Avian’s Blog

Two weeks after I enabled comments on this page (and after two actual comments posted), I caught the first (futile) attempt to exploit an XSS vulnerability in my comment system. This is quite an improvement over my previous try with the stock cgicomment installation which took only a couple of days before bots started posting unmanageable amounts of spam and other nasty stuff.

Hi, Jonny. You'll have to try harder than that this time.

Today I went on a tour of the power plant Krško, the only Slovenian nuclear power plant (the trip was organized by EESTEC).

We weren't allowed to bring anything with us on the tour (hence a lack of pictures of the most interesting stuff in this post). They also performed some kind of a biometric scan of my right hand, gave me a temporary RFID tag and assigned an armed guard to each group of visitors. No dosimeters were given though (like when I visited the small research reactor in Podgorica a few years ago). But then we also didn't visit any of the areas where they handle radioactive stuff.

A model of the fuel bundle that contains pellets of enriched uranium oxide.

They gave us a complimentary nuclear power plant starter kit containing three (fake) fuel pellets. Now I only need some 7 million more of them to get enough for the whole bundle :)

Our guides really took their time and answered all of our questions, even some that probably weren't very pleasant from their point of view. It seemed to me that they are working hard to present the power plant to the public in the friendliest way possible. They explained the basic principles (Krško uses a PWR-type reactor) and also told us quite a few things that were surprising to me.

• Every physics book I read says that the power output of the reactor is controlled by moving the control rods in and out of the reactor. I learned that this isn't so, because having control rods only partially inserted would burn the uranium fuel at the bottom of the reactor faster than at the top. This is a bad thing (they try very hard to keep the reaction rate constant throughout the volume of the reactor). So a better way to do this is to vary the concentration of a neutron absorber in the coolant. Rods are only used for fast changes (like in an emergency). Interestingly the Wikipedia article linked above gets this right.
• The plant was designed by the Westinghouse company, while the actual work was done by workers and specialists from local (then Yugoslavian) companies. The design specifications were so strict that all markings on the equipment, pipelines, etc. are in English. For example even the color codes of the wires used in the buildings are according to American standards instead of European. I'm surprised they managed to adapt the plans to the 50 Hz frequency we use here :)
• Hruske writes how nuclear power plants run on QNX. Krško actually doesn't run on anything. All control loops are still implemented with analog electronics (which is considered a Good Thing by the maintenance engineer I talked to). They recently installed a new digital monitoring system though. It runs on some software that nobody there actually knew how it was called (but they all agreed it wasn't Windows). The main control room is full of switches, analog gauges and warning lights.
• New fuel elements are transported to the plant from USA first by a ship and then by trucks. A couple of times they also flew them in on an airplane. It's interesting that usually our newspapers are full of reports how a convoy of trucks carrying nuclear fuel will be traveling through this-or-that distant country but I never seen an article about how they transported fuel for Krško. On the second though this may be a good thing or there would be people chaining themselves to trucks and things like that.

I've just blacklisted the following blocks of IPs on mail.tablix.org:

196.217.0.0/16
196.206.0.0/16
41.250.0.0/16

According to the Whois database, they all belong to some Moroccan ISP.

Last week I've been getting incredible amount of spam from there. I can handle spam sent to my personal email account (it gets filtered anyway), but these guys have been sending spam to the tablix.org's mailing list manager address with my address in the Sender field. That way I got some 200 bounces from my own machine telling me I issued a wrong mailing list command.

Luckily they only used my address. I would hate to think that other people would be getting such mails from my server because of some stupid spammer.

So, if you're from Morocco and can't send mail to any mail accounts at tablix.org, go bug your ISP or something. I'll try to remove the blacklist entries when I don't see any rejected messages in my logs for a month or so.

By the way, these IPs are from the same block as those that caused problems with the mailing list manager this June.

Almost every single time I hear a discussion about the OLPC project somebody says that people should focus on providing food and other things that are more urgently needed by children in undeveloped countries before giving them a computer.

I think that this comment on Slashdot should give those people some idea why such arguments don't hold water.

Hello, world!

First post using Nanoblogger 3.3RC5.

RSS feed seems to be broken. Sorry, will look into that soon. Upgrade also changed the URLs of all archived posts. On the other hand I can now enable user comments.

Please report any broken links. Thanks.

My recent work on the redesign of the Galaksija computer reminded me of some old books about Spectrum I have on my bookshelf. One of them is the The Complete Spectrum ROM Disassembly. I never actually read it before. Back when I was playing with Spectrum the lack of pictures and confusing assembly language was enough remove all my interest.

Now when I'm reading it, I am surprised at how much has changed since then. Here is a book with the complete contents of Spectrum's ROM (its "operating system"), together with detailed commentary for each function and each line of code. All that from authors that were not affiliated with Sinclair in any way (as far as I can see from the book).

Publishing a book like that today would probably be like taping a big sign "Sue me" on your forehead. Reverse engineering is not something people usually like being connected to (not to mention that it can be even illegal in some weird countries). It is either done in laboratories of big companies (from which you don't hear much but I'm sure they exist) or more or less anonymously on various internet forums (just look for example at game console modders). Imagine how a XBox complete ROM dissasembly book would be welcomed by Microsoft.

In a few days I will stop receiving mail sent to my old email address tomaz.solc@siol.net.

So, from now on please send any mail to my new address (replace siol.net part with tablix.org).

In case you don't know already:

I've managed to read those old floppies without errors and I've put the contents on-line. You can download them from here.

All games more or less work in DOSBox. Nethack has some problems with display (it says it needs some driver loaded in config.sys to work correctly). Spacewar works best if you use Hercules emulation (dosbox -machine hercules)

To run Ranadinn in DOSBox, follow these steps:

1. Run DOSBox: dosbox -machine hercules
2. Mount C drive in DOS: mount c /path/to/ranadinn
3. c:
4. Run SIMCGA (for some reason DOSBox CGA emulation isn't good enough for this games): simcga40.com
5. Run Ranadinn. You will have to type this blindly because SIMCGA messes up DOSBox' prompt: ranadinn.com
6. Enjoy.

I just found this set of 5¼" floppies with shareware games on them. I remembered that my father bought a couple of these many years ago.

We had a 286 computer with monochrome graphics at home back then (our first Intel compatible). It was made by Mlacom, a Slovenian computer company that still exists (although I think they lost all good reputation by now).

These disks contain three games: Ranadinn was a graphical role-playing game (there's a screenshot on the cover) where you wandered around in forests, entered cities and usually died of hunger after playing 5 minutes. Spacewar was something similar to KSpaceduel. Then there was Hack, a predecessor of NetHack.

Apart from multiplayer Spacewar I don't remember playing these games a lot. That's interesting because Nethack is now the only game that I still occasionally play. It seems that I met my favorite Unix game long before I even knew what Unix and internet was.

I wonder if these disks can still be read. I don't have a 5¼" drive at hand so I can't try them out.

Previous weekend we removed a large window in the room with my server because some wooden parts of the roof had to be painted.

It's interesting how this was reflected on the lm-sensors graphs. The CPU has a temperature controlled fan installed, so its temperature is nearly constant. The "system temperature" (I don't know where this is measured on this motherboard) on the other hand fell for almost 10°C during the night.

What Colour are your bits? by Matthew Skala is a very interesting article about the so-called intellectual property and copyright of data in digital form. It explains in a simple way why copyright is a purely subjective matter without any logic background and why people that do not have a background in mathematics and computer science usually do not see that.

The Monolith project (linked from the article) is also worth a look.

Believe it or not, this is an automatic cat feeder constructed by my dad and my uncle.

The electronics is made of a modified light timer, battery charger, 12V sealed lead battery and some custom electronics. It is fully programmable and can cope even with the most complicated diet a cat can have.

The food is dispensed by an old drilling machine that opens and closes a valve in the stainless steel container.

• Loot: $152
• DHL shipping: $70
• Having to explain to the customs officer what a generic humanoid carbon unit is: priceless

Seriously, Thinkgeek should give better descriptions of the package contents to the DHL. T-shirts, for example, were labeled only by the text that is printed on them. There was no obvious sign that the package contains apparel and not some weird organic stuff. The same goes for the plush microbes. The customs officers would probably understand the English word "toy", but "giant plush kissing disease" probably rang some alarm bells. So I had to describe each and every item in the package to the (otherwise very friendly) customs officer.

But I guess it could also be worse. After all, the package also contained stuff that could be misidentified as a plastic explosive, an ignition device and a book with "for the evil genius" in the title.

OTP-EPROM: One Time Programmable Erasable Programmable Read Only Memory

You are likely to be eaten by a grue.

Seriously, this scene from the Mahorčičeva jama near Škocjan reminds of old style adventure games. You know, the ones where you would have to go back to the village and give this-and-that to the local carpenter so he can then give you some planks which you could then use to repair this path and get to the next screen.

I'm waiting for my bus to Birmingham which leaves in approximately an hour and a half and I'm slightly bored. I've been browsing through my photographs and I found this photo of a poster I saw in a window in the administration building. I guess Japanese animation also has its fans in Lancaster.

It was really nice staying here and I hope I will come back soon.

Every time I go to sleep in my room and turn off the lights I get the impression of seeing little points of light on the ceiling. They look a bit like stars on the night's sky, but are a lot fainter.

At first I thought it was only my eyes adapting to the dark. But once I turned off the light when I was still standing and the lights remained on the ceiling, not floating in front as a typical residual image on the retina. I decided to investigate this further.

First I had to make sure this wasn't some kind of an optical illusion. I took my digital camera, put it on my bed facing upwards and set it to maximum sensitivity (ISO 800) and longest exposure time (4s). I then took two pictures: one with the protective cap on the lens and one without the cap. This way I could then remove any camera artifacts by subtracting the first image from the second. Following is the resulting image:

There is still a lot of noise in the image, but you can clearly see some green dots. These dots are larger than one pixel across so I guess they can't be caused by CCD sensor noise.

So if this is not an illusion, what is it? I can't see any significant features on the ceiling when the light is turned on. It could be light from a room above coming through the floor. This would be possible if the ceiling would be wooden and had cracks in it, but as far as I can see, the building is concrete and also all rooms have thick carpets on the floor.

Another possibility would be that the ceiling finish contains pieces of some phosphorescent material. I believe this is more likely, since I don't remember seeing these dots when I wake in the middle of the night when the light has been switched off a couple of hours. However in that case it is weird that the dots only appear above the bed. I looked for them on other parts of the ceiling but haven't seen any.

I think the following advice should be added to the next revision of the Lonely Planet guide:

Do not photograph large buildings standing in the middle of nowhere. It may turn out that one of them is a nuclear power station and then otherwise friendly local people start to ask weird questions like "Are you going to blow it up?".

In my case, it later turned out that I had found the Heysham power station while walking along the coast from Morecambe. I guessed it has something to do with large amounts of electrical power since there was a large high voltage line coming out of it, but as far as I knew it could have been a fish processing factory (I could see no cooling towers or containment buildings that you usually see near things that run on nuclear power). Anyway, only the third person I asked actually answered me what the building is (right before asking me back what my evil intentions are). The previous two just looked at me suspiciously.

Last Sunday I went to Lake District with my supervisor and two other professors from the Department of Physics. We had plans to go hiking, but it turned out we were a bit short on time because we had to go to Manchester in the evening. So in the end we only went for a short walk around the Castlerigg stone circle. Any way, the view very still fantastic.

Lake District is a national park in Cumbria, north of Lancaster. I guess the name is quite appropriate because basically every deeper valley between hills is filled with water. The mountains aren't very high compared to what you see in Slovenia and you hardly see any exposed rocks. If there's no grass the rocks are covered with moss.

The weather changes very rapidly here. In an interval of two hours it went from snow through rain to a clear sky and back to rain again.

This is the view of some of the higher peaks from the Castlerigg stone circle.

University underpass (bus for Lancaster city center stops here).

Somewhere between Engineering department and Grizedale college.

A view of Grizedale college from the building I'm staying in.

Yesterday I finally got internet access here at Lancaster University. It took me three days to find the proper way of doing this: first I tried to get my Eduroam account working here (University's web page said that visitors should use it to get internet access), but after some phone calls and visits to various help desks it turned out that the English Eduroam database isn't actually connected to the continental one. So much for the universal access provided by Eduroam.

Now I have a temporary computer account for visiting students and for which I got some 15 pages of instructions on how to use it (it includes stuff like email, Unix shell and Windows accounts, storage space, etc.). So far I figured that the internet access I'm used to doesn't exist here - web traffic is only allowed through a proxy (which may be monitored and/or recorded). Anything that is not HTTP traffic is mostly filtered out. Email for example is only allowed through their official web interface (SMTP port is blocked). Luckily SSH appears to work.

Here are some pictures of the room I'll be using for the next two weeks. It has a network connection (20-something pounds per day extra if you want to use it - I don't), phone (I don't even want to know how much that costs) and a cleverly placed electrical socket behind the curtain, which I only found yesterday.

This is an interesting way of forcing the user to use as little hot water as possible. Give exactly two choices: cold or scalding hot water. Guess which one you're going to use? Oh, and the green faucet is for the hot water.

At least they tell you about it...

Fabricated using 1.2 micron CMOS technology, this chip holds a unique Cyberpipe logo on the 3rd metal layer. The logo measures approximately 20 μm on the side and is located over some of the digital signal lines entering the integrated circuit in hope of guarding it from the bad carma outside.

The chip is a must-have (in much the same way as the CP Cap) for any Cyberpipe member and is available for sale at the manufacturing price of 2.5€ per die. Only one wafer of these one-of-a-kind chips have been fabricated, so hurry while the stock lasts.

Yes, this was a joke. The image was made in GIMP, using this microphotograph as background.

My server that runs services at tablix.org will be down tomorrow for a few hours because of a number of hardware and software repairs. DNS resolving problems, bounced emails, unavailable web pages and all sorts of other weird phenomena can be expected.

It turned out that my EN-X temperature control circuit has a bug and can burn out one of the operational amplifiers if you turn the temperature setting knob too fast in either direction. This problem didn't look very serious when I first noticed it a few months ago because temperature regulation still seems to be working. However I'm not really happy with one of my machines running with a half-working CPU fan. And since I now also have a list of some other tasks that involve a reset (like a kernel upgrade and some UPS rewiring) I decided it's finally time to trow away those 200 days of uptime.

My refrigerator light bulb works better with the door closed.

-- jcostantino on Slashdot

Robotfindskitten (apt-get install robotfindskittten on Debian) is another free software game with a profoundly deep background. The similarity of the default user interface with Nethack is just another plus that adds to the experience (although there are a lot of other interfaces available, for example a fancy 3D accelerated OpenGL port for the ASCII challenged).

"GO!" I told the box as it began to roll out of my workshop into the frozen desert beyond. "FIND KITTEN!"

-- The Book of Found Kittens, pages 43-4, author unknown

Good stuff :)

While I agree that these videos look really great and all I think GNOME developers should really take John Williams' proposal seriously.

Making "an intuitive and attractive desktop" isn't only about removing some buttons from dialog boxes and reordering the rest. A usable desktop is also a desktop in which you don't have click through a couple of error messages before you can begin working.

Such a desktop also doesn't replace a working PDF viewer with a broken one just because the latter is "simpler and more convenient".

Compare a PDF document in Gnome 2.10:

with the same document in Gnome 2.12:

LEGO Technic difference engine

GNOME Bugzilla presents the Ultimate Bug Report.

I have just found an amusing little program by Taper Wickel on my (half-broken) hard disk. It uses some simple math to predict how long it will take for humanity to destroy itself. And if your simulated little earth manages to survive the next 500 years (which in my case is yet to happen) you win.

It's a fine replacement for fortune in your .bashrc.

Here is the link to source (it is in the public domain). Have fun.

Last August a 6 year old 13GB Seagate Medalist failed in my server. In October, a 5 year old 30GB IBM Deskstar failed in my desktop.

Now, my desktop won't boot from a three months old 200GB Western Digital because of the infamous { DriveReady SeekComplete Error }

Before August, none the disks in my machines have ever died. Should I expect this row of failures since I was cheating on the specified mean time between failures for the drives I have been using?

The road to hell is paved with NAND gates.

-- J. Gooding

I still can't find the cause of XKB error messages, but I'm getting close.

I suspect there is a race condition when something runs xkbcomp to compile the XKB keymap, but I can't find who exactly is doing that. After going through the source or gnome-settings-daemon, libgswitchit and libxklavier I filled a bug report at Debian BTS.

To be continued...

This is the condition of the road to Javornik near Črni vrh (that round thing is the top of a road sign).

One fact about computers that even technical people often forget: The job of the computer is to make your job/task easier - it is not the other way around.

-- Joeytsai on Slashdot

My friendly incompetent internet service provider is again blocking all external mail traffic. So don't be surprised if you can't send me a mail or use one of the mailing lists at tablix.org.

This time I'll restrain myself from calling their support line. Everytime I call and complain about mail problems the marketing droid on the other side wants me to open my Dial-up networking properties and Outlook express settings. After I tell him that I don't have this stuff on my machine and that their mail server is returning 450 Unable to find domain after MAIL FROM SMTP command he simply rewinds his tape and starts again: OK, I understand. Now click on the Start button and open the Control panel...

After some reading of the stuff in /usr/share/doc/udev I found out that putting lp into /etc/modules brings back the lp0 device. Suddenly things don't look that bleak anymore.

Still have to figure out how to get /dev/ttyS0 and friends back though.

It's almost midnight, I've just spent 5 straight hours writing a report for one of the term projects I've been working on and now I just have to print it.

No problem. I click on the Print button, but the printer remains silent. Great, why must things always break when you need them the most? Some 5 minutes later, I dig out this from /dev/log:

Jan 18 23:38:50 orion lpd[9645]: /dev/lp0: No such file or directory
Jan 18 23:39:30 orion lpd[9669]: /dev/lp0: No such file or directory
Jan 18 23:40:18 orion lpd[10641]: /dev/lp0: No such file or directory

WTF?!

Ok, no problem. Some broken script obviously wiped out some of my device files. Luckily, I still have MAKEDEV, right?

orion:/dev# MAKEDEV lp
.udevdb presence implies active udev.  Aborting MAKEDEV invocation.

Argh. So MAKEDEV thinks that udev should take care of device files and udev obviously only cares about fancy new stuff like USB and Firewire. How on earth am I supposed to use my printer on the parallel port? I don't remember the major and minor numbers for the lp0 device and I don't feel like greping the kernel source right now. Maybe there's another lp0 somewhere?

avian@orion:/dev$ locate lp0
/home/avian/morphix/live/dev/lp0

Finally! I knew that all that junk in my home directory will someday turn useful. A quick edit of the /etc/printcap and the printer begins churning out first pages of my report.

I guess it's time to switch to CUPS or something.

Today was the last day of lectures in my five year undergraduate study program. Basically this means that I'm getting near the end of step one in the three-steps to profit of the new economy:

1. Study
2. ????
3. Profit!

The only problem is that I still haven't quite figured out what step two is supposed to be.

I heard several (mostly older) people say that the electric current is "stronger" in the evening. This was usually said in reference to cooking on electric stoves - for example you need to set the stove on a lower setting in the evening to get the same effect as with a higher setting in the morning. Recently I got into an lengthy argument about whether this is true or just an urban myth, so I decided to check it myself.

The UPS on my server keeps logs of the RMS voltage in the electricity distribution network so this seems to be a good point to start.

As you can see the voltage in the socket can vary as much 15V during a day and is far from the declared 240V.

On the other hand the digital voltmeter in the UPS can't be trusted, since it wasn't designed for such precise measurements (its only purpose is to tell the UPS when to switch to battery power). Because of this I double checked these values with a true-RMS voltmeter and it looks like the UPS measurements are constantly off by 10 volts (UPS shows 228.0V, voltmeter shows 238.0V), however the relative values seem to be quite accurate.

So let's say that I want to boil 2 liters of water on a 1kW electric stove. Such an electric stove has a heater with resistance of R=(240V)²/1kW=57.6Ω. To heat 2 liters of water from a room temperature of 20ºC to 100ºC you need Q=m·c·ΔT=4200J/kgK·2kg·80K=672kJ of heat. If the electric stove is operating at its declared power it will provide this quantity of heat (if we ignore the heat loss) in t=Q/P=672s=11.2min.

Now let's see how much time it would take to boil the water if we take into the account the swings in the line voltage. The graph shows that the lowest voltage measured is 222.2V. This is U_min=232.2V when corrected for the UPS measurement offset. At this voltage, the stove has a power output of P_min=U_min²/R=936W. It will heat 2 liters of water to boiling point in t_max=12.0min.

On the other hand the highest measured voltage is U_max=247.5V. This gives the power output of P_max=1063W and the water will boil after t_min=10.5min.

The maximum difference is obviously 1.5min or around 13%. This is value falls somewhere in the middle: I can't say that I would notice if the water for my spaghetti would begin to boil 90 seconds too early, but it is quite possible that some people can notice that. On the other hand most modern electric stoves have thermostatic control. This means that the temperature of the stove is controlled with some kind of a feedback loop. This loop should attenuate any effects of the line voltage swings to such a degree that I'm sure noone could notice (even if with a stopwatch I don't think you could measure any effects but I'll try to prove that at some later time).

Happy winter solstice everyone.

It seems that with the recent addition of Tablix package to Debian Unstable, http://www.tablix.org finally became the first hit on Google for tablix. Until now the first hit for this query was the old page at www.kiss.si which only contains a link to the new page.

It took 10 months for Google to find out that the page has moved, which is kind of interesting. Also interesting is that Tablix Systems, which appears to be a Danish software company that has nothing to do with my timetabling software, barely gets on the first page of results.

Tablix is now also the 19th hit (second page) for query timetabling, a great deal behind FET, a free software project similar to Tablix, which is on the 8th place.

I guess the author of this NetHack patch died of hunger one time too many in the Mazes of Menace. Now if you are feeling hungry, your deity is angry with you and there are no edible creatures around you can simply pull out your cell phone and order a pizza. Wonderful.

I was etching the printed circuit board for my keylogger project today. When I tried to turn the board over in a bath of hydrochloric acid and hydrogen peroxide I got this warm feeling in my left palm. I quickly withdrew my hand from the acid and saw that the protective glove I was wearing had three little holes near the center of the palm.

I quickly washed my hands under running water and everything turned out fine in the end (the circuit board looks great), but I'm never going to buy another pack of "beste qualität" Spontex disposable latex gloves again. When I checked four other gloves in the box, two disintegrated when I just tried to put them on. And this was a freshly opened pack I bought a week ago (no "best before" or production date printed anywhere on the box).

I wonder if these guys are also making condoms of such fine quality.

We get snow.

...if looked at through the Internet Explorer 3.0 on Windows 95.

It seems google.com is just about the only page that still renders correctly with this browser.

I guess this computer is also safe from worms and viruses these days. I'm sure any self-respecting malware needs at least Windows 98.

Make backups

Today I had the opportunity to visit my Faculty's integrated circuit fabrication line. It was quite fascinating although not exactly what I expected

My first reaction was that it looks like a set for an 80s science fiction movies. They are using some quite old technology - the line was originally meant for a 5μm process, but thy gradually upgraded it and today they can manage everything up to around 1μm. Green monochrome monitors, big clicky keyboards, 5.25" floppy disks, blinking lights, large racks of equipment, people in protective suits, low vibrating hum everywhere...

There were of course a few objects that quite obviously didn't fit there. First a modern washing machine with a box of detergent (?! - probably for washing protective suits, forgot to ask, but it was like seeing a ghost). Then there was a single modern PC (running Windows XP) - Excel was running when I saw it so it probably isn't running anything critical.

The whole process isn't as automated as I first thought. Silicon wafers are transferred by hand between various furnaces and machines. Etching is also done by hand as far as I could see. In temperature controlled baths and with carefully controlled chemicals but they still put the wafers in and out by hand.

Also I expected I wouldn't be allowed in the actual rooms where the fabrication takes place. I was told that those places need to be kept very clean and that anyone entering the room can bring dust particles. But in the end I only had to put on a protective coat and that was it.

Hint: sending spam to a lot of free software authors and maintainers is not a good way to start a new open source community web site. Words like "15% referral bonus" on the second line in my opinion further reduce the number of people that will actually read past the first paragraph.

It's perfectly plausible of course that they really wish to support open source software with this move like they say in that email. But the mere fact that they are advertising their site by sending spam greatly reduces their credibility. If the people behind makezero.com are really involved in open source movement they should know that their target audience won't take spam lightly.

I'm just getting too many emails from people that are offering to "help me make a little money", not to mention helping me "grow my distribution". If it weren't for the words "open source" I'm sure this one would end in the "Junk" folder like the others.

I've just got my first case of bad electrolytic capacitors.

I am trying to reuse an old 486 as a wireless bridge. Since this computer served as my ADSL router and web server for four years I thought it was going to work without problems. I only inserted a new wireless ethernet card into a free PCI slot and turned it on. I've heard a small click and thought something got into the CPU fan. I turned the computer off, cleared cables from the vicinity of the fan and tried again. This time there was a series of loud bangs coming from the power supply and a strong smell of burning plastics and simultaneously all LEDs on the computer went on.

It currently looks like the two large high-voltage 330µF capacitors in the power supply blew. They were both too hot to touch when I opened the PSU and one of them has an opened top.

I just hope nothing else got fried in the computer.

What are we going to do tonight, Brain?

Same thing we do every night, Pinky. Try and take Cyberpipe's link down!

A quote from Visual C# .NET Step by step, page 262, chapter 14:

• Keep Visual Studio .NET running and turn to Chapter 15.

• On the File menu, click Exit. If you see a Save dialog box, click Yes.

This suggests that there are people who have already read 260 pages about C#, understand "properties to access attributes" and who have to be explicitly told to turn the page of this book if they want to continue reading. Obviously the average Windows programmer must be in even worse shape than I thought if the instructions on how to exit Visual Studio must be repeated for him at the end of each chapter.

Today is a "no car" day in Ljubljana. This is a way city authorities are using to promote public transportation.

So far so good. I'm all against cars. I think they are an fundamentally inefficient way to travel. Public transportation isn't very popular in Slovenia and basically everyone drives a car, so in my opinion such an initiative is a good thing.

What I don't get is the way the city implemented this idea. First, they closed the city center for all traffic. And I mean all traffic, including buses (which are the only public transport in Ljubljana). This of course means that all major bus lines have been diverted to unusual routes. Causing confusion among regular users of bus lines is certainly a great way to promote them. The bus rides are free though, but what good are they if they don't take you where you want to go?

Another weird idea is that they closed the dedicated bus lanes on the major roads in Ljubljana. Buses now have to drive on the inside lanes, which means that you have to walk a couple of meters over an improvised platform to the bus (in the rain). I can't find any reasonable explanation why they did this? (perhaps they wanted to broaden the sidewalks?)

And of course it seems that the city center is closed only for those people that respect the road blocks. Because all the "smart guys" obviously ignore them.

When in doubt, use less technology.

I've read this simple piece of advice somewhere on the net a while ago. I think it nicely complements the Keep it simple rule.

It seems that more and more people rush into using a computer for tasks that are currently accomplished faster and better for example with a pen and a piece of paper or an old-fashioned blackboard. Computer-human interfaces have come a long way, but let's face it: the old way is still better in a lot of cases.

Paper doesn't need be recharged, it boots in less than a second and (best feature in my opinion) can seamlessly integrate different types of media: text, drawings, photos. Blackboards always work, aren't vulnerable to viruses that crash your computer in the middle of the lecture and won't suddenly turn all yellow/purple/blue. Also they are far cheaper than computer projectors and in my experience every single lecture given by someone using the blackboard was better and more enjoyable than a lecture given by someone using a computer presentation.

hda: dma_intr: status=0x51 { DriveReady SeekComplete Error }
hda: dma_intr: error=0x40 { UncorrectableError }, LBAsect=54455098, sector=54455086
ide: failed opcode was: unknown
end_request: I/O error, dev hda, sector 54455086
hda: dma_intr: status=0x51 { DriveReady SeekComplete Error }
hda: dma_intr: error=0x40 { UncorrectableError }, LBAsect=54455098, sector=54455094
ide: failed opcode was: unknown
end_request: I/O error, dev hda, sector 54455094
hda: dma_intr: status=0x51 { DriveReady SeekComplete Error }
hda: dma_intr: error=0x40 { UncorrectableError }, LBAsect=54455098, sector=54455094
ide: failed opcode was: unknown
end_request: I/O error, dev hda, sector 54455094

Two hard disks in two months is just too much...

I guess it's time for another put-hard-disk-in-freezer-and-try-to-copy-the-most-important-documents trick. Sigh.

I play with 3D modeling and ray tracing software now and then and I found this short video to be absolutely hilarious.

It obviously received quite a few awards, so it seems that I'm not the only one that likes it.

Wikimania got some media coverage in Slovenia and on several occasions I had an opportunity to listen to a conversation about Wikipedia between people that are not exactly computer enthusiasts. I was surprised to hear that many people seem to think Wikipedia is some kind of a search engine, just like Google. After all, you can go to Wikipedia, enter a subject you are interested in in a search box, click a button and if you are lucky, you will get a lot information about that subject printed on your screen. As far as these people are concerned, Google works in exactly the same way.

I have installed a second hard drive in my computer some time ago. Unfortunately it seems that the two discs in my computer spin with frequencies that are very close to each other. The result is that you can hear the interference as a low "brrrruum brrrrrum brrrrrum" sound that is coming from the computer.

Now this wonderful Chieftec case I have (it boasts as having a "Special spring design for anti-vibration.") has hard plastic feet that are wonderful at transferring this low frequency sound to the structure of our house. And the result of that is that "brrrrum" sound could be heard in almost all our rooms (and I'm sure the neighbors also get their share of it).

So I bought a piece of a thick carpet and made some soft pads to isolate the case from the floor and it seems to be working.

Who knows, maybe I'll even get the extra benefit that my MP3s will "sound more immediate and crisp" :)

The Debian Quiz has this to say about me:

You got 20 of 37 (54.1%) right.
Your Debian knowledge is good, but you lack some deeper insights.

I guess I have been overestimating my knowledge of Debian :)

I can't belive that a company could do something like this.

Just played a game of ATCA with Loopus while listening to the recording of the Humppa radio from What The Hack.

We experimented with various options - mostly on the OHare map which we both know well. The default setting of the number of planes is too low for two players, but with the setting increased by one step it is again quite a challenge. The Killer map is still unplayable, even with two players.

The greatest problem seems to be the coordination between players. Some kind of airplane marking needs to be implemented. This way players won't be confused which one is focusing on which airplane (right now airplanes crash when by chance we are both focusing on one group of airplanes while ignoring some other group).

Also SNL seems to have this problem when some times it will say that it can't open a socket if I start the server immediately after the game ends. I have to wait a few minutes and the problem mysteriously solves itself.

If there is a properly configured IEEE 802.11a access point in the forest and there is nobody there to use it, can you still say that the wireless network works?

-- Loopus at What The Hack

After some more thought, I don't think you can compare middle ages before the invention of the printing press with the age of computer networks like Mr Stallman did. People that were writing books at that time were either already rich enough so that they didn't care if they get any money for their work or were monks in monasteries who didn't need money. It is no surprise that no one cared about copyright.

I've listened to Richard Stallman's talk about copyright and community today. He tried to apply ideas from his free software movement to other kinds of works like literature, movies, music and physical objects.

He rejected the idea that free software ideas could be applied to hardware because you don't have a compiler or a copy machine that will cheaply manufacture a piece of hardware for you from a blueprint. I don't agree with that. If he claims that everyone has the right to learn from software by looking at the source an changing it, then I don't see why it would be OK for companies to hide the physical design of a device. 20 years ago it was quite common for companies to include full schematics with an electronic device you bought, but now the insides of some of the chips are the biggest company secrets. You won't find a television set today that will come with any sort of schematics or connection diagrams. Stallman said that physical objects are already as free as they can be, but many companies are deliberately making their electronic products so that they can not be tampered with (like embedding circuits in plastics so that you can not access them without destroying the device).

His interpretation about how the age of computer networks reestablished the situation that existed before the invention of the printing press was quite interesting. Before the invention of the printing press the ability of making a copy of a piece of written work was uniformly distributed between people. Everybody could make a copy and making 10 copies was about 10 times more difficult than creating one copy. With the printing press only publishers were able to make a large number of copies because they owned the expensive equipment. This uneven distribution of ability to make copies made copyright law necessary (or at least not harmful). Computers again enabled anyone to make large or small numbers of copies and Stallman believes that this makes conventional copyright harmful.

As expected he said that "digital rights management" devices are evil and that everyone should stay away from them. It seems interesting that he even acknowledges the existence of DRM (DRM is mentioned in the GNU Free Documentation License, so that definitely means that FSF acknowledges its existence). There are numerous examples that show that it is almost impossible to clearly define what exactly is a way of enforcing copyright (or "digital rights"). DMCA law in US is often criticized because of this, because by definition a lot of everyday items can be classified (like black markers for example) as devices that circumvent copyright protection schemes.

I also didn't like his idea that the author has absolutely no right to charge anything for his or her work. He said that if nobody asked the author to produce a piece of literature then he has no right to request money from people to read it. As far as I don't approve lengthening of the copyright that is happening today and huge amount of money publishers are getting I find this statement nonsensical. Someone manufacturing a physical object in advance and then selling it has in my opinion the same right to get some money for his work as a musician or a writer. I'm not saying that they should both use the same ways of getting money though. I also find his support of taxing empty CDRs as a way of financing musicians a bit weird. There are huge practical problems with that (like who will gather all the money from all around the world and who will then decide what part of this money an artist would get?) not to mention that the bulk of CDRs are used for everything else except music.

I was also surprised at the way he answered questions at the end of his talk. I would expect someone who is defending the right of free speech at least let people finish their questions. Most of the time he interrupted them in the middle of their questions and said that this or that assumption is wrong or that they are pronouncing their words wrong or that they aren't speaking clearly and so on.

This article about remotely controlled humans looks quite scary at the first glance (direct link to video). On the other hand, application of this to flight simulators would be quite cool.

Once you read the article it seems that technology behind this effect is quite simple (just stimulate some part of the skin with electric current). It looks like we won't be able to do anything really interesting with the human nervous system for a long time. The current methods remind me of hacking a microprocessor with a hammer - if you punch it on just the right spot, you might knock out a specific part of it, but you really don't know anything about what happens inside.

This morning I attended the keysigning party at Wikimania. I organized a similar event last year at Haip (I wanted to follow the keysigning party HOW-TO), but it failed miserably then because nobody had the patience and/or time to check all the passports and identity cards and then compare endless strings of public key fingerprints. So everyone just went like "OK, I trust all you guys without checking you IDs and fingerprints, now let's sign these keys and go on with the party". In the end nobody remembered to submit the signed keys to the keyserver anyway, so it was just a big waste of time and nerves. So I really wanted to see how a keysigning party organized by some one else would look like.

Hanno 'Rince' Wagner, who led the GPG workshop and the keysigning party didn't exactly follow the HOW-TO. What he did is that he gathered key IDs and fingerprints from everyone and compiled a list on the wiki. Everyone showed his or her passport or ID card to him and everybody else so we could confirm their identities.

This way of doing thing greatly simplifies the keysigning, because everyone doesn't need to check everyone else's fingerprints (the algorithm is O(n) instead of O(n^2)). However in this case you must trust whoever is compiling the list that he won't change the key fingerprints on it, so it isn't a technically correct way of doing it.

One bottle of drinkable water (1 liter) costs 1.80€ in Frankfurt. Quite a shock comparing to free drinkable water at What The Hack.

Tcie and I just arrived at Wikimania. Another change of plans. I know I said I won't be going, but there was enough time to dry myself up at home and then go to Frankfurt by airplane. We checked in the Haus der Jugend (a huge, modern hostel) and paid the conference fee. We got a nice printed conference programme and a printed version of selected articles from Wikipedia on Frankfurt.

People at the Wikimania check-in had problems entering our proper names. Funny, since čžš characters are in the UTF-8 character set and MediaWiki uses UTF-8 by default. I'm used to spell my name in plain ASCII anyway and they promised to make our conference name tags with proper characters.

They have a wireless network set up here. It is working right now (so I hope I will be able to post this), but it already seems that the number of users is approaching the critical mass (people at the counter said that there are currently around 160 attendees, each one using her or his own laptop I guess).

I'll be posting short descriptions and comments on the talks I will be attending at the Cyberpipe web site, since tcie and I are after all representing Cyberpipe at this conference.

Here are some instructions how to get from Ljubljana to Klagenfurt airport for any other Ryanair customers from Slovenia (it may look simple, but it is far from obvious once you are actually there)

V Ljubljani kupiš vozovnico do Klagenfurt Hbf. Direktne linije ni. Prestopiti moraš v Beljaku - Villach Hbf. Vožnja z vlakom traja okoli dve uri. Na glavni železniški postaji lahko nato na avtomatu kupiš vozovnico za primestno železnico do Klagenfurt Annabichl (karta stane 1.60€). Vlak, ki pelje do tja, ima končno postajo v Wien Südbahnhof ali Bruck a.d. Mur. Klagenfurt Annabichl je druga postaja po vrsti (prva je Klagenfurt Ost) in je v bistvu samo peron, ki stoji ob progi (nekaj takega kot Ljubljana Tivoli). S perona greš po stopnicah navzdol do ceste, ki gre pod progo. Tam zaviješ desno in slediš cesti do letališča.

I traveled from München to Ljubljana with a Slovenian Railways night train. So I got on the train and tried not to be upset because I will have to share my seat with my luggage for the next six hours (luggage compartments were full, go figure). As always I waited for the conductor to first check my tickets and reservations and then put on my headphones and went to sleep.

Next thing I remember is five faces staring at me. My first though was that I did something stupid while asleep (wouldn't be the first time), then I noticed that the owner of one of the faces was the conductor and then I noticed that his face is all red and that he is shouting something. So I removed my headphones and slowly began decoding his German language. Obviously he wanted to see my tickets again and was furious because I didn't show them immediately. I tried to explain to him that some people occasionally experience a period of inactivity we call sleep, but he didn't seem interested. After he checked the tickets one more time he was satisfied simply left. When the conductor returned for the third time, I was already awake because of the passport control.

I traveled by day for longer through Netherlands and Germany, but I never had to show my ticket more than once (and ICE tickets are quite a bit more expensive than this night train ticket). Why on earth do they want to see your ticket three times on a night train when people probably want to (gasp!) sleep? Add two more sleep interruptions on the border to show the passport (once for the Austrian and once for the Slovenian border police. We are in EU, right?) and I really don't see why I would like to spend extra 20-something Euros for a bed, when I couldn't get any sleep anyway?

1 August 2005, 14:52

I left the camp and went by train to Amsterdam today. The official part ended yesterday around 18:00 with the closing talk. After that we began clearing the camp area. I helped to pick up the garbage and carry chairs and equipment from one of the larger tents. The net and What The Bar were still operational yesterday evening, but when I woke up today all of the big tents were practically cleared up and I could only get a breakfast and lukewarm tea.

Loopus and I planned to go to the Wikimania, but wet clothes and equipment, bad weather and a promise of a hot shower (in two out of three cases the water in the showers at WTH was freezing) changed my mind and I decided to end my journey a bit early. Loopus stayed behind (has went to sleep about the time I got up) and is planning to camp for a few days more in some camp near Amsterdam.

I'm currently on the train to Duisburg and I don't have net access, so I'll post this as soon as I get home.

20:25

I'm just past Stutgart and thinking about WTH

Last year when I was on 21C3, I tried to attend as many talks as possible - almost every hour there was a talk or two that was simply to interesting for me no to attend it. But when the conference was over I realized that I hadn't actually had any longer conversations with other visitors of the conference. I decided that when I'm on the next such event, I'll spend most of my time socializing and only attend a few of the most interesting talks and workshops.

Now that I can look back at my visit to WTH, I see that I made the right decision. I learned a lot just by walking around, talking to people about their projects, sometimes helping them with a specific problem and most importantly, it was much more fun than just listening to the talks. I was really surprised how friendly everybody was. If I asked something I always got a friendly answer. I got the nice feeling how everyone here is really part of a great community where each member is really doing her or his best to contribute.

I got into a conversation with a guy sitting next to me. We talked about this and that and then I mentioned I'm just returning from a "computer related" camp. The conversation went fine until I mentioned the term "hacker conference". At that moment he noticeably became uncomfortable and asked if I was talking about people that break into FBI databases and similar stuff. And then I spent the rest of our journey explaining what I understand under the term hacker. My little contribution to undoing what the mass media has done.

One thing that did bother me in the camp is that I really didn't sleep much these four days. We set up our tent in the area A1 right next to What The Bar and Foo Bar, which turned out to be a bad idea. Ear plugs don't help much when you can feel the drums in your stomach. And if we would move our tent to other areas, we would probably drown instead of just getting wet in that storm.

2 August 2005, 18:45

There's no place like $HOME :)

It began raining again over the night. Fortunately the ground is absorbing water fast enough this time and the tent is dry for now. I also dried all of my stuff yesterday on the sun, so I have very little motivation to get out of the sleeping bag or our tent.

I've attended a talk about exploiting PocketPC. Collin Mulliner showed us what an insecure platform Windows CE and all of its variants are. Because of certain features of this operating system (static addresses of the functions in libraries that are the same across all devices of the same type, even user processes can directly access hardware, etc.) writing exploits is quite easy. He showed a few examples ranging from displaying a message on the screen to making a call on a Windows CE based smartphone.

I've also been at the GNU radio workshop. After discussing basics of software defined radio, they showed us a basic FM radio receiver (they have basic building blocks like decimators, demodulators, etc. written in C for speed. These blocks are then connected with Python). Demodulation of the stereo signal used 25% of a 400 MHz processor if I remember correctly. A more interesting example was a transmitter that continuously transmitted words "All your frequencies are belong to us" on all CB channels. So no matter on which channel you tuned your receiver on your heard that voice.

Hacking on a traffic shaper configuration for the Meshnode people was also quite fun. I've set up a basic shaper, but the the configuration they would like to have is well beyond my knowledge.

Yesterday I also had my first experience handling a professional video camera. Since the video crew is running out of people I volunteered to record a talk by Indymedia people. I hope the recording is satisfactory.

Anyway, the official part of the camp is ending today. We plan to stay another night until Monday, perhaps helping to clean up the camp site. After that we're going to spend a few days in Amsterdam and then head back to Frankfurt to Wikimania.

Here are some assorted photographs of interesting things I found around the camp.

It seems some people we ignoring warnings about different host keys when logging into their machines through the network. Yesterday, there was something like this written across the big display in What The Bar: "If you find your password here, change it quickly" followed by a list of passwords.

Unfortunately, a lot of POP3 and IMAP mail servers still don't support any kind of secure authentication (our beloved Siol included)

Obviously, they buried Bill Gates here back in 8 August 1997 (that is the date of Hacking In Progress, one of the previous camps)

These people have planted some kind of a grass made out of fibreoptic cables. The sight is really impressive in the dark when each strand of grass emits blue light. Unfortunately, my camera can't take that kind of pictures.

Some people have brought their kids with them. There are a lot of workshops organized for them (I wouldn't mind joining some of the workshops myself - water rockets are fun)

This the place I spend most of my time. Meshnode, ASCII, GNU Radio people are all gathered here, building bamboo antenna towers, broadcasting various stuff (for example: "All your frequencies are belong to us" simultaneously on all CB channels) and generally having a lot of fun. They are organizing a big party tonight that will include dancing for rain (like we don't have enough of it).

This is the rocket of the Chaos Computer Club. They have set up a system of running landing lights that look like the beginning of a runway. I'm just waiting when one of those AWACS will try to land here.

Following two innovative transportation devices appeared near the What The Bar today. I'm not exactly sure if the exoskeleton is functional (it's built out of stuff commonly found in kitchens like utensils, part of washing machines, etc.) but the quadruped bike has certainly seen some use because someone broke of a pedal.

Some parts of the camp are connected to the net through equipment conveniently stored inside toilets. First time I saw that sign about dropping cables in the toilet, I thought someone was joking. Network admins said that toilets were the cheapest weather-proof cabinets for network equipment they could get.

Remember what I said what a nice place we found for our tent? Well, I take back everything I said. It was raining hard all night and I just woke up in water. Most of it is fortunately still below the bottom of the tent but speaking from experience it won't stay there for long. It looks like the whole field is under a few cm of water. My backback with laptop and other electronics is already soaking wet and my sleeping bag is also getting wet. My bed now has a squishy feeling and if I drop something on the floor in one corner of the tent I can see ripples in the water below going right to the other corner.

Perhaps taking that emergency shelter wouldn't be such a bad idea.

An AWACS-like airplane (big circular radar antenna on the back) just flew quite low over the camp. If the army is warflying I'm afraid they will be disappointed because the wireless network (at least IEEE 802.11b/g) continues to be in a very bad shape :) Unfortunately I was t