cells, Physiology of Organisms, Animal Biology and possibly Evolution and Behaviour (although the latter might be too introductory).

Shopping, King's college

Yesterday, I went to the main site of the Engineering Department to get an access card and another computer account. I waited for 40 minutes and when I asked whether they forgot about me it turned out the person I was supposed wasn't actually there. Doh! So I went to my work place, connected my laptop with the IP address I stole from a Windows client there and found out that I had been assigned an IP address for my room. My two supervisors also scheduled a meeting with me for the next day, at 9 in the morning (ouch). After catching up on e-mail and having dinner, Daniel Silverstone kindly came by with his car and took me shopping to the big Tesco. I bought various things, like Pint glasses and some food. I was surprised how little money I spent for all the things, and the whole thing didn't take too long even if Daniel wants to make people believe otherwise (actually, let me mention that he claimed he was patient, but then was crying like a baby after shopping with me for 5 minutes!). Anyway, during the day (after lunch), I stopped for a minute to take some pictures and thought that I can be really happy to live in such a beautiful city.

Today, got up at about 8.30 and left at about 8.57 for my meeting. Talked about my research interests for an hour, and then my supervisors had to go to another meeting and scheduled another one with me for the next day (guess when? Right, 9am). Went to the main Engineering site, and got an access card and signed up for an account.

Signing up for uni

Got my student card and signed up for an account at the central Computing Service (got a very unaesthetic username but I won't use it much anyway). Back at King's, also signed up for an IP address for my room, and I asked the network guy who long it will take. He said he ran out of IP addresses, but he'll look into it in the afternoon (alas, this obviously didn't happen). Food at college seemed good and cheap. After lunch, I went to the bank to change my address and pay in some money, and then went to the Centre for Technology Management (where I'm doing my PhD) to talk to my supervisor and sort things out. As it turns out, they don't even provide me with a dedicated PC. It's not really a problem since I have my laptop, but I think this is really lame. Every PhD student should get their own PC. Anyway, I have to go to the main Engineering site tomorrow to get an access site and an IP address.

One more paper to go…

Yeah, finished another paper; only one more to go (next weekend). I might go the beach tomorrow or on Monday. It has been over 30 degrees the past few days.

Valedictory Dinner

Today was Valedictory Dinner at my college. Since I'm finishing my degree this semester, I counted as a valedict. Of course I've only stayed here for a year whereas most other valedicts are finishing three or four year degrees. Some guys had prepared a slideshow of photos of the last few years and a movie which was rather amusing. Someone also played a nice prank — he suddenly removed some flags from the wall of the dining room (we have flags of all the nationalities from our college there) and behind he had written farewell.

Now I have to prepare for a presentation on my research which I'll have tomorrow morning. The slides are fortunately done, but I have to fix the handout one of my colleagues has written and add some references.

Debian & uni work

It's almost mid October. I wonder where the time has gone. The semester is drawing to an end and I realize that I've spent much more time on Debian than on uni. I think I basically treat Debian as a full-time job and uni as a part-time job. I suppose that explains why I'm busy all the time. ;-)

There were some interesting events connected to Debian recently. A few days ago, I got up and found a nice message in my DPL mail folder informing me that Debian has won the Linux Journal Readers' Choice Award as the favourite distribution. I also managed to get James to announce how New Maintainer rejections would happen and to actually get some done. It was really time this happened. I also chased more inactive maintainers. It's amazing (and depressing) how many inactive developers there are in Debian. However, I'm actually more annoyed by those developers who are basically inactive but don't think they are. Unfortunately, they are not up to date with what's going on at all and don't produce any good work at all. At the same time, they don't want to give their packages away because "I'll find time to fix it this weekend" (sure…). The problem is that I have heard this from so many people and then seen the (non-existing) results that I know it will never happen - but they really think they'll find the time. I've also spent a fair amount of time helping some folks in Italy prepare a proposal to get funding from the EU for a project involving mobile Linux and Debian.

I've also done some work for uni. I've finally managed to compile a listing of 40 successful and 40 unsuccessful random projects from SourceForge. This wasn't as easy as it may appear because I had to make sure that the two groups don't differ in lines out code, age and category (development status). Now my other other team members can analyze those projects. We're doing a blind study so I won't tell them which projects are successful and which are not.

Today it's actually sunny and pretty warm. Unfortunately, this is the exception rather than the norm. The weather has generally not been very nice in the last few weeks, and I've been told by a local that it is likely to stay like that for the next couple of weeks. The photo show the view from my window.

Open Source research in Cambridge

I received an email a few days ago from an Open Source researcher from Cambridge University (Judge Institute of Management) who wanted to interview me about Debian. I found this pretty ironic. In any case, I just spent an hour on the phone with him talking about management and coordination aspects of Debian.

Uni, team work

I had a meeting at uni today with the folks with whom I'm working on some research (basically, it's forced teamwork in a class on Software Engineering case studies). It seems to be going fairly smoothly; apparently we're ahead of most other teams. Later, I wrote an outline of the PhD I'm going to do because I need this for a scholarship I'm going to apply to; I also got all of the other bits and pieces together which I need for the scholarship application. After finishing that, I spent a couple of hours on sorting out Debian related mail. Unfortunately, no matter what I do, it seems to be getting more and more. Even though I currently have "holidays", I'm at least as busy as normally, if not even more so.

SourceForge and data collection

I started writing some scripts to extract various data from SourceForge which I'm going to use for some academic research. Unfortunately, I had to discover that they don't provide mbox files of their mailing list archives so I wrote a script to snarf the postings from the web and create an mbox (of course some headers are missing, but they are fortunately not crucial). The FAQ says they don't provide mbox files on the web because of e-mail address harvesters, but they could at least provide them to registered users. I'm also a registered user now. I'm not totally comfortable with this given SourceForge's increasingly proprietary nature, but SourceForge currently has more data to offer to a researcher than services like Savannah. I expect this is going to change eventually, though.

Cambridge

Hmm, apparently I've been offered membership by King's College.

Prospective supervisor

Met with my prospective supervisor in the afternoon and had a nice chat. In the evening, went to the cinema and watched Bruce Almighty with cjwatson.

Uff

Okay, completed my exam. I'm happy it's over. While the sample exam contained some knowledge questions, the real one put the emphasis on understanding, which is good… but now I feel as if I've wasted so much time cramming in all those facts. Anyway, it's over, which is good.

Now I have a busy weekend preparing all kinds of stuff because I'm leaving for my big journey on Tuesday. I have to do the laundry, clean up my room (put everything in the closet so they won't charge me for the room during the holidays), and do some more stuff for uni. At the moment, I'm trying to catch up on e-mail.

Exam, Friday the 13th

Just realized I have my exam on Friday the 13th. Fortunately, I don't believe in this stuff…