2010

Thu, Feb 25

<12:57 EEST> Linux 2.6.33 was released yesterday, but the lid status problem of 2.6.32 remains. The old patch should work in principle, but someone had switched a logical statement from A && B to B && A, so it takes some hand tweaking >.< However, I assumed the solution would have evolved along with other kernel developments, and I found an i855 specific solution that had not made it into 2.6.33. It still has the 'old' ordering of the logical statement, but after that simple fix the patch applies. Haven't booted into the new kernel yet, as I am currently at the university; the remote boot itself should not be a problem, but the screen status is another matter :-j

Yesterday I also got a new 2.5'' 320-GB SATA hard drive, destined to make Hoo a little faster and more power efficient. I decided not to transfer the OS quite yet, as I am also expecting another piece of Ebayed hardware, a 64-bit processor.

Sun, Feb 21

<13:24 EEST> On Friday, after a somewhat quirky exam in Physical Chemistry, I was browsing Ebay for some interesting and quality hardware. This meaning not x86 or x86-64, more specifically ARM, PowerPC and MIPS, and with the latter keyword I came across a couple of deskside supercomputers by Sicortex. Whisper-quiet 72-CPU clusters drawing 300 watts were being sold for around $2000. I did not recall hearing anything from a company producing such interesting technology, only three instances of which were available on Ebay that day.

Later browing Slashdot, I searched the site for Sicortex; I almost never use the /. search, and I had not even searched the company anywhere else, but it felt like the appropriate place in this case. An article from last November came up, along with news of demise. Sad, but not surprising, in this world where technological superiority is no match for marketing and industrial inertia. I felt the urge to get my hands on one of these machines, if only to increase the amount of Quality around me, as it had been drained away by the evils of x86.

It is noteworthy that this was not just some geek fantasy of technology that nobody wanted to buy. Sales were booming, but the start-up was not quite yet standing on its own, and during the financial crisis some venture capital was withdrawn. So this was a rather more striking case of good technology losing out to market forces.

Not surprisingly, most supercomputers today use x86-64; not because it is particularly efficient, but because it is available and somewhat cheap in the short run (somebody has to pay the power bill after all). Admittedly, the efficiency of x86[-64] today is not half bad, because of the huge amount of development resources poured into it. I like to think the reason behind this is the market share of Windows and its applications, which are hopelessly locked into this architecture.

But supercomputers do not generally run Windows. Unfortunately, they often run closed-source applications. Having used some of these applications, I can tell that the vendors are generally more flexible and sensible than those of your typical Windows application. They usually provide optimized binaries for many different variants of x86[-64], but apparently the whole point of portable software is not quite there yet.

I have often thought, one way in which open source can speed up technological development, is by allowing the use of new and better hardware architectures with minimal effort. With today's ARMs race in netbooks and smaller portable devices, this should be even more apparent. Windows users can stay with their trusted old x86s draining batteries, while the open world whizzes by into a more efficient and environmentally friendly future. Or so I can wish.

The key phrase "trusted old" brings to mind the word "conservative". One of my pet problems with the world is that "conservative" in the political and economical senses has nothing to do with actually "conserving" resources like energy and the environment. It is only about conserving practices, doing things the way we have always done, no matter how stupid and inefficient they may be today. The detail here being that some practices were great in the 1800s, for example, but they are not necessarily so great today.

Sun, Feb 14

<22:34 EEST> The coincidence of St. Valentine's and the Chinese New Year's Day may be somewhat notable already, but in Finland today is also the winter festival of laskiainen. The root of the word is "laskea", which in this context means going downhill, especially by some sliding means down a snowy hill. However, it also means to count or to calculate. Thus the triplet of events also marks my newly found academic interests of number theory and Chinese language.

The idea of doing minors¹ at all stemmed from the fact that my subject studies in chemistry are nearing completion, at 54 of 60 credits. The two remaining courses are spread out over the spring term, meaning either plenty of extracurricular activities, or something within the university. I figured I might as well exploit my precious time in the free education system, and take up some of my long-time interests with academic rigour.

Number theory is something I dabbled with in 2007 while exercising my then-idle brain with Project Euler. Actually, I regard it as a rather central part of mathematics that touches many other fields, and I have even covered some of its aspects while teaching at Voionmaa. As a field in its own right, it has the curious character of having easily understood problems that may require hard-core solutions. For example, the Pythagorean theorem is a nice little investigation when you consider integer solutions only, but its extension into higher powers turns out quite elusive.

As a physical scientist, I already know my way around plenty of mathematics with a mix of intuition and exercise routine. However, a lot of this intuition comes from the physical reality. For example integration, to me, is a physical process, and the math is just a convenient way of writing it down. However, pure mathematics has definitions for integration that do not depend on the fleeting existence of a universe, as perceived by mere mortals.

This discrepancy between views is probably one reason why I am still humbled and enchanted by the miracles of mathematics, after all these years of work and study in nearby fields. Of course, number theory is probably one of the most obvious branches in which to feel this way. The experience reminds me, in many ways, of my years in the IB where I discovered my talent and passion for mathematical thinking. Seeing how my intuition has so far played out, in the somewhat foreign territory of pure mathematics, not all of it has been lost in the sea of applied and approximate. Quite the contrary.

In some ways, I think Chinese language is an excellent complement to hard-core pure mathematics. It also happens to be one of the exotic choices I wanted to study as my first foreign language, but back in a small town of 1987 it was obviously not available. Today the language has huge global importance, and schoolchildren are likely better off in this respect.

Like number theory, this Chinese course is mostly about doing something cerebral in my non-chemical time, and in reconfiguring the brain it has already succeeded to a great extent. After studying and working for years in mathematical and physical sciences, this is a refreshing breeze of doing something completely different. Frankly, a lot of my subject studies in chemistry have felt like revision, not that there's anything wrong with that.

The writing system is just one of the new and exciting aspects, the use of tones is another. Occasionally, written and spoken Chinese feel like two different languages, and even English starts to look completely phonetic in comparison. Of course, we also use pinyin extensively, so it is like two written languages for one spoken. Not counting simplified vs. traditional Chinese and other character systems :)

[1] Pun intended. However, this does not mean that I fully appreciate acting out the secondary meaning of this double entendre. It means that, while writing about the original meaning, I noticed the secondary meaning, and I was not afraid to leave it there, as I appreciate a good pun. Nevertheless, many people would note "no pun intended" in such cases. So they notice the potential pun, do not want it interpreted thus, yet they choose not correct it using some alternative wording. I like to think that everything I write is intended, especially if I write something clever enough to have non-trivial meanings.

Tue, Feb 9

<12:41 EEST> Eikö olekin ärsyttävää, kun joku hehkuttaa keksineensä uuden sanan, jota sinä olet käyttänyt jo iät ja ajat? Jokin aika sitten googlasin lumen satamista merkitsevää verbiä, jonka yksikön kolmannen persoonan indikatiivin preesens on "lumisee". Sana on pyörinyt ainakin minun ja äitini keskusteluissa vuosikausia, enkä muista mistä sen alunperin nappasimme, mutta näemmä Lotta Backlund on keksinyt sen uudelleen.

Levittääkseni tällaista keksintöärtymystä laajemmalle ajattelin kertoa sanasta "yhdyssanaliitto", jota on tänään käytetty eräällä IRC-kanavalla. Sana tarkoittaa kahden sellaisen yhdyssanan ristisiitosta, jossa ensimmäisen yhdyssanan loppuosa on toisen alkuosa. Esimerkiksi lapsivesi + vesipuisto = lapsivesipuisto on monelle tuttu esimerkki, joka ilmentää tällaisissa sanaleikeissä usein haettua häiritsevää, ristiriitaista efektiä.

En siis väitä ilmiötä uudeksi, mutta en ole tähän mennessä kuullut sille nasevaa termiä. "Yhdyssanaliitto" on sikälikin toimiva nimi, että se on itse yhdyssanaliitto. (Vastaavalla metatasollahan pelaa esimerkiksi "sananmuunnos".) Sanaliiton kanssa termillä ei ole varsinaisesti mitään tekemistä, mutta liitto se on kahden yhdyssanankin liitto, tai sosialististen neuvostotasavaltapelilautamiesten. Idean ketjuttaminen tällä tavoin on siis enemmän kuin suotavaa, kenties nimeäkin voisi tällöin laajentaa vaikkapa yhdyssanaliittotasavallaksi. Verbaaliharrastuksille ei tosin liikene kaikkea aikaa, kun menossa on esimerkiksi matematiikan kurssi jolla vaaditaan hyvää alkulukupäätä.

---

Risto A. Paju