This reminded me of a paper I had written for a course I followed during my master’s programme at Leiden University, Politics of Bureacracy. For my argument I summarily investigated political appointees in the USA. American federal executive officials are nominated by the president and confirmed by the senate.

Political appointees in the USA

When Leon Panetta (Democrat) was appointed as the Director of the Central Intelligence Agency under Obama (Democrat), the appointment raised questions if Obama was trying to politicize the CIA; Panetta had no experience in intelligence work. Allegations of partisanship are not new, as even stronger criticism was aimed at the appointment of Porter Goss (Republican), the Director of Central Intelligence under George Bush (Republican).

Lewis (2009) argues (use this link for a PDF behind a paywall but with decent layout) that these political appointments are made for two reasons. They allow the president to control the bureaucracy, with appointees being more responsive to the wishes of the political leaders than public servants. They are also made for patronage, i.e. to reward the members of the president’s party with lucrative positions. He goes on to explain that too many appointees ultimately hurt the performance and control of federal agencies.

Appointees lack the experience of career public servants and stay for short tenures, impeding long term planning. Lower performance in turn makes the agencies harder to control for the president, for example because they are more likely to make mistakes. On the other hand, as outsiders appointees can also give new energy to agencies. It’s a matter of balancing the ratio of appointees to public servants to get optimal agency performance. The USA evidently has far too many appointees, more than 3.500, as opposed to a number between 100 and 200 for other developed countries.

The situation in the netherlands

While I don’t have exact numbers, I assume the Netherlands is in the latter category too. Our General Intelligence and Security Service (Algemene Inlichtingen en Veiligheidsdienst in Dutch) is led by a career public servant. The CPB would even have a serious credibility problem if it would be led by a political appointee, it is imperative that it is an apolitical organization.

I think the difference between the USA and the Netherlands might partially be explained by our different political culture. We have coalition governments instead of the Democrat-Republican duopoly. Even the Balkenende-III cabinet in which the Labor Party didn’t participate had no problem with appointing a vocal economist with Labor Party sympathies like Coen Teulings as head of the CPB.

What I see as a notable exception to the non-political nature of appointments in the Netherlands is the appointment of Piet Hein Donner (Christian Democrats) as vice-president of the Council of State (Raad van State). The cabinet insisted that the procedure was open, but many opposition figures thought it was a farce and Donner’s appointment was predetermined.

The man surely is an expert in law and seems capable for the job, but the appointment had a semblance of politicization to me as well. Just like the CPB advises on economic policy, the Council of State has an important advisory role on law towards the government. All the more reason to make it just as apolitical as the CPB.

Obama and partisanship

What caught my attention regarding the politics of the USA is the appointment of the Republican Secretaries of Defense Robert Gates and now Chuck Hagel by Obama. This is different from this discussion of the relationship between the public service and politics because a Secretary of State is always a politician, but related to the partisanship issue. Why not reward a Democrat with position? Obama assures us he chose Hagel because all he cares about is having the right person for the job. Some journalists think it really is a clever plan to screw the Republican Party.

What if Google deiced that you can only send e-mails to other users of GMail and not to users of other e-mail providers? There would be outrage over Google’s anti-competitive move and GMail users would switch to other e-mail providers. Yet this is exactly what LinkedIn is doing: it’s impossible to communicate with users of other professional social networks, such as XING and Viadeo.The key difference between GMail and LinkedIn is that LinkedIn has a very dominant position in the Netherlands, almost 3,8 million members as of November 2012. That’s probably why we tolerate this “lock-in” from LinkedIn.

I don’t like monopolies, dominant market positions or anti-competitive practices. If I would create a LinkedIn account myself I would reward LinkedIn for its behavior and strengthen its dominance on the Dutch market. I’m desperately looking for job, but not desperate enough yet to part from my principles. LinkedIn and its competitors should meet and devise an open standard so I can communicate with users of competing social networks. until then, I want none of it.

But how does that compare to other laptops, in particular my Acer TravelMate TimeLine 8371 which still serves me nicely? In the end of 2009 I paid € 532 for the Acer, minus € 70 for the Windows Vista Business license refund and plus the € 187 for the Intel X25-M 80 GB SSD it set me back € 649 total. Obviously there is a large difference in price, but what’s the justification for the much higher cost of of the Dell and Samsung?

Notebookcheck reviewed both the Acer, the Dell and the Samsung, albeit in slightly different configurations than I’ve seen. The key differences for me are that the Dell and Samsung have much higher resolutions: 1920 by 1080 pixels and 1600 by 900 pixels respectively. The Acer has 1366 by 768 pixels, and all three have a 13.3 inch monitor. The lower resolution of the Acer is by no means annoying to me, but I’d appreciate more. Another key difference is the weight: 1.73, 1.4 and 1.19 kilo for the Acer, Dell and Samsung respectively. If you take a laptop with you on holiday in a backpack filled with clothes like me, the difference in weight is notable. Both the Dell and the Samsung also have more durable enclosures than the Acer.

The Dell and Samsung have much faster CPU’s, more RAM, larger SSD’s. The only downside is that the Dell has no ethernet port at all and the Samsung requires an adapter, because both laptops are too thin for a full port. The Dell has DisplayPort and the Samsung has a VGA connector requiring an adapter which is not included. The Acer does have a normal VGA connector and only the Samsung and the Acer have a card reader.

I think I like the Samsung most. Even though the limited connectivity is a downside – Samsung should include a DisplayPort connector too and include adapters for VGA and HDMI – I don’t think it’s a big problem. However, the Samsung shoves Windows 8 down my throat, and I’m not so eager to go through all the trouble to get a refund for the Windows license. Choices, choices! For now I think I’ll keep using my Acer until it’s spent, by that time there will be better hardware in the form of Intel’s “Shark Bay” platform.

I also wouldn’t mind less powerful CPU’s so the price could be lowered. All I want is a laptop which is portable, with a monitor size between 11 and 13.3 inch, a keyboard (no tablets for me), a long battery life and a low price. That might happen in the form of Intel’s “Bay Trail” platform for its Atom architecture. Of course, some competition from AMD would be much appreciated.

It frustrates me that in practice they can’t check the scientific sources of statements made on Wikipedia or in other media unless they have deep pockets. Especially if you live in a third world country, scientific knowledge is only available to the rich. This is strange, because the scientists who contribute to and edit for scientific journals don’t get paid for their work by the publishers of those journals. The universities where the scientists are employed are funded by the taxpayer. We pay twice, first for scientists who produce scientific knowledge and then to the publishers to acquire that knowledge.

Why scientific journals charge (high) fees

A detailed explanation can be read here, but I’ll summarize. In the past publication of scientific journals was done by scientist-driven organizations and it was rarely a profitable activity. This changed when the Science Citation Index was started in 1960, which measured how many times journal articles were cited to determine the influence of articles and their authors. This in turn gave rise to the impact factor, the average number of citations of all articles in a journal to determine the influence of that journal. As a consequence scientists started to prefer publishing in the most prestigious journals.

At this point of the story the commercial publishers make their appearance. They saw profit, gradually acquired more and more journals and started charging disproportionate prices for the prestigious journals. The prices for subscriptions increased with 215% from 1986 to 2003, while the inflation rose with merely 68%. In the same period digital distribution became popular, which should have led to a reduction of distribution costs according to common sense. Scientists have agitated against high journal prices, especially those owned by the publisher Elsevier.

Open Access publishing as an alternative

But then the concept of Open Access arose to fight this evil, enabled by the Internet. Open Access publication of scientific articles means that they are freely available and carry no restrictions on their use. It has been estimated that 7,7% of all articles published in peer-reviewed journals were Open Access in 2009. And that rate is still growing. The best example for this is the Public Library of Science (primarily natural sciences).

Open Access has found appeal in the Netherlands too: my own university (Leiden University) and several other organizations collaborate on the Journal of Archaeology in the Low Countries. A professor of Psychology at Leiden University also recognizes the importance of Open Access publishing. In fact, all Dutch universities committed to Open Access when they signed the Berlin Declaration.

But 7,7% still means that the vast majority of scientific knowledge is locked away behind the paywalls of the publishers. Also, the adoption of Open Access publishing in the Netherlands has been stagnating since 2007. One solution which has been proposed to elevate Open Access publishing is to make it mandatory, just like Harvard University and the Massachusetts Institute of Technology have done.

My stance in this

If research is funded by the taxpayer the taxpayer should have the right to access that research freely, anything else is an injustice. If Dutch universities don’t take action by themselves the government should. But since there is no indication that is going to happen anytime soon, the more important question is what I can do.

Right now I’m at the verge of pursuing a scientific career. I’m soliciting for a PhD position at Leiden University and job as a researcher at Erasmus University Rotterdam, both in the field of public administration. It seems like I’m already getting off on the wrong foot, because I submitted an article for review to the aforementioned Journal of Public Administration Research and Theory just a week ago! This is my master thesis, which I’ve rewritten in the form of a shorter, publishable article.

I keep telling myself this is justified because I’m at the beginning of my career and can afford to be picky in where I’ll publish, but I feel like a hypocrite. I don’t really have an excuse, because there are Open Access journals in the field of public administration. They’re not prestigious and I’m not sure what to do with my article – let’s see if it’s accepted for publication first – but if I do land at a job at either university Open Access publishing will be my first priority.

What I’ve done

It started when I decided to look up the article for the e-mail client I use daily, Evolution. The version I encountered before I started working on it was quite outdated, so I’ve improved it considerably to a state which is more or less finished.

As a Classical Antiquity nerd I’m quite interested in ancient Greek and Roman cities, so I ended up reading the article on Sybaris. This article had serious issues, because like many other articles on ancient cities its content was copied from a 19th century dictionary. That dictionary didn’t mention the city was excavated in the 20th century of course. After finding sources about its excavation and figuring out what the obscure abbreviations for references to ancient works meant, I give the article a rigorous update.

But as you can see it’s not finished yet because good photos are still missing. Note that three partially overlapping cities were built on the site and that it’s difficult to be certain which remains are on the photos. I only managed to add one which was sent to me by an Italian scientist I contacted, but the quality isn’t great. I’m playing with the thought of getting a ticket to Bari for € 28 with Ryanair, a second hand Nikon D5100 and the 35mm prime lens to make some pictures of this site and others for use on Wikipedia myself.

There are many more articles of ancient cities and other archaeological sites which need a lot of work. Apart from the Mediterranean world, there are also many articles about cities in Mesopotamia and South Asia. Unless I find a way to clone myself it’s simply too much, so I try to focus on a few articles and make minor improvements to the rest. This includes things such as adding the the geobox river and infobox ancient site templates to articles.

Why I do it and you should help

But why are you doing this, you might ask? Not only is Wikipedia’s mission to spread knowledge freely totally cool, it’s exciting that thousands of people read what you write. You become an authority to them to some degree, you feel like you have some power over them. If you get killed by a lightning strike tomorrow, the information you contributed to Wikipedia will still live on.

Even though I mostly edit the more obscure articles – popular articles don’t have much room for improvement – an article like Sybaris gets 5.000 visitors a month. Babylon on the other hand, to which I’ve made a minor contribution, gets 170.000 a month.

To give a motivation which is less self-serving and based on illusions of grandeur, the desire to contribute is reciprocal for me. I benefit from good Wikipedia articles written by others, so I wish to return the favor. And because I’m slightly perfectionist, it is simply frustrating for me that some articles are so inadequate. Also, contributing to Wikipedia allows me to compensate for having lost the opportunity to do volunteer work in India to some degree.

Considerations

Nowadays you can get 15 inch laptops for little more than € 300, but if you buy cheaply, you pay dearly. Most of those cheap laptops have relatively low display resolutions (with 1366 by 768 pixels being most common) and because they’re meant for consumers they almost always have glossy displays. Build quality often isn’t good either.

That’s why it’s smart to buy more expensive laptops which are more sturdy, often the models intended for business customers. These also have matte displays, which don’t turn into a mirror if there’s a lot of sunlight. But most importantly, if you’re willing to pay a premium, you can get a high-resolution display of 1600 by 900 pixels or even more. Quite a difference in screen real estate.

The choice for the L530

A few weeks ago, one of the cheapest 15 inch laptops available in the Netherlands with a 1600 by 900 pixels display resolution was the Lenovo ThinkPad L530 for € 752. Even though this is “budget” in terms of ThinkPads and this model has an ordinary plastic case, the reputation these laptops have for good build quality still stands. And it has a matte screen. As an added bonus it also came with Windows 7 instead of the disaster that is Windows 8. My brother was discouraged by the high price compared to the usual low-end consumer notebooks, but I managed to convince him.

I won’t review the laptop in detail because that has already been done by NotebookCheck. Like they say, you shouldn’t expect it to be as sturdy as the premium ThinkPads with its plastic case, but it’s certainly a good deal for € 752. After buying and unboxing the L530 however, we were surprised to find out that the touchpad had a problem: the mouse cursor moved like a person stuck in quicksand.

When my brother read the review on NotebookCheck again, he noticed they had also experienced a “major problem with the touchpad’s responsiveness”. They assumed this problem was restricted to their test laptop. Curiously, I couldn’t find anyone else experiencing the same problem on the Internet. It was time to contact Lenovo and get them to fix the laptop.

Lenovo technical support

The first impression wasn’t so positive. We had to sign a form to give them permission to re-install Windows 7 if necessary and some other trivial matters, but this form had many Dutch grammar and spelling errors. I still remember my ordeal with Acer’s helpdesk, but expected Lenovo to do better. Apart from the form though, Lenovo did its job properly: when the laptop was returned the touchpad worked fine.

Only the reparation report raises some eyebrows. It claims they replaced the display and the “kbd (keyboard?) bezel”. I have no idea what those have to do with the touchpad.

Steam has been packaged for Fedora too (albeit not by Valve) so the most convenient way is to add the repository for it as instructed here. Because these instructions for Fedora 17 are slightly outdated I stopped following them after the yum install steam command. The more recent Steam package from the repository pulled in all of the dependencies needed to get Steam itself running. But while Steam itself might run, the games did not run for me without some more effort.

Some more dependencies are needed. The problem is that Steam is 32-bit, so if you’re on the 64-bit version of Fedora you need to download a lot of 32-bit packages of which the 64-bit equivalent is already present on your system. The first step is to install the 32-bit version of your display driver libraries, which I did with yum install xorg-x11-drv-catalyst-libs.i686 (the i686 suffix tells yum to get the 32-bit version). That’s for people who have AMD video cards, those with NVIDIA video cards should use the yum install xorg-x11-drv-nvidia-libs.i686 command.

After doing this my games would start, but I didn’t have sound. To get that fixed I simply installed the 32-bit versions of all packages which I thought to be audio-related. From this topic I came up with the command yum install pulseaudio-libs-glib2.i686 libao.i686 esound-libs.i686 alsa-oss-libs.i686 alsa-plugins-oss.i686 alsa-plugins-pulseaudio.i686 audiofile.i686 which fixed the problem for me. But before entering this command I had downloaded some other packages I don’t remember anymore. Not sure what’s exactly needed, for me it was a matter of a few Google search queries and trial and error.

Hopefully someone else can clear this up or the package could be made to pull in all the necessary dependencies. Preferably Valve will support Fedora 18 officially soon and make a 64-bit version. Valve is totally awesome for porting Steam to Linux. Maybe I can even remove my Windows 7 partition completely in the near future, since the games are the only reason Windows 7 is still on my hard disk drive.

I don’t expect the improvements I’m hoping for will arrive soon because Evolutions suffers from a shortage of developers. But at least I’ve filed bug reports for everything which bothers me now. You can’t criticize free software without bug reports.

I haven’t reported all bugs I have experienced: bug #687360 for example has already been fixed for 3.6.3 apparently, but that version hasn’t landed in Fedora 18′s updates yet. I noticed some nasty bugs with editing contacts in address books too. In some cases the name in the “File As” entry would change without my input upon saving a contact when using the older Evolution version in Fedora 16. But I haven’t been able to reproduce this in Fedora 18.

• Bug #692531: dialog appearing for invalid SSL certificates isn’t pretty
• Bug #692533: Edit Rule dialog has drop down menus with a lot of blank space
• Bug #692535: option to have filters active on an account should be enabled by default
• Bug #692541: tab widget in Preferences dialog looks wrong
• Bug #692542: never loading images in HTML e-mails is a bad default setting
• Bug #692555: “Personal” address book does not appear in birthdays calendar by default
• Bug #692557: sorting by date descending disrupts thread view of messages
• Bug #692558: language detection for spelling checker
• Bug #692572: “Messages” column should use space more efficiently
• Bug #692573: use vertical view as default instead of classic view
• Bug #692574: remove status bar
• Bug #692576: use list view as default instead of address cards
• Bug #692577: list view shouldn’t have columns for faxes by default
• Bug #692578: display “File As” column by default in list view
• Bug #692579: make view settings apply to all address books
• Bug #692636: doesn’t delete certain contacts
• Bug #692713: calendar widget arrows don’t look correct
• Bug #692716: all calendar items of an icalendar file disappeared
• Bug #692743: e-mail shows preformatted lines which do not word wrap if replying

Fedora 18 hasn’t been received well by some. A review of the KDE spin of Fedora 18 has been quite negative and the same reviewer is also very critical of GNOME 3.6, I didn’t try the KDE spin but I’m quite positive about the GNOME spin. I did a fresh installation, not do an upgrade.

Installation

This release was postponed multiple times because the redesign of the Anaconda installer wasn’t ready yet. It is still problematic in the final release: as soon the live desktop had started up and I chose to install to the hard drive to start Anaconda, it crashed. I suspect this happened because a connection to my router was established just while Anaconda was starting up. I had already witnessed a similar crash when trying the alpha versions on my laptop.

It wouldn’t start again so I had to reboot, after which I waited for the network connection to come up before starting Anaconda. This time everything went fine. Maybe I’ll try to reproduce this later so I can file a bug report. I agree with those who think Anaconda still needs a lot of work. But I also think the new design is a step in the right direction, it’s a diamond which still needs polishing.

Post-installation guide

First of all let Fedora download the latest updates and then add the RPM Fusion repositories. I wanted to get the proprietary AMD drivers first, there are instructions for that here. Take notice of what I wrote there in the comments. You need to follow the instructions here and add another repository because RPM Fusion apparently doesn’t have the package for the Flash plugin anymore. However, only install flash-plugin, the rest is not necessary and nspluginwrapper – while necessary if you want Flash working with GNOME Web – drives SELinux crazy. Also read these instructions to get the FreeType patches so you get better font rendering, this involves adding one more repository.

After all this I use one more command to download all the remaining stuff. This command downloads the GStreamer packages you need for proprietary audio and video formats, the GNOME Tweak Tool, the GNOME Web web browser, the GIMP and the Google Droid fonts (with GNOME Tweak Tool I change my default, document and monospace fonts from the default Cantarell, Sans and Monospace to Droid Sans and Droid Sans Mono, I also the font size of all fonts from 11 to 10):

yum install gstreamer-plugins-bad gstreamer-plugins-bad-nonfree gstreamer-plugins-ugly gnome-tweak-tool epiphany epiphany-extensions gimp google-droid-sans-fonts google-droid-sans-mono-fonts google-droid-serif-fonts

To download the XeLaTeX packages I need (which include support for the Dutch language, the APA style and the Linux Libertine font) I give the following command:

yum install texlive-xetex texlive-collection-langdutch texlive-memoir texlive-biblatex-apa texlive-libertine

Bugs

Here’s a list of bugs in either the Red Hat or GNOME bugzilla, for which I found an existing report or filed a new one:

• Bug #893218: Anaconda crashes when it starts
• Bug #854201: Ask users for Country / Language / City to determine the correct locale settings
• Bug #690750: System Settings doesn’t change locale completely
• Bug #904000: free ati driver not enabled with HD7850 after installation, uses LLVMpipe instead
• Bug #692518: processes can use more than 100% CPU according to System Monitor
• Bug #904014: Anaconda doesn’t ask for host name
• Bug #878433: Removing default English keyboard layout doesn’t have effect in the installed system
• Bug #904052: Shotwell should not be the default application for viewing images
• Bug #904055: only root can mount an external hard disk connected through eSATA
• Bug #692519: hidden drop down menus for keyboard shortcuts difficult to discover
• Bug #692520: graphical corruption if taking screenshot of single window
• Bug #905103: Firefox occasionally freezes unpredictably and takes out X with it
• Bug #692923: no unlock button for hostname in “Details” (System Settings)
• Bug #693057: implement options to change font and font size
• Bug #693058: option to suspend a desktop computer should be visible without Alt key

There are more bug reports I have to file for Evolution, but that would make this post too long.

Since then I’ve lowered my consumption of meat, but it was not an issue for my family. They think they don’t eat it much, while I do. Because I wasn’t motivated to make my own vegetarian food when my mother prepared non-vegetarian food for the whole family, I used to eat along.

Recently I changed my behaviour because I thought I was lazy. Being served non-vegetarian food by others was no reason to stray from the virtuous path of sustainable living. This led to some discussion at the dinner table, giving rise to the question: at which quantity is meat consumption sustainable?

Statistics and some assumptions

The Economist has statistics on meat consumption per capita for 2007. In the complete data they also provide the world average, which is 38,7 kilo or 106 gram a day.

I make two premises. The first is that everyone should not consume (much) more than this average. If you look at the statistics you see the developed world consumes a disproportionately high amount of meat while most of the developing world consumes very little. This means the developed world puts the greatest strain on our environment. I don’t think we have more right to burden the environment than the developing world, so what we are doing now is unethical.

The second premise is that no further environmental degradation should occur. If that is what we want, we should not increase our total meat consumption above the 2007 average. However, I’d say that even in 2007 the global environmental damage caused by meat production was already too much, and it needs to be even lower. I’d say an arbitrary number such as 30 kilo a year seems acceptable.

Solutions

Ultimately, reducing meat consumption is not going to happen with a growing world population. At a certain moment meat prices will increase because of supply and demand, but when that happens the environment has probably gone to hell already. I’m sure many people are not as concerned over this issue as I am, so they’re not going to change their behaviour out of their own motivation.

This means government has to intervene. The best solutions would probably be taxing meat (just like we levy an excise tax on petrol in the Netherlands because of its environmental impact) and drastically reducing the world’s population. But if you want to change the world you should take responsibility and start with yourself.

How much do i and my family consume?

According to the statistics the average Dutch person consumed 71,3 kilo meat in 2007. I think my family is probably below this average, so in that regard they are probably right when they say they don’t consume much.But I suspect my family may consume more than the global average of 38,7 kilo.

But this is something which needs to be measured. I will do so by calculating the total weight of all meat products appearing on our supermarket receipts for one month and then extrapolate that to a year. I estimate my own consumption is probably no more than 21 kilo a year (based on 400 gram a week), but this is also something I will measure.

Update 14 February

After measuring the meat consumption of my family for 19 days and extrapolating that to a year, it turns out we consume far less than I expected. I consume 19 kilo a year, the rest of my family 22, 29, 31 and 34 kilo. All far below the average for the Netherlands and below the world average.