Nothing bad has happened to me because of this, and one may argue that targeted ads are better than non-targeted. But I do not like the idea of being tracked – and I shut off all web ads anyway with AdBlock. Additionally, I have NoScript always on (and allow sites selectively each time when “some site does not work”).

Today I have been told that there is a way to track me even without JavaScript and tracking images from spyhouse sites. It is demonstrated here:

https://panopticlick.eff.org/

I have quite long “accept-language” header set in my browsers, as I can read web pages in several languages. panopticlick shows that I’m one such user out of about half million (it might be that I’m just the only one with this value of accept-language, who made a check there).

I’m not feeling paranoid because of this. But I’ll be happy to know, is there a way to pass my (complicated ) language preferences without allowing for easy fingerprinting.

But what if the ice breaks and you fall in the zero degrees cold water?

There is a simple answer: try this before it hits. Your chances to get out alive become higher. Read on for the boring theory and a short report of how I did this recently in a mild -17°C afterwork evening.

The first thing to really understand is that this can happen. If you ever move on the ice, this can happen to you. I learned it from my own experience.

So, the second thing is to be prepared. In this case, being prepared is relatively easy.

There is a general “smart” advice: do not panic, know the theory, act calm and fast. Anyone who had ever been in a stress situation knows that the latter can’t be just easily followed. When the life is at stake, all your skills can degrade in one moment down to what is learned in practice. So – just do it.

• Find an ice opening, where you can safely get out. In Finland, we have maintained winter swimming spots which are perfect for this task
• Call a friend or two who will pull out your body on a rope in case something gets really wrong.
• Make sure they actually can do it – test this on the ground first; one should be reasonably fit in order to be able to lift a person of similar weight.
• Prepare full change of the clothes, including shoes. (Never attempt such fun disrobed. You can cut yourself with ice edges.)
• Read more theory below.

It’s nice to practice next to a warm place – a house or a car, which you can jump into after the exercise. But in fact I recommend doing it without such possibility. Getting out of the water is just first stage of your game; second, maybe equally dangerous and sometimes underestimated stage is recovering your warmth afterward. Climbed out, one may fall in an euphoric feeling “I did it” and end up with frostbitten extremities. (For ones looking for real horror pictures of frostbitten hands, see bottom of this page. And the story started innocently: small dipping in water in the cold.) You’ll know fascinating details: wet shoelaces freeze and you can’t take the shoes off; zippers freeze and do not open; and so on. Reading about such issues from a computer screen does not always deliver the idea that this is serious.

One more “do-and-do-not” list:

• always carry awls with which to climb out of the opening back to the ice (pictured above; jäänaskalit in Finnish; I wonder how is it called in English or Russian). In some cases, it may be impossible to climb out, as the ice may be slippery and your hands just slide helplessly. They must be hanging on the neck, not in a backpack or a pocket.
• always carry some spare clothes. Full set is the best; at least something is a must. It must be packed so that it stays dry after your potential bathing
• climb out to the direction you came from
• when got out, immediately do whatever you can to the wet clothes (in 15 minutes, it may be already too late) – change it if you have change, wriggle it as good as you can otherwise
• never take alcohol or anything like that.

So, how was it in practice? Pretty cool. Real winter fun (with necessary precautions). Unfortunately, the water level in our test lake was lower than I expected – just up to shoulders, so I could actually stay and push myself up. This made climbing out easier than it might have been. I climbed first time without pulling the awls out; but I guess the possibility to jump on the bottom makes a huge difference. After that, I decided to still test the awls and crashed ice edge the second time. Naturally, it was even easier.

Now I would like to repeat the exercise in a place where I do not reach the bottom with feet. And of course getting a bit nicer pictures does not harm. It may be not immediately clear without explanation, that the picture to the left shows a floating object in the ice opening, and that this object is in fact me giving a short lecture on how it feels.

• “Epson tests and certifies most common Linux and Unix platforms with our printers” (quote from here)
• “The HP Linux Imaging and Printing project provides printing support for 1,949 printer models” (quote from here)

Finally, my choice has fallen on HP c8180. At this moment, openprinting.org, the source for Linux printer support information, does not list this model (it’s quite new). But having configured it as “HP C8100 series” in Ubuntu 9.10, both printing and scanning work just fine right out of the box without any additional installations or configurations.

Other riders carry water bottles in their back pockets, but my jacket has pockets which are not really suitable for it. So I decided to try a camelbak. But models in the store make my digestive tract to work in the wrong direction. Such simple thing as bag for the plastic water bladder is made of tens of various colorful cloth pieces, contains plastic foam padding everywhere, several strips and bindings all over the body, and weighs up to 500g! I’m not much of a weightweenie, but half-kilo just does not make any sense here. All this plastic foam and nylon base becomes completely wet during the ride, I’d have to carry half kilogram more of own sweat each time, and wash it away at home. And for these rides, grams actually matter: for me, it easily can mean keeping up with the group until the end or dropping out somewhere in the middle.

I realized that I’m not taking any of these fancy ready camelbaks even for free. I bought just the plastic water bladder (it was marked “Platypus”, weighted 120 g.) and sew quickly an ultra-thin nylon cover for it together with two thin, minimally padded shoulder straps. The latter attach directly to the top closing of the bladder. As I said, all my creature weighs 20 grams and does not collect more sweat than the inner lining of the jacket. In fact even much less.

Last weekend this masterpiece was successfully tested. I didn’t drop out of the ride. Well, the pace was a bit more peaceful than it had been some times before. Here is the map.

View in a larger map

It worked! It takes 2l of fluid and weighs less than 2 x 1l. standard water bottles with holders. I’m taking it next time.

See you on the ride!

• There are two planet carrierrs. Call “first” the one which is closer to the sprocket.
• Both rings have two suns, two gears on the planets, and only one ring gear. The ring gear meshes with the smaller gear of the planets.
• In lower gears 1 and 2, power input applies to the ring of the first planet carrier. This carrier lowers down the transmission ratio, depending on which sun is locked to the axle by a pawl. The second carrier is not involved.
• In higher gears 6 and 7, the first carrier is not involved. Power is applied to the planet ring of the second carrier. Depending on which sun is locked to the axle by a pawl, two step-up ratios are obtained.
• In middle gears 3, 4 and 5, both planet carriers are involved.

The table explains which exactly rings are engaged, and gives tooth numbers:

Note that in the step-up ring (Ring2), 30+19+14=63, while the ring has 66 teeth. Also, 36+14+14=64. I saw one tooth differences between sums of sun and planets and the ring, but at the moment I do not understand how so big difference (3 teeth) can make working combination.

I found the service instruction in German, explaining which planet rings are transmitting the torque in which gear. This, plus the transmission ratios with three decimal digit precision, plus some general theory proved to be sufficient. Update: No! The numbers below are wrong. You may think, how this can be seen from them. See the real counts from the disassembled hub for correct counts.

To understand the numbers below, you must know that Nexus-7 has two planet carriers; each has planets with two bound gears. One of them meshes with the ring gear. Each of planet gears meshes with corresponding sun gear. The latter can be rotating freely, or fixed to the immobile axle with pawls, activated by the shifter. In gears 1 and 2, only the first step-down ring is engaged. In gears 6 and 7, only the second step-up gear is engaged. Read the service instruction linked above for further information.

I give values for the active sun, planet meshed with the sun, planet meshed with the ring, and the ring. Next, the exact fractional ratio is shown, and it’s decimal value with 5 digits. Last in the line is the 3-digit decimal value from the manuals. We see the rounding errors (and sometimes inconsistencies).

First ring, step-down:
Gear 1: 31, 17, 22, 69. Ratio: 69 * 17 / ( 69 * 17 + 31 * 22) = 1173 / 1855 ~ 0.63234 ~ 0.632
Gear 2: 24, 22, 22, 69. Ratio: 69 / (24 + 69) = 23 / 31 ~ 0.74193 ~ 0.741

Both rings engaged:
Gear 3: G1*G6. Ratio: (1173 / 1855) * (617 / 462) = (17 * 23 * 617) / (2 * 5 * 7^2 * 11 * 53) = 241247 / 285670 ~ 0.84449 ~ 0.843
Gear 4: G2 * G6. Ratio: (23 / 31) * (617 / 462) = (23 * 617) / (31 * 2 * 3 * 7 * 11) ~ 0.99085 ~ 0.991
Gear 5: G2 * G7. Ratio: (17 / 11) * (23 / 31) = 391 / 341 ~ 1.14663 ~ 1.145

Second ring, step-up:
Gear 6: 31, 21, 15, 66. Ratio: 1 + (31 * 15) / (21 * 66) = 617 / 462 ~1.33549 ~1.335
Gear 7: 36, 15, 15, 66. Ratio: 1 + 36/66 = 17 / 11 ~1.54545 ~1.545

The inspiration for this feat was John Allen’s table of tooth counts for various hubs. John did this by some guessing-and-trial-and-error, putting numbers in a spreadsheet with the formulas. Unfortunately his data does not include the two hubs I have, Nexus-7 and Nexus-8. I decided to leave the guessing to the computer and wrote a script, which does exhaustive search over all reasonable combinations. There are two restrictions on numbers of teeth in a working planetary gear (diameters should add up in a known way, and some counts must be divisible by number of the planets for them to be equally spaced; ask me if you want to get techy). Allowing the precision of the resulting ratio down to the last decimal number given in the documentation, the search returned just one result for both planet rings of the Nexus! I was so delighted to get such a consistent answer.

Update once again: this “consistent answer” is wrong. Here is the right one.

1 (I knew this earlier) “Dried cucumbers”:
Cucumber contains 99% of water. A box of cucumbers weighs 100kg. During winter storage, cucumbers dried out so that they now contain only 98% of water. How much does the box weigh now?

2 (I got this today, which inspired me for this post) “Probability for doctors”:
Assume a test for a rare disease (which affects 1 person in a million) is 99.99% accurate. A patient tests positive. What are the probabilities he actually has the disease?

So I made my own from a 2mm thick rubber sheet. It is heavy and non-aerodynamic. It is heavy-duty! It makes difference when riding in the rain. It is tested. Highly recommended for bike commuters.

First of all, what I see when trying to access the site, is this great information:
They kindly inform me that it’s “not yet available on UNIX”. And that I should use Firefox 3. Which I am, in fact, using now. But Ovi does not work anyway. Google maps have been working for my platform with Firefox already for years.

The first dumbest thing to do is to use a user agent switcher add-on, which just tells to the server “I’m Internet Explorer”, and see if that works. What a surprise, it does. So, these Ovi maps just check what my user agent is saying and refusing to serve other agents without visible reason. Disgusting. OK, I still continue.

The main screen has… about 1 cm margins on all sides. What the hell. Screen space is a valuable resource, and everybody wants the map to be larger. In fact that’s one reason why I still buy paper maps: there is no screen (and certainly no portable screen) which can show as large area in as much detail, as paper map does. Now, they just made my screen 2 cm less in each dimension. I’m not sure I want to continue. Google does not do that – I come for maps, and I get maps. Not white margins.

My message to Ovi managers: you are not only several years behind in time. You are also way behind in the implementation. I don’t see how you are going to catch up.

And, sure, I’m not working on either of the companies I compare.

What to expect there? A sporty ride for 2-3 hours; this time we rode 87km in 2:40h, which makes 32.8km/h average.

View in a larger map

Almost 33km/h average speed may sound too sporty for many. But it’s a group ride, and the power for 33km/h in draught is comparable to maybe 27.5-28km/h average alone. Also, the group motivates! I’m unlikely to pedal the same distance even at 27.5km/h average because of my lack of motivation – I need a reason to push myself. When there are 15 persons hanging around, I need a reason to drop out. Sometimes there is one (too fast! ), but now I holded until the end. Of course there is absolutely no problem if anyone wants to stop or continue own way – just inform somebody in the head so that others do not think that you have some technical problem and do not stop to wait.

It was sunny, dry, and a bit over +20°C. We do not have that good cycling weather too often in Finland, so it must be used effectively!