Neverending Dream

What we lost

noreply@blogger.com (sothim) — Sun, 08 Nov 2009 18:59:00 +0000

Just to recall that we had a free, generic, distributed and simple file exchange framework with a broad user basis for exchange of documents, pictures, music and video. Namely Napster, eDonkey and co. It is all gone now due to the greed of the big music labels, only on the basis that some files for which they had licenses for were shared on those networks. Think about it the next time you buy something from them and when you are searching a rare old track that you cannot buy nowhere nowadays. Napster and co would likely have solved your problem.

A case for LaTeX against WYSIWYG

noreply@blogger.com (sothim) — Tue, 27 Oct 2009 17:39:00 +0000

Nobody would argue that Word, WordPad and OpenOffice writer are easy to use tools that quickly produce output. In comparison with LaTeX even more so!

However, often hidden information is transported in files of aforementioned editors like the username, date of creation, even version histories with previous edits. With LaTeX you really see and permanently work on the (hopefully) only input to the tools that then actually produce the pdf or other output format. Thus it eliminates hidden information in your own documents. Last week a professor recommended copy and pasting the text into a new WordFile before submission - what an utmost ugly solution, and ineffective with for instance usernames and "white font" text etc.

Another thing that I just recognized is that LaTeX sources might enable an easier semantic search on text documents. Like in programming you abstract when you write documents. I just finished some document where I used the dialogue package to state that this part of the document forms a dialog between persons, but not just by formatting it like:

A: lorem ipsum
B: ipsum lorem
C: A habla lorem ipsum

but:

\begin{dialogue}
\speak{A} lorem ipsum
\speak{B} ipsum lorem
\speak{C} \refer{A} habla lorem ipsum
\end{dialogue}

This has the advantage that a software can automatically deduct that a dialog is actually going on and which persons are speaking, etc; in contrast to the first part where it is only deductable that some colon was used.

Of course the dialogue package only uses this info for formatting reasons. But imagine searching semantically for dialogues where Socrates has some say and getting better results through such a technique. Of course some tasks would be harder in WYSIWYG editors.

Video Surveillance

noreply@blogger.com (sothim) — Sun, 25 Oct 2009 18:06:00 +0000

With all the growing video surveillance in the cities and everywhere else, one has to wonder which spots are actually free as in freedom? Also of interest, how are the videos retained and who has access to the tapes?

If there isn't one yet, one should create a database with the cameras and overviewed places.

Right to Left assignment (Left := Right)

noreply@blogger.com (sothim) — Mon, 02 Feb 2009 22:29:00 +0000

Quite everyone seems to have settled on the convention that "further-right-things" are built of "further-left-things" according to the left-to-right reading order. Grab any scientific paper and check for that. However the most basic mathematical building blocks say "let x be -y-" where the order really is reversed, as in most equation/assignment formulations and definitions.

Are the roots of that found in the Arabic influence of Al-Khwārizmī with the Arabic "right-to-left" writing? Fact: I don't know.

Anyways, I find it quite irritating to have right-to-left assignment with left-to-right reading order. So one of those should change, choose one. You should read a construct (expression, definiendum, etc) and understand it before a name is chosen, because the first is required to be understood to get an idea what the last will stand for.

Put in other words: The "information flow" direction (in an very informal manner) should stay the same throughout any text for easier comprehension.

So how can any mathematician justify that very basic notation order besides tradition?

Been a long time

noreply@blogger.com (sothim) — Mon, 02 Feb 2009 22:16:00 +0000

Simple reminder to make the world work: Don't let yourself tell what is best, think for yourself and do the thing that YOU think is best for everyone. That is the only way community works!

Deprecate return values?

noreply@blogger.com (sothim) — Fri, 19 Sep 2008 21:32:00 +0000

Return value handling in imperative languages is really awkward.

In many you cannot have multiple return values. I propose a syntax for making in and out parameter handling symmetric:

(type1 in1, type2 in2) myFunction (type3 out1, type4 out2)
{
in1 in2 + =: out1;
}

whereas in function composition, all out parameters form the input parameters for the next function:

( ( in1, in2 )myFunction )myFunction2 =: res;

Control codes and Caret Notation

noreply@blogger.com (sothim) — Wed, 28 May 2008 19:06:00 +0000

For some time I have been confronted with ^C and the like, well, right now I was wondering and finally looked it up. It is officially called the Caret Notation.

Furthermore you can use

cat -v [FILE]

to make cat output Control Codes in Caret Notation for a bit of readability of binary files.

Wikipedia has some more articles on the topic: C0 and C1 control codes and Control code.

For instance, did you know that Control subtracts 64 from the parameter character (uppercase) ASCII value, which for ^C (C=67) results in End of text(=3).

La música electrónica

noreply@blogger.com (sothim) — Sat, 29 Mar 2008 12:48:00 +0000

Been there (Annual Pressure Festival), heard that (Klaudia Gawlas). She does techno the hard way, had a great time listening. Also take a look at the videos for a sound test!

There is also a 'Close your eyes (CYE)' mix by her travelling around, find it or request per mail address from the hompage. Thats a great one. Looking forward to seeing her again, in Linz or next Pressure. Give her a try.

Road: 730, Terror: 0

noreply@blogger.com (sothim) — Fri, 08 Feb 2008 20:51:00 +0000

Terrorism is currently seen as an imminent threat to Europe and thus also Austria. The possibility of terrorist attacks is widely used in political context to push privacy and civil right reductions.

I thought about that argument and looked up traffic deaths. According to the official source, the Statistik Austria, in 2006 (latest available numbers) 730 people died due to traffic accidents. Compared to 0 terrorism deaths (I could not find any statistics, but am quite sure, I wouldn't have missed a terror attack).

So I claim, that there doesn't exist any measure on terrorism attacks, thus the effectiveness of counter-measures against such cannot be proved.

Okay, I do not want to say that there does not exist any threat. But I first do want to say, that there are other more important problems and threats to life and that terrorism threats are gaining too much attention and people argue as if we would face daily bombing attacks, worse than in Baghdad.

Second, I think terrorism has to be fought on a very different level, a much earlier stage than surveillance of suspects. One would have to avoid that people even want to make terrorist attacks, so prevent their grief and anger.

Let me quote the Verfassungsschutzbericht 2005 - they watch threats to the constitution, terrorism and such - on the 2004 Madrid bomb attacks:

Begründet wurde der Anschlag mit der Stationierung spanischer Truppen im Irak.

So what are the reasons for terrorist attacks? You just read one. How to dissolve them? I propose:

Not fighting causeless prevention wars

Help those in need in war-troubled reasons - eg: sending aid to Iraq, if necessary with some publicity

Repeatedly and officially (by government) opposing and criticizing unjust behavior such as the US attack on Iraq and the situation in Guantanamo

Sending objective watchdogs to war regions and announce their reports

Work closely with immigrants and others that may have emotional bindings to conflicts in the world and provide an open forum for exchange how to improve the situation

This behavior would be much more effective than sending troops supporting illegal wars. If somebody would invade and occupy Austria on such a reasoning basis like the US did, I would fight them and their supporters too.

Musikindustrie bestiehlt Österreicher um 15 Millionen Euro

noreply@blogger.com (sothim) — Sun, 03 Feb 2008 00:21:00 +0000

Dem Bericht der Austro Mechana Verwertungsgesellschaft wurden 2006 erfolgreich 15 Millionen Euro von den österreichischen Bürgern im Rahmen der Leerkassettenvergütung (auf CDs, MP3 Player, DVDs, uvm) geraubt.

Protesting data retention wasn't in vain

noreply@blogger.com (sothim) — Tue, 29 Jan 2008 15:54:00 +0000

Seems like the demonstrations against data retention in Germany were not in vain. They seem to have convinced German Justice minister Brigitte Zypries.

Heise runs a story about her on the police congress backing her arguments against the Transparency of the citizen with the recent Germany-wide demonstrations:

Gegen diese ihrer Ansicht nach ausufernde Sammlerei verwahrte sich Justizministerin Brigitte Zypries (SPD). Unter Verweis auf die Proteste von über 10.000 deutschen Demonstranten, die gegen die Vorratsdatenspeicherung auf die Straße gegangen sind, erklärte Zypries, eine Fluggastdatenbank mit einer Eingriffstiefe von 13 Jahren sei der falsche Weg.

Thanks to all protesters, organizers and to Misses Zypries for trying to uphold the constitutional right for privacy.

Landtagswahl Hessen 2008

noreply@blogger.com (sothim) — Tue, 29 Jan 2008 11:14:00 +0000

"Dafür [Ann: dass die Wahl manipuliert wurde] gibt es keinerlei Anhaltspunkte, da das vorläufige Wahlergebnis in der Stadt Niedernhausen dem Hessentrend entspricht." Nur wenn das anders wäre, könnte man überhaupt über eine Neuwahl in diesem Wahlkreis nachdenken.

(aus Spiegel Online im Interview im Landeswahlleiter Wolfgang Hannappel)

Interessant, wenn ich das richtig lese sind also nur jene Wahlergebnisse gefälscht, welche nicht im Trend liegen? Wieso wählt man dann eigentlich noch? Man könnte gleich die Statistiker um das Ergebnis befragen und so viel Geld und Zeit sparen.

Something like democracy (€50 per vote)

noreply@blogger.com (sothim) — Sun, 27 Jan 2008 20:32:00 +0000

The CCC tried to independently observe elections in Hesse, Germany. Well it turned out, not everybody likes to be observed - so they have something to hide, don't they?

Read reports, especially about the probably manipulated elections in Obertshausen, where it seems they have something to hide. Something big enough to justify turning away independent observers based on constructed reasoning and stalk them to make sure that they are gone.

Deprecate Arrays?

noreply@blogger.com (sothim) — Thu, 17 Jan 2008 13:39:00 +0000

Most (imperative) languages got some built-in support for array types and I was asking myself, if it wouldn't help us all to remove the special type from the language and instead treat them what they really are, to help us recognize them as that and get a better understanding of collections types?

What I want to say, would it make sense to replace

String[]

with

Map< Integer,String >

For clarity, the accessing of an array through the indexes could be viewed as a Dictionary=Map type usage, mapping an Integer to some other Type. To actively counter first no votes, you can also specify the initial size in the constructor type and I think optimization algorithms could deduce the original situation from the Map.

Of course some Generics/Template implementations right now are kind of verbose, but let's leave that detail out and assume some type inference.

Java Definition order matters

noreply@blogger.com (sothim) — Mon, 14 Jan 2008 23:12:00 +0000

Ever encountered a illegal forward reference in Java? Thought that those days were gone and that compilers finally managed to get the initialization order right. Well, turns out not in all cases.

The Java Language Specification tackles that topic in section 8.3.2.3. It basically says the declaration has to textually appear before another if it's used in a member initialization expression.

The static initializer also has this restriction. Well perhaps thats a reason, why many Java programmers declare Variables at the beginning of their class. I still think thats not the best approach, as the public interface - which should give the best overview of the class - should appear first (textually) to ease comprehension.

24C3 Proceedings

noreply@blogger.com (sothim) — Mon, 14 Jan 2008 14:58:00 +0000

Just came across the 24C3 proceedings.

OTR

noreply@blogger.com (sothim) — Wed, 09 Jan 2008 20:21:00 +0000

Off-the-record.

Open Source Encryption extension for (over) IM Protocols with plugin for pidgin!

http://www.cypherpunks.ca/otr/

aptitude show pidgin-otr

Enjoy the privacy! If you hadn't known yet, at least ICQ is transmitting messages in plain text or the Internet and the local network, so anyone can listen in or probably even inject messages without you knowing it.

OTR seems to solve this problem.

Quicksort in C#

noreply@blogger.com (sothim) — Wed, 09 Jan 2008 16:48:00 +0000

Amazingly short in code and even intuitive if you are a bit into functional programming:

Func< intlist, intlist > Sort = 
  xs =>
    xs.Case(
      () => xs, 
      (head,tail) => (Sort(tail.Where(x => x < head)))
                             .Concat
                           (Single(head))
                             .Concat
                           (Sort(tail.Where(x => x >= head)))
    );

Taken from this presentation, also find more interesting features from C# 3.0 there.

Now argue with me, that some languages are not clunky compared to this!

Chained comparisons

noreply@blogger.com (sothim) — Wed, 09 Jan 2008 16:09:00 +0000

Have you ever wondered about writing

if( 0 <= x < max ) { ... }

in your program? Did it work? Like you intended?

The (by me) expected semantic for this expression is the same like the mathematical, or if you want to talk in C:

if( (0 <= x) && (x < max) ) { ... }

However in C, the <= operator is left-associative meaning, 0 <= is evaluated first and the integer 1 is the result if the test is true and the second test against max is made to this 1, thus the result is most likely a bug. C++ upholds this C legacy.

Java doesn't make it work, but at least detects it as an error through the dedication that comparisons in general result in an result of type boolean, thus you get this:

operator < cannot be applied to boolean,int

compiler error.

Python (tested) and Ruby (read) implement the "right" behavior, Perl 6 is about to implement it. Concerning D, I came across this discussion.

The legendary BTX Hack

noreply@blogger.com (sothim) — Sun, 06 Jan 2008 19:29:00 +0000

The Chaos Computer Club back in the eighties hacked the BTX - Bildschirmtext to demonstrate the vulnerability of the system through transferring 130.000 DM indirectly to their own bank account.

The recording of the back-then news journal report has been put up at Google Video.

It features an interview with Wau Holland.

Shared source

noreply@blogger.com (sothim) — Fri, 04 Jan 2008 00:00:00 +0000

Shared source programs allow restricted auditing of normally closed source code. This is mostly done by software vendors to get contracts from governments for sensitive application areas. Examples include Microsoft's Windows OS and Voting machine vendors.

The point I want to bring up is that though you can use that kind of audit to further assure that the software vendor didn't make any unintentional errors, normally this doesn't help you against intentional backdoors.

Why that? Well unless you really build the software yourself, the vendor could show you any source code and deliver any malicious software. This is the point where I doubt, that the auditor really takes the given source and builds the product from there on after the audit.

So those audits still imply trusting the vendor. Thats the interesting point, do you really trust the voting machine vendor? Imagine what you could do with the control over elections in a country. Or the control over a government computer through controlling the operating system.

Let's drop a tear

noreply@blogger.com (sothim) — Sat, 22 Dec 2007 02:34:00 +0000

Due to rights and licensing issues, in most countries a piece of music named "House End Credits" is used,...

Wikipedia on House (TV series). So the song Teardrop by Massive Attack only appears in english versions of the series.

You wouldn't expect that the music industry's shitwalk in the recent years would affect your day-to-day life. Well as I just recognized, it does. I hated the change of the title theme of my currently favorite TV series and wished Teardrop back.

Unbelievably they have been able to fuck up even that. At least I can smile about their loss during the christmas sale on Amazon as some kind of revenge.

Microsoft Wireless keyboards hacked

noreply@blogger.com (sothim) — Mon, 03 Dec 2007 23:09:00 +0000

Microsoft's Wireless Optical Desktop 1000 and 2000 products (probably more) are affected.

Synergy

noreply@blogger.com (sothim) — Sun, 02 Dec 2007 14:48:00 +0000

I think I haven't been writing about Synergy, but as it is an essential tool for me, I want to introduce it to you too. No I don't mean the buzzword thing, I'm speaking of:

Synergy Keyboard and Mouse Sharing

Open source and cross-platform software (Windows,Linux,Mac), for cases where you have multiple screens on your desk. You can configure Synergy, to move onto the other PC, when you move the mouse over the left/right end of your main computer. And then directs Mouse and Keyboard input to the other PC!

So my setup here is basically a Laptop and a desktop PC, where I'm using the desktops keyboard and mouse to work on both of them and thus expanding my work area.

It's a good replacement for other remote softare, when you have the "remote" PC's screen on your desk anyways.

However, it's missing a maintainer to eliminate the existing minor bugs and port it to Vista the proper way, searching for a project? Give it a try, it has got very well structured C++ source and architecture as far as I can tell from the first glances.

What if the music industry...

noreply@blogger.com (sothim) — Sat, 01 Dec 2007 21:37:00 +0000

okay, I think this will be the first one from a series of posts, so let's get started.

For those series, we pretend the music industry is already dead (this might become true while the series is running, so at that point, you may ignore this premise).

Today:

What if the music industry... would have done youtube?

Especially for music videos and perhaps with an option to buy music titles/albums right from the place.

Looks to me, like that would have been a good deal for the transition to the digital age and might even have resulted in higher profits.

However, that has not happened. And instead of reacting to youtube in a senseful manner, with respect to licensing and so on - they are suing the digital media companies, instead of making deals with them! Thats like hating cars and suing everybody to ride on horses again!