Perl is nifty. I used it to write a little converter script to let me import
bank statements from Bank Austria [DE]‘s web
interface (available only in CSV or the obscure href="http://xml.coverpages.org/ifx.html">IFX/XML [EN] format) in href="http://www.gnucash.org/">Gnucash [EN]. It was even quite easy to add a
secure programmable auto-categorizer that lets me write the rules for assigning
transactions to various accounts in restricted Perl in a configuration file.

Then, I got into a problem when dealing with split transactions. Whenever I
added split transactions to my QIF file, the final QIF generation would fail. I
suspected my data structures first, or some weird interaction from the Safe
module that executes the categorization code. Debugging it was impossible since
the code that ran fine when executed normally would fail when executing the
categorization code:

themel@sophokles:~/ba-ca$ perl ba-ca.pl -f examples/bank-account.pl < split.csv > split.qif
Loading categorization code from examples/bank-account.pl...
Not an ARRAY reference at /usr/local/share/perl/5.8.4/Finance/QIF.pm line 296, <> line 2.
themel@sophokles:~/ba-ca$ perl -d  ba-ca.pl -f examples/bank-account.pl < split.csv > split.qif 

Loading DB routines from perl5db.pl version 1.25
Editor support available.

Enter h or `h h' for help, or `man perldebug' for more help.

main::(ba-ca.pl:34):    my $qif ;
  DB<1> r
Loading categorization code from examples/bank-account.pl...
Categorisation code failed: Undefined subroutine &Finance::QIF::Transaction::add_to_splits called at /usr/share/perl/5.8/perl5db.pl line 3636, <> line 2.

 for transaction:
$VAR1 = bless( {
                 'amount' => '323.40',
                 'date' => '27.10.2004',
                 'memo' => 'Gutschrift a/BM F.INNERES            DVR:0000051          ZIV-Zahlung 11/04                                           ZIV-Zahlung 11/04                                           (222179 THEMEL Thomas)                                  Fahrtkostenersatz: 41,70                                  Wohnkostenbeihilfe: 281,70'
               }, 'Finance::QIF::Transaction' );
Debugged program terminated.  Use q to quit or R to restart,
  use O inhibit_exit to avoid stopping after program termination,
  h q, h R or h O to get additional info.
  DB<1> q
themel@sophokles:~/ba-ca$

I have no idea what causes this, but it meant I was back to printf-style
debugging. Using this, everything seemed normal – Data::Dumper showed that the
split member of the transaction actually was an array reference:

$VAR1 = bless( {
                 'splits' => [
                               {
                                 'amount' => '41.7',
                                 'memo' => '',
                                 'category' => 'Travel Cost'
                               },
                               {
                                 'amount' => '281.7',
                                 'memo' => '',
                                 'category' => 'Rent'
                               }
                             ],
                 'amount' => '323.40',
                 'date' => '27.10.2004',
                 'memo' => 'Gutschrift a/BM F.INNERES            DVR:0000051          ZIV-Zahlung 11/04                                           ZIV-Zahlung 11/04                                           (222179 THEMEL Thomas)                                  Fahrtkostenersatz: 41,70                                  Wohnkostenbeihilfe: 281,70'
               }, 'Finance::QIF::Transaction' );

Of course, after banging my head on the desk long enough and nearly
despairing, I took a closer look at the offending line of code:

        for (@$self->{splits}) {

And finally noticed that the problem wasn’t that $self->{splits} wasn’t
an array reference. Perl attempted to make $self into an array
reference (so that it then could attempt to dereference it using a hash
accessor? I have no idea what this is supposed to do). A tiny change to

        for (@{$self->{splits}}) {

fixed this, and now I can have split transactions, too... Another case were
Perl's illegible crap syntax cost me time and nerves - but who am I to rant, I'm
still too lazy to solve my problems in a proper language...

Anyway, if you're a fellow victim of BA-CA's web interface and non-support of
HBCI, have a look at the
result of my struggles.

Update 2006-08-03: Look for updated versions here.

I can’t make sense of href="http://www.google.com/search?q=georgewbush.com+access+denied">this
[EN] – what’s the point? Anyone who actually wants to see the damn thing can
simply use the excellent tor [EN]
network (many of whose exit points seem to be located in the US) or a number of
less high-tech alternatives to get at it.

In my post entitled href="http://home.themel.com/weblog/archives/permalinks/2004-10-24T17_27_12.html">"Being
Rational Is Depressing"[EN], I stated that I thought rationality’s
missing emotional appeal is the reason for a lot of the things that go wrong in
world politics. While catching up with href="http://www.cypherspace.org/adam/cp.html">cypherpunks [EN] today, I
found one of these gems that actually make wading through all the bullshit
flame wars (and reminiscing for The Real href="http://groups.google.com/groups?q=tim%20may">Tim May [EN]) worthwhile.

John Kelsey produced an excellent three paragraph summary of why neither
Islamic fundamentalists nor rabid terrorist hunters (unfortunately, the only two
kinds of people who seem to have grown in power, or at least mind-share) are
what we want:

What I discovered is that the "coming up with your own purpose"
thing is a bit hard (and thus avoided by many people, see success of above mass
movements) – but I still find it vastly superior to having someone else’s
purpose imposed upon me against my will or through lies, well-meaning or not.

Retailers have been pestering us for some years to join all kinds of reward
programs designed to track customer buying habits. Usually, what they offer in
return for my shopping anonymity is not of interest to me – exclusive offers of
stuff that would be EUR .40 more expensive when bought elsewhere or utopic
premium items (see: href="http://www.wiwo.de/pswiwo/fn/ww2/sfn/bm_artikel/bmpara/1012/bmpara/415254205374616e64617264/id/127/id/33000/bt/2/fm/0/artprint/1/SH/0/depot/0/">wiwo.de
- Deutschland wird zum Punktemekka [DE]

Payback und HappyDigits versuchen Emotionen dagegen mit hochwertigeren Prämien
zu schüren. Selbst wenn diese kaum zu erreichen sind. Wer das
Team-Telekom-Mountainbike bei HappyDigits ohne Zuzahlung haben will, muss dafür
beispielsweise rund 145 000 Euro bei T-Mobile vertelefonieren.

– muhaha). However, while doing some calculations about my personal finances
today, I thought of something that would convince me: Give me online access to
my transactions! I’ll happily let you track my purchases if it reduces work for
me while doing the same. Currently, I have to keep receipts and
manually type them into gnucash [EN], which
feels entirely prehistoric to me (and is the reason I have three-digit amounts
of “mystery money” every month, which disappear without me having any idea where
they went).

Wouldn’t it be great if there was a standard web service so that I could
simply tell gnucash to aggregate data from href="http://www.ba-ca.com/">ba-ca.com [DE], href="http://www.billa.at/">billa.at [DE], href="http://www.johnnys-pub.at">johnnys-pub.at [EN], href="http://home.themel.com/weblog/archives/permalinks/2004-08-12T13_29_41.html">amazon.*
[EN] and all the other sources? I can’t think of too many technical
difficulties at first, and it’s definitely one of those cases where I’d give up
privacy to obtain more information about my personal cash flows.

I’m not much of a sports person. This can probably be blamed on my father,
who taught me to read at an early age, after which I couldn’t really figure out
what the fun in running around and hurting yourself in the dirt was. Except for
some trouble with PE teachers in various stages of my education, I never quite
got to regret this, until I recently got my chance to help in some href="http://www.zivildienstverwaltung.at/informationen/informationen.html">government-organized
slavery [DE] and noticed that my physical condition was rather badly
adapted to eleven-hour shifts of physical work.

Since I felt that I needed some kind of activity to balance the pointless yet
exhausting work I do there [DE], I
took up swimming. One of the benefits of living in a city that has been href="http://www.wien.gv.at/ma08/bgmliste.htm">governed by social democrats
[DE] for most of the democratic parts of its history is that there’s a good
number of public baths [DE] in
excellent condition (if you don’t mind the occassional homeless person sleeping
in the warm shower) and with rather cheap access (EUR 132 for a whole year).

One of the first things I noticed is that swimming laps tends to get boring.
I can only concentrate on executing exact strokes or counting strokes per lap
for so long before it loses its appeal. This is not in itself a new situation -
on most days, I spend an hour on public transit, or waiting for something.
Usually, I have two things available to mitigate this – my href="http://derstandard.at/">newspaper [DE] and my href="http://www.apple.com/ipod/">MP3 player [EN].

At first, both of these seemed out of the question, but while I haven’t yet
found a good solution for reading large-format newspapers with both of my hands
busy and my body mostly submerged, there’s a company in Florida that sells
waterproof bags
[EN]and headphones
[EN]. While there are numerous solutions that claim water resistance, this was
the only that claimed to be usable for swimming and even submerged operation.
The price tag let me hesitate a bit, but after a few days, nagging curiosity
beat out fiscal conservativism and I ordered it. It took some time for the
package to arrive since God was once again punishing
Florida [EN] at the time, and there was another EUR 28 added to the bill by
those bastards at the customs
office [DE], but I finally got it last week. The package’s contents inspired
a bit of skepticism in me – the headphones are basically your average no-name
phones with the insulation and ear-pieces grafted on with silicone gel, and the
manual is ink jet printed on a single page. The bag, however, looks a bit more
industrial, and even comes with a printed manual. I followed the recommendation
to perform a "bubble test" with some worthless stuff in the sink, and
it emerged perfectly dry.

I then took the thing to the pool and actually put my iPod inside. To keep
you from skipping to the end: it worked. It doesn’t look all that elegant when
I’m toting the rather large bag around my waist, but I was surprised that it
doesn’t seem get in the way at all when I’m swimming. I haven’t yet discovered
a completely successful technique for putting the headphones in my ear so that
they stay there for longer periods – at the moment, I have to readjust them
every five laps or so. Sound quality is rather abominable compared to my usual href="http://www.amazon.de/exec/obidos/ASIN/B00008XYJL/302-1099565-5853646">Sony
[DE] or href="http://www.shure.de/produkte_psm/product-shr-472-56-119.htm">Shure
[DE] headphones, but that’s probably to be expected due to the price and
grafted-on insulation. Still, it’s a lot better than listening to the screams of children
and the other usual pool noise.

Reflecting on the story, it’s a rather fine example of the virtues of our
connected society – I’m pretty sure I would never have found out about the
company in Florida that offers this thing if I had to scour obscure catalogues
or read swimming enthusiast newspapers. And while grafting plastic pieces on
headphones is probably hardly ever used as an example for an innovative product,
there were these people across the Atlantic Ocean who had built a solution that
quite exactly matched my requirements, and it took me about five minutes of work
to have their product delivered to my doorstep. Yay!

This is one of the fundamental problems in my life, and I suspect it’s the
root of a significant part of humanity’s problems (since it keeps most people
from rationally considering the world). As an (I believe) deeply rational
person, I have been in for a number of unpleasant discoveries. Some of these
involve:

• There Is No Final Truth.

  I firmly believe in href="http://en.wikipedia.org/wiki/Empiricism">Empiricism [EN] as the
  best basic approach to understanding the world. Since all knowledge about the
  world must ultimately be based on observation, we can never be sure that
  our understanding is final. It is rather more likely, given our historical
  record of being wrong countless times, that the current understanding is
  going to be replaced some time in the future.

  The only discipline where absolute and "eternal" truths exist is
  mathematics, but these are never applicable to the real world without an
  empirical model (Standard example: Mostly everyone knows that 1 + 1 = 2. So,
  your typical 17th century philosopher would "deduce" that 1 liter
  of water + 1 liter of alcohol = 2 liters of mixture. He would be wrong, and
  this is why physicists experiment.)

  While the empirical approach and an epistemology based on falsification can
  do a great job when exploring (probably fairly static) laws of nature, the
  challenges are much greater when trying to deal with social problems – human
  beings and their relations seem so dynamic that I tend to believe it’s impossible
  to predict anything useful on a large scale. Of course, this thought or a
  similar one has occurred to philosophers a long time ago, and gave us href="http://en.wikipedia.org/wiki/Pragmatism">Pragmatism [EN], which
  is probably the only usable approach to dealing with social problems.

  While I believe in pragmatism because social problems are much too complex and
  dynamic to be solved by strictly adhering to any preconceived doctrine without
  adjusting it to experience, it is unatrattactive emotionally. I see a
  strain in myself that longs for being securely right, something to believe
  in. As David Brin [EN] put it in
  The Transparent Society [EN]:

  It seems that the Pleasure Principle is back from the dead. This once
  discarded theory held that humans and animals behave in certain ways because
  these actions feel good. For much of the twentieth century, the
  notion was abandoned in favor of behaviorist concepts propagated by
  B. F. Skinner, or more recently cognitive theories. Only now are
  neurochemists finding that a great many of the things we do are reinforced by
  the release of psychoactive molecules in our brains.

  [... pleasure & drugs stimulate endorphin release ...]

  There is a third way to excite these pleasure centers. Humans can learn to
  trigger the release of endorphins and other natural chemicals in the body by
  voluntarily entering certain mental states. For instance, ever since the days
  of William James, it has been known that deep religious faith can have
  similar psychological effects on believers, across nearly all boundaries of
  doctrine and creed. Studies of experts in the arts of meditation have shed
  light on the agreeably detached conditions that they enter by volition,
  enabling some aficionados to develop shortcuts using computerized biofeedback
  equipment, training themselves to achieve pleasurable levels of
  (non)consciousness at will.

  One particular mental state seems quite effective at enabling individuals to
  self-administer these psychotropic chemicals on a massive scale.
  Self-righteousness is an especially heady condition that all of us
  have experienced at one time or another. Those who are honest will admit there
  is something sickly-sweet and alluring about knowing you are right
  while all others are terribly wrong. However many unpleasant or anxious
  feelings may accompany a crisis, righteousness is one side effect that can
  imbue a person with sensations of romantic virtue, and the satisfaction of
  feeling like a martyr, or the lonely champion of a pious cause.

  You entirely miss out on all this if you admit that your theory is just as
  good as the next guy’s until it has been proven to work better in practice
  (and even then, it might not be the correct choice for tomorrow). Notice the
  complete absence of this admission in real-life political debate – nobody
  admits to the possibility of being wrong, and discussion is generally framed
  into good-vs-evil us-vs-them terms. I am quite sure that the blame for this
  can be put on the missing appeal of rationality – would you vote for the guy
  who says "Let’s try this, it might work for the next few months" vs the one
  who says "This will be the best thing since sliced bread, I guarantee"?

  Without any real evidence to back this up, I tend to believe that the
  political class does not just blur this distinction out of propaganda
  reasons, but that they actually believe in their world views, and
  many of them do so in the deeply dangerous way that make them ignore
  empirical evidence that they are wrong. The current USian administration
  provides a good example, see Brin’s href="http://www.davidbrin.com/neocons.html">excellent [EN] href="http://www.davidbrin.com/neoromantics.html">write-ups [EN] on the
  topic. However, the same thing can be observed in European politics as well -
  it seems that rationality and pragmatism are losing ground for a recurrence
  of faith and traditional belief in restrictive laws, repressive order and the
  "natural" authority of a ruling class.

• I’m Insignificant.

  I may be smart. I tend to believe that I am, since I always did well in
  school and do a good job beating computers into whatever it is that I want
  them to do. However, that doesn’t mean much – I might be able to do these
  things better than most other people on the planet, but take a handful of
  them, and they’ll almost certainly do them better than I can.

  The same holds true for most people. Even the ones that have risen to
  fame in world history mostly were just playing roles in their environment,
  determined by uncountable factors – if Hitler had actually href="http://www.radioislam.org/historia/hitler/mkampf/ger/02.htm">been thrown
  off a scaffolding (p. 42)[DE], I’m quite sure that wouldn’t have avoided
  the general course that history took (of course, this is kind of a non-issue
  since the question is fundamentally undecidable unless someone comes up with
  a way to scale back time and do historical experiments).

  Even if I come up with something smart that causes people to name 2105 href="http://www.einsteinyear.org/">Themel Year [EN] (something along
  these lines has played the role of "lifetime achievement fantasy"
  for me for some time), the same probably applies – someone else would have
  come up with it, a year later at worst, but it doesn’t matter all that much.

• I, The Individual, Will Die

  I used to be a radical individualist – I had somehow (Thanks, Ayn Rand!)
  conceived of the vague notion that it was morally questionable to exert
  coërcive influence over others Consequently, I thought that science or
  something technical would provide an occupation fitting my notions of
  morality – mastering some programming problem or discovering some physical
  property would not coërce anybody, and at the time I considered myself
  sufficiently good at IT stuff to make it my occupation for life.

  This radical approach suffers from a number of impracticalities. Trying to
  live according to non-coërcion principles in a society that employs coërcive
  conflict resolution is tricky – with to Libertarian ethics, withdrawal is the
  only acceptable strategy for conflict resolution where there is no simple
  decision according to the few basic tenets (property right, non-coërcion
  etc). This, in turn, produces the oft-commented upon need for a frontier -
  Libertarian ethics only function when the grudging individualist has new land
  to settle when civilization moves in and needs the space that he has claimed
  “from nature” for hordes of despised “normal citizens”.

  One of the problem with this withdrawal and the resulting disconnect from
  society is that even the smartest and most independent individual will die
  some day. There is a number of common ways of dealing with this issue –

  1. Religion: $HIGHERBEING exists and a part of me will live on after my
     body dies.
  2. Procreation: The purpose of my life is to produce offspring and model them
     (as far as this is possible) to become better versions of myself.
  3. Transcendence: The purpose of my life is to produce something that will
     outlive me, and attach my name to it.

  I believe Number 1 is soundly defeated by philosophy and science – I can not
  imagine a theory where the proposed $DEITY does not fall to
  Occam’s Razor [EN].

  Number 2 seems like an annoyingly common option to me – after all, it’s the one
  adopted by most multi-cell organisms, and it requires nothing uniquely
  human nor any mental skills to procreate.

  Number 3 used to be my approach of choice, but it suffers from a number of
  problems. First, it’s comparatively hard to pull this off in a way that will
  make me remembered/notorious for more than a human life span. Second, what
  will survive is really a lot short of what makes up me, the individual -
  probably it’s just a damn picture and some euphemisms/lies about my life as
  they fit the agenda of each particular observer. Most importantly, it requires
  someone to implant the meme into – so if I don’t dig culture and despise
  society, I can consider myself dead. Even if I manage to blow up the
  whole damn planet and end humanity while forming a cloud in the shape of my
  name, as long as there is nobody to observe it, it’s over. I’m dead,
  forgotten and gone.

  Nowadays, I still believe in Number 3, but the depressing insight above
  caused me to adjust my view on individualism and society to something a lot
  more moderate. The "I’m Insignificant" point above also reinforces
  that shift. And, romantic bullshit aside, I’ll probably end up pulling a
  Number 2, and not being very good at it because it’s not something I want to
  do. What makes this so depressing is that after a whole huge tour through
  science and philosophy, I’ll end up just next to every common idiot who never
  thought for a moment about deeper issues than "What’ on TV
  tonight?". There goes all the romantic elitism, and that sucks.

So, it seems that after much studying, I have reached a point where I’m not
all that happy with what I know. Was there a point to it? Well, I think so.
Bertrand Russell
[EN] wrote in his href="http://www.users.drew.edu/~jlenz/decalog.html">Liberal Decalogue [EN]:

Do not feel envious of the happiness of those who live in a fool’s
paradise, for only a fool will think that it is happiness.

I somewhat share this sentiment – although I don’t think it makes me a better
person (for there is no such thing as a one-dimensional good person/bad person
metric), having reached my conclusions (which, of course, are anything but
final) via the long and agonizing yet enlightening and entertaining way is
preferable to simply copying what my parents or peers do.

However, it remains to be seen if there will ever be a tangible benefit to
knowing that many of the things others never have questioned are just
conventions instead of natural laws. Maybe I just discovered a really tedious
way to trigger my endorphin emitters…

Now, it seems everybody hates pedophiles (and in cases where actual
children get hurt/abused/harrassed, I do understand), which is why they got their place
among the four href="http://lists.virus.org/cryptography-0410/msg00010.html">Horsemen of the
Infocalypse [EN] and served as an example in the famous case of href="http://kris.koehntopp.de/inkomploehntopp/00078.html">de.alt.sex.pictures.children
[DE] ("Köhntopp’sche Idiotenfalle", which was a nice idea but
failed miserably).

A debate on whether having href="http://www.k-otik.com/exploits/09272004.JpegOfDeathM.c.php">a certain kind
of JPEG [EN] on your hard drive should get you thrown in jail aside, we have
a new twist on the issue: href="http://www.local6.com/news/3835658/detail.html">This kid [EN], age 13,
faces charges for posting pictures of himself. Funny detail: 60% of the
voters in the attached poll think this is justified.

What’s your opinion?

I was kind of surprised I couldn’t find an RSS feed on href="http://www.kernel.org/">kernel.org [EN]‘s main page. I was even more
surprised when I couldn’t find anyone roviding one on Google. Since I no longer
consider "finding out about things by regularly looking at a web page"
an acceptable method, and "reading lkml" isn’t an option on my time
budget, either, I just built my own
[RSS].

It currently scans the 2.4 and 2.6 archives on kernel.org, including
snapshots (-bk patches) and testing (-rc kernels) branches. I probably should
add a CGI interface to allow requesters to select trees, but I’m waiting for
someone to actually pop up and request it before putting in the effort – the
current version fits me alright.

The actual FTP and RSS bits were quite trivial, thanks to href="http://www.perl.org/">Perl [EN]‘s href="http://search.cpan.org/dist/libnet/Net/FTP.pm">Net::FTP [EN]
and XML::RSS
[EN] modules. Building the intermediate data structures in Perl was a
pleasure as always, given the fun little details of reference semantics,
and it’s subtle syntactical sugar for accessing nested data structures…

${$releases{$1}}{$field} = "$branchdir/$name";

I should have a look at Python [EN] one of these days.

Regardless of my opinion on state-imposed television fees to maintain a public broadcasting service, this [DE] is simply asinine. More than EUR 200 a year just because I could use my Internet-enabled PC to access some program produced by the state’s pet television company? And of course, they already consider imposing the same tax on UMTS mobile phones. Consider the fact that most providers manage to sell you access to their (well, implicitly, THE network) on a highly subsidized phone and some free call time for less…

The funny thing is that I’m hardly the only person who doesn’t even own a TV set because there’s very little worthwhile stuff on TV that doesn’t also come in written form or as a torrent [EN] or DVD. Due to the fact that I didn’t have to pay, I wasn’t too annoyed by the whole “TV tax” issue, but if they ever consider doing this in Austria, it’s
time to pull a Tim May [EN] on this place [DE] (Note to whoever does it: use precision technology, you wouldn’t want to unnecessarily diminish choice in local universities, would you?).

On rereading this, I might seem a bit agitated. Well, I am.

My contribution to the href="http://piskernig.at/blog/index.php?p=114">recent wave of desktop wallpaper
exhibitionism [EN]: hearty contempt!

Desktop wallpapers? I’m not wasting screen real estate for fancy images. As
a user of the One True Window
Manager [EN], I don’t have background most of the time. Switching to ion made
me realize how much time I actually spent moving, resizing and otherwise
adjusting Windows (z-order on an xterm cluttered desktop can be a real pain!) in
Enlightenment [EN]. Ion does
most of it for me, plus it’s highly customizable and I don’t need to take my
hands off href="http://shop.store.yahoo.com/pfuca-store/haphackeylit1.html">the keyboard
[EN].

Here’s an impression of the workspace I spend most of my non-browsing time in:
1280×1024, 4 same-size xterms, Psi when I’m feeling particularly communicative.

The few apps that are unusable with ion (xmms, oo.org, gimp on the rare
occasions I need it) live in a FloatWS without any background – no need to make
them feel comfortable about it, after all!

Currently, I’m quite happy with ion, but there’s already the next revolution
in X window management lurking around the corner – href="http://wmi.modprobe.de/">WMI [EN] combines the best features of ion and a
number of other window managers. Its most obvious advantages over ion are support
for mixing floating and framed windows on the same workspace, and a dock (called
“slot”) for applications like IM clients or MP3 players that can be recalled on
any workspace. It also has a really nifty vi-like command interface, and it’s
written in cool, STL-enabled C++! I haven’t quite had enough trouble with ion
yet to hack together an equally optimized WMI config, but it’s definitely the
way to go.