Was im Internet steht ist bekanntermaßen wahr, und was in Zeitungen geschrieben steht noch wahrer. Also muss, was in Onlineauftritten von etablierten Zeitungen steht, die Quelle aller Weisheit schlechthin sein. Natürlich ist klar, dass auf einem Onlineauftritt der einige Angebote umfasst, die in der gedruckten Ausgabe nicht vorhanden sind, manche Dinge nicht ganz genau geprüft werden. Trotzdem ist es immer wieder ein erhellender Aha-Effekt wenn man direkt auf einen Fehler stößt wie hier.
Im konkreten Fall geht es um die Rubrik Die Auslese der SZ, in der eine Meldung über ein Depeche Mode-Konzert in Peru wiedergegeben wurde. Ich zitiere: "Musiker wissen bei Auftritten nicht immer, wo sie gerade sind. Diesmal war es Dave Gahan von Depeche Mode, der sich beim falschen Land bedankte. 'Vielen Dank, Chile', rief Gahan den Fans in Lima zu - dabei mögen sich Peru und Chile nicht gerade gerne."
Zum Glück wohnt Lisset in Lima und ist zudem Depeche Mode-Fan(in) und war auf besagtem Konzert - und was Dave Gahan wirklich gesagt hat, war "Thank you, children". In der Folge haben das peruanische Journalisten dann wohl falsch verstanden - ob absichtlich oder nicht - und, da es tatsächlich eine gewisse Nachbarliebe zwischen Peru und Chile gibt, die ganze Geschichte künstlich aufgeblasen.
Bis dann eben ein Redakteur in München in einem Nachrichtenticker darübergestolpert ist und die Meldung ohne große Prüfung übernommen hat.
Lerne, wie die Welt wirklich ist, aber vergiss niemals, wie sie sein sollte.
Freitag, Oktober 30, 2009
Samstag, Oktober 24, 2009
Eine elegante Definition von Mathematik
Die meisten Nichtmathematiker haben nur eine sehr schwammige Vorstellung davon, was Mathematik eigentlich ist. Man muss es ihnen verzeihen, schließlich tun sich Mathematiker selbst auch schwer damit, eine klare Definition zu geben. Einen sehr schönen Versuch einer Definition davon, was Mathematik eigentlich ist, las ich kürzlich in einem Artikel von William Thurston:
Mathematics is the smallest subject satisfying the following:
Manch Futurist mag sich an der diskriminierenden Verwendung des Wortes "human" stören. Diesen Mangel könnte man vielleicht zum Vorteil umkehren, um den Begriff der Intelligenz zu klären: wir könnten sagen, die Objekte einer Klasse C sind mindestens so intelligent wie Menschen, wenn in der obigen Definition die Klasse der Menschen durch C ersetzt werden kann, ohne das dadurch definierte Feld der Mathematik zu verkleinern.
Viel Spaß bei der Spekulation über diese Definitionen.
Mathematics is the smallest subject satisfying the following:
- Mathematics includes the natural numbers and plane and solid geometry.
- Mathematics is that which mathematicians study.
- Mathematicians are those humans who advance human understanding of mathematics.
Manch Futurist mag sich an der diskriminierenden Verwendung des Wortes "human" stören. Diesen Mangel könnte man vielleicht zum Vorteil umkehren, um den Begriff der Intelligenz zu klären: wir könnten sagen, die Objekte einer Klasse C sind mindestens so intelligent wie Menschen, wenn in der obigen Definition die Klasse der Menschen durch C ersetzt werden kann, ohne das dadurch definierte Feld der Mathematik zu verkleinern.
Viel Spaß bei der Spekulation über diese Definitionen.
Sonntag, Oktober 18, 2009
Politik im Entstehen
Ab und zu kommt man in das fragwürdige Vergnügen, beobachten zu können, wie in der Politik gezielt die Manipulation der öffentlichen Meinung mit Fehlinformation gesät wird. Sich darüber zu beschweren, dass Google Bücher einscannt, ist vollkommen Fehl am Platz. Um gegen unerlaubtes Verbreiten von Büchern übers Internet vorgehen zu können bedarf es keines besonderen Urheberrechts. Dafür würde auch das Urheberrecht des 19. Jahrhunderts ausreichen.
Genauso unsinnig ist die Formulierung, sich für Urheberrechtsschutz im Internet einsetzen zu wollen; dass sie überhaupt genutzt wird zeugt davon, wie wenig weit das Verständnis für die Thematik in unserer Gesellschaft gediehen ist. Würde die Kanzlerin von sich geben, sie wolle sich für Urheberrechtsschutz auf Papier einsetzen würde sie zu Recht lächerlich gemacht werden.
Es ist richtig traurig, was hier passiert: ein namenloser Lobbyist sorgt aus dem sicheren Hintergrund dafür, dass der Kanzlerin Worte in den Mund gelegt werden, die ein Strohmannargument aufbauen. Es wäre in der Tat nicht akzeptabel, wenn Google einfach so Bücher kopieren und verbreiten dürfte. Die meisten Menschen werden aber nicht durchschauen, dass es sich hier um ein trojanisches Pferd handelt. Letztendlich wird die öffentliche Stimmung dann genutzt werden, um kulturell destruktive Gesetzesänderung durchzubringen.
Natürlich gibt es Schwächen im heutigen Urheberrecht, von der viel zu langen Laufzeit bis zur desaströsen Klausel über "wirksame technische Schutzmaßnahmen". Aber die einzige Branche, die legitimerweise den Umwälzungen durch das Internet mit Sorge entgegensieht ist der Journalismus. Alle anderen erhoffen sich doch nur, das Urheberrecht als Mittel zur Kontrolle und Zensur mißbrauchen zu können. Es ist die Pflicht unserer Volksvertretung, dagegen auch präventiv vorzugehen.
Genauso unsinnig ist die Formulierung, sich für Urheberrechtsschutz im Internet einsetzen zu wollen; dass sie überhaupt genutzt wird zeugt davon, wie wenig weit das Verständnis für die Thematik in unserer Gesellschaft gediehen ist. Würde die Kanzlerin von sich geben, sie wolle sich für Urheberrechtsschutz auf Papier einsetzen würde sie zu Recht lächerlich gemacht werden.
Es ist richtig traurig, was hier passiert: ein namenloser Lobbyist sorgt aus dem sicheren Hintergrund dafür, dass der Kanzlerin Worte in den Mund gelegt werden, die ein Strohmannargument aufbauen. Es wäre in der Tat nicht akzeptabel, wenn Google einfach so Bücher kopieren und verbreiten dürfte. Die meisten Menschen werden aber nicht durchschauen, dass es sich hier um ein trojanisches Pferd handelt. Letztendlich wird die öffentliche Stimmung dann genutzt werden, um kulturell destruktive Gesetzesänderung durchzubringen.
Natürlich gibt es Schwächen im heutigen Urheberrecht, von der viel zu langen Laufzeit bis zur desaströsen Klausel über "wirksame technische Schutzmaßnahmen". Aber die einzige Branche, die legitimerweise den Umwälzungen durch das Internet mit Sorge entgegensieht ist der Journalismus. Alle anderen erhoffen sich doch nur, das Urheberrecht als Mittel zur Kontrolle und Zensur mißbrauchen zu können. Es ist die Pflicht unserer Volksvertretung, dagegen auch präventiv vorzugehen.
Dienstag, Oktober 13, 2009
Radeon 3D Wiki drive
Outsiders of the group of developers regularly complain that it's very hard to get an understanding of what happens in development of the Radeon driver(s). They have convinced me.
So I've spent some time to refactor the documentation on the DRI Wiki, creating a new, cleanup up and up-to-date portal page in the process. However, this refactoring is not complete, and the amount of information is not yet entirely satisfying. The real test now is whether this can be more than a one-shot effort by me, so this is a shout out to others - particularly to the users who have been asking for this - to help improve the documentation.
Partly, this can be done in very mundane ways by normal users. The application support matrix needs to be updated for Mesa 7.6 (there have been a lot of significant improvements), which mostly requires testing by a lot of people who own the necessary hardware/software combinations.
I have also created a stub page for 3D troubleshooting and known problems, and I want to rely mostly on users to help filling that with useful information. After all, the kind of trouble I run into usually has to do with me breaking the driver during development, which is on a rather different level from the kind of trouble users run into.
Now my main goal is actually to improve the way we present ourselves to potential new developers. This is also quite tricky because after having been immersed in this stuff for so long, it is hard for me to judge what kind of information and documentation would be most helpful to newcomers. If you have a concrete question, or set of questions, I'd be curious to know, and I may end up blogging on it, or documenting it on the Wiki, or documenting it in the source code. However, please give me more than a muddy "this stuff is complicated, can you hold my hand through it?", because yes, unfortunately, this stuff is complicated.
In my experience, the first step to understanding the 3D driver is to understand OpenGL. After all, the purpose of the 3D driver is ultimately to implement an API - and you just won't understand the driver if you don't understand that API. So if you're curious about 3D driver development, hack on little OpenGL samples first, or even at the same time! It doesn't have to be a large game engine; the samples that come with Mesa may be more than enough as a starting point. Then maybe you want to step through your program and what the driver does in a debugger, or experiment with other little ways that can help you understand the driver.
I hope that we will be able to welcome you into the fold one day.
So I've spent some time to refactor the documentation on the DRI Wiki, creating a new, cleanup up and up-to-date portal page in the process. However, this refactoring is not complete, and the amount of information is not yet entirely satisfying. The real test now is whether this can be more than a one-shot effort by me, so this is a shout out to others - particularly to the users who have been asking for this - to help improve the documentation.
Partly, this can be done in very mundane ways by normal users. The application support matrix needs to be updated for Mesa 7.6 (there have been a lot of significant improvements), which mostly requires testing by a lot of people who own the necessary hardware/software combinations.
I have also created a stub page for 3D troubleshooting and known problems, and I want to rely mostly on users to help filling that with useful information. After all, the kind of trouble I run into usually has to do with me breaking the driver during development, which is on a rather different level from the kind of trouble users run into.
Now my main goal is actually to improve the way we present ourselves to potential new developers. This is also quite tricky because after having been immersed in this stuff for so long, it is hard for me to judge what kind of information and documentation would be most helpful to newcomers. If you have a concrete question, or set of questions, I'd be curious to know, and I may end up blogging on it, or documenting it on the Wiki, or documenting it in the source code. However, please give me more than a muddy "this stuff is complicated, can you hold my hand through it?", because yes, unfortunately, this stuff is complicated.
In my experience, the first step to understanding the 3D driver is to understand OpenGL. After all, the purpose of the 3D driver is ultimately to implement an API - and you just won't understand the driver if you don't understand that API. So if you're curious about 3D driver development, hack on little OpenGL samples first, or even at the same time! It doesn't have to be a large game engine; the samples that come with Mesa may be more than enough as a starting point. Then maybe you want to step through your program and what the driver does in a debugger, or experiment with other little ways that can help you understand the driver.
I hope that we will be able to welcome you into the fold one day.
Montag, Oktober 12, 2009
Math reading #1: Volume bounds for lattice polytopes with interior lattice points
So I got a bit into the mood of writing down mathematical stuff for the web. Since reading and trying to understand stuff is a good exercise anyway, I decided to start an experiment in which I will attempt to digest and explain some mathematical paper, or chapter from a book, or something similar on a semi-regular basis.
I hope that these digests will be useful to other people, but in the end their real purpose is to force myself to organize my own thoughts a bit better - and perhaps to serve as an archive for myself where I can look things up in the future.
Without further blah-blah, here's the first installment: Volume bounds for lattice polytopes with interior lattice points
I hope that these digests will be useful to other people, but in the end their real purpose is to force myself to organize my own thoughts a bit better - and perhaps to serve as an archive for myself where I can look things up in the future.
Without further blah-blah, here's the first installment: Volume bounds for lattice polytopes with interior lattice points
Samstag, Oktober 03, 2009
Ehrhart polynomials and integer points in polytopes
It's about time for me to write about something mathematical on this blog. I used the opportunity to experiment with MathML and SVG. Unfortunately, the state of these technologies is rather horrible, which is why I can't write the actual entry in the blog itself. HTML 5 promises to improve things, but it's not quite there yet.
So here is a link to my text on Ehrhart polynomials.
There's a little bit of backstory here which I should probably mention. I was reading up on Ehrhart polynomials a while ago, and in particular I was looking for a proof of their existence. Unfortunately, the proofs I found immediately by perusing literature used rather abgewandte Mathematik, which made me sad. So, in a moment of the kind of hubris which is necessary to do these kinds of things, I decided that I could find an elementary proof on my own. I succeeded, and I thought to myself, "Hey, that proof is actually rather simple. I've been looking for something mathematical to write up on my blog, let's just use this."
So I started, and I had this goal in mind that I could explain my proof in a way that is understandable to ordinary laypeople. In the process, I had to admit to myself that the proof is probably not that simple.
You see, I am not writing for the kind of people who are uninterested in mathematics - that would be futile - but I do want my writing to be interesting and useful for other students of mathematics and interested laypeople. Sometimes, I like to try to write a text where my yardstick is, "Would I have been able to follow and appreciate this text at the beginning of my university studies?" Of course it is not always feasible to write texts like that, and it is actually incredibly hard to tell whether I achieve this goal because I have mostly forgotten who I was five years ago. Trying to see things from that older perspective is not easy.
I do hope that I have succeeded, and while the MathML was annoying to write, it was ultimately enjoyable because I could touch a large number of ideas and areas that are relevant to my daily work.
In the future, I will probably experiment with ASCIIMathML, which I discovered a bit too late. It appears to offer a reasonable solution to the verbosity of MathML.
So here is a link to my text on Ehrhart polynomials.
There's a little bit of backstory here which I should probably mention. I was reading up on Ehrhart polynomials a while ago, and in particular I was looking for a proof of their existence. Unfortunately, the proofs I found immediately by perusing literature used rather abgewandte Mathematik, which made me sad. So, in a moment of the kind of hubris which is necessary to do these kinds of things, I decided that I could find an elementary proof on my own. I succeeded, and I thought to myself, "Hey, that proof is actually rather simple. I've been looking for something mathematical to write up on my blog, let's just use this."
So I started, and I had this goal in mind that I could explain my proof in a way that is understandable to ordinary laypeople. In the process, I had to admit to myself that the proof is probably not that simple.
You see, I am not writing for the kind of people who are uninterested in mathematics - that would be futile - but I do want my writing to be interesting and useful for other students of mathematics and interested laypeople. Sometimes, I like to try to write a text where my yardstick is, "Would I have been able to follow and appreciate this text at the beginning of my university studies?" Of course it is not always feasible to write texts like that, and it is actually incredibly hard to tell whether I achieve this goal because I have mostly forgotten who I was five years ago. Trying to see things from that older perspective is not easy.
I do hope that I have succeeded, and while the MathML was annoying to write, it was ultimately enjoyable because I could touch a large number of ideas and areas that are relevant to my daily work.
In the future, I will probably experiment with ASCIIMathML, which I discovered a bit too late. It appears to offer a reasonable solution to the verbosity of MathML.