weblog by Urs Gehrig

A weblog about libre software, law, technology, politics and the like.
Academic, Android, Apache, Apple, Art, Best Of, Biotech, Civil Society, Content Management, Cooking, Copyright, Creative Commons, Crosspost, Culture, Database, Deutsch, DRM, Economy, Education, Event, Gadget, General, Geodata, Government, Health, Howto, Humor, Innovation, Intellectual Property, Java, Language, LaTeX, Law, Linux, Media, Moblog, Mozilla, Music, Office, Open Content, Open Source, P2P, PHP, Podcast, Politics, Privacy, Projects, Random Thought, Rant, Science, Search, Social Network, Software, Sport, Talks, Technology, Technology Transfer, Travelling, Weblog, Wiki, Wireless and Mobile, XML

17. November 2005

Modifications on the $100 Laptop
@ 22:25:16

Last Monday I read the first time about the $100 Laptop of the MIT to be built for the One Laptop per Child (OLPC) initiative [1]. There will be no shop that is going to sell that Laptop, as I read on the introductional page. The Laptop roughly will be
a Linux-based, full-color, full-screen laptop that will use innovative power (including wind-up) and will be able to do most everything except store huge amounts of data. This rugged laptop will be WiFi-enabled and have USB ports galore. Its current specifications are: 500MHz, 1GB, 1 Megapixel.
So far quite impressive, if you keep $100 in mind. Nevertheless I would like to add the following;
  • To add a solar panel on the top of the Laptop case [2].
  • To add a foil keyboard to make it more robust [3].
  • To add more fixed Memory (>1GB).
  • To set up a Hardware-Kit to let students build their own Laptop.
  • To find a way to license the schematics, Hardware-Kit etc. under a non-commercial license.
I am not yet sure about what an Open Hardware License would look like and even if it even is nescessary. As said, it is just a random thought. But I can imagine that it would push the innovation in the soft- and hardware industries, even though I tend for a non-commercial license. The idea behind is that companies should not re-produce the hardware and sell it at exorbitant prices. The Laptop should be available for approx. $100. So if they are innovative, they produce it for less.


Comments (3) Permalink

15. November 2005

Post-Grokster world with a closed P2P eDonkey
@ 22:12:37

Sam Yagan, President of MetaMachine, Inc., explains his companies "capitulation" to the RIAA and the conversion of eDonkey to a "closed" P2P system [1,2]. Regarding the creation of software, that can be used for legal and illegal purposes Sam Yagan says:
Because the Grokster standard requires divining a company's "intent," the decision was essentially a call to litigate. This is critical because most startup companies just don't have very much money. Whereas I could have managed to pay for a summary judgment hearing under Betamax, I simply couldn't afford the protracted litigation needed to prove my case in court under Grokster. Without that financial ability, exiting the business was our only option despite my confidence that we never induced infringement and that we would have prevailed under the Grokster standard.
Later in the testimonial he states:
Imagine if for the last ten years we had been able to convert just 1% of the estimated tens of billions of shared files into paid downloads. There is a market solution to be found - it may well be one that fits in to the business model of the incumbent entertainment industry, but it also may not.
If all those P2P software companies die, there will be not less file-shareing but less control by the record industry. As experianced earlier, users will simply turn to other services with similar offerings. What might help is to turn this P2P software to be used for something "good". Therefore some random thoughts:
  • What about a P2P network that only accepts open content licensed files?
  • And what about heading towards that 1% conversion, if the price per song - if we take music as an example - is considerably lower than e.g. in the iTunes Music Store?
  • What if the price is not fixed but depending on supply and demand?
  • What if some start an Open Music Stock Exchange to produce appropriate revenue?
I appreciate your response!


Comments (0) Permalink

06. October 2005

QR coded blog entries
@ 00:23:26

Innovations is - if it works, but you do not have the technology to prove it. The blogpost title, the blogpost date, the URI to each blogpost and some copyright ("CC") information of this weblog are QR coded. Actually the images have not been decoded yet, since the writer of that code has no QR code scanner nor has he a mobile phone with an appropriate software to decode those codes.

But anyway. QR codes are kind of twodimensional barcodes [1].
While conventional bar codes are capable of storing a maximum of approximately 20 digits, QR Code is capable of handling several dozen to several hundred times more information.
20 digits are not enough to encode a typical blog title, a link, a date string and some copyright information. As an example, the QR coded string for this blog post is: "QR coded blog entries, 06.10.2005 00:23:26,, CC-by-sa_2.5". The QR code library behind my implementation is written by Y. Swetake [4].

If it makes sense, to link from a QR code to a blog post, is not yet plain clear. But I could imagine a simpler, maybe automated navigation for mobile phones:
  • Reading a blog feed (e.g. qr.xml or qr img link in the RSS feed) with QR codes
  • The phone decodes the images and displays the title to the blog entries
  • A single click on the auto-linked title is enough to access the full blog post
At least in Japan there are various applications that support QR codes [2]; e-commerce e.g. is a big issue [3]. Is anyone out there with an opensource QR code decoder? ;)

Update: The ETH Zürich has a project running called "Visual Code Recognition for Camera-Equipped Mobile Phones" [5]. They also provide so called Recognizer software for Symbian OS based mobile phones.


Comments (5) Permalink

07. September 2005

Get mapping coordinates from an address
@ 14:41:57

randy [1] found a way to get the longitude and latidue out of an address like "1600 Amphitheatre Parkway, Mountain View CA 94043". He uses an XML-RPC call to and gets an array of mapping coordinates in return. Awesome.

As the current petrol prices are rising, I would like to see an SMS service, where you can send something like "Lorraine, Bern" to a certain number and in return you would get an MMS with a map with a red outlined track to the next petrol station with the lowest prices on petrol. ;)


Comments (0) Permalink

01. September 2005

A Plaze to be with a signed Java Applet?
@ 09:23:08

chregu [1] recently bookmarked Codeplaze [2], a code repository for the Plazes site. Instead of using their Javascript Badge I went for the XML-RPC variant to link to the current Plaze of being. Unfortunately there is no way to run a Launcher on some (restricted) systems, and therefore one turns out to be plazeless sometimes.
You haven't used the Launcher to log on, therefore you are "plazeless". You cannot search for people or Plazes by radius.
Apart from binaries there is also a Python launcher [3]. That is ok, if Python works on your system. So I was looking around for a signed Java Applet which gets access to the machines MAC address [4]. I am not sure, if at all, a launcher is programmable as signed Java Applet [5], but it would be fun to have a platform independent Launcher anyway. Maybe there will even be a non-Java Plazes Launcher Firefox extension one day, but that is just a guess.


Comments (3) Permalink

30. August 2005

Freetag - a tagging and folksonomy system
@ 23:08:58

Or more detailed:
Freetag [1] is a module that implements a simple, fairly robust beginning of a tagging and folksonomy system. It works with PHP4 and MySQL 3.23 so far, and i imagine that if it gets popular, it should be easy to port to additional databases and/or languages.
Maybe a way to build up its own tagging repository. Acutally I haven't tried, but according to the docs it is promising. As to feed the database, one could use his own tags or category names first.

There are probably at least two ways to tag a blog post. Either tags selected by hand at the time of writing or as a self-tagging system (algorithm), that takes e.g. the content of a blog post, analyzes it and applys the tags to the post.

Some more thoughts on tagging and retagging at [2].


Comments (0) Permalink

29. August 2005

Blog backend with recommended tags per post
@ 12:57:59

I like the support of the Firefox extension for' interface with clickable recommended, own and popular tags. This is something I would like to see in a backend/admin of a blog software. So, the code under GPL is there [1]. I am wondering, if the javascript could be used more or less instantly. The are various needs. E.g. as recommendations per post for Technorati, Flickr or tags.


Comments (2) Permalink

18. February 2005

Google search plugin based on XMLHttpRequest?
@ 00:03:07

Bertrand today told me about Google Suggest (Beta) [1] and just right now I was wondering if out there someone already implemented a search plugin for Firefox that makes use of XMLHttpRequest [2]. Chregu [3] called it LiveSearch and that is what drives this blog's search engine too ;)


Comments (0) Permalink

31. December 2004

CC's second anniversary
@ 17:51:21

Creative Commons turns 2.0 [1]. Congratulations! There will be a party January 6th 2005 in San Francisco. We definitely should celebrate locally too. I will have a look where we could meet in Bern - what do you think about the Kornhaus Bar Turnhalle, Speichergasse 4, Bern [2]?


QR Code for your mobile phone.
Comments (5) Permalink

30. December 2004

JPEG image problem with PHP GD
@ 23:15:04

I wrote a PHP script based on imagecopyresampled() to resize images coming from mobile phones. Some phone models seem to produce corrupt JPEG files. As on calling imagecreatefromjpeg() a message gets produced looking like: "Corrupt JPEG data: 164 extraneous bytes before marker 0xd9".

Imagemagick handles such problems in some - to me unknown - way and resizes the image properly, regardless of the problem pointed out. But there is currently only GD available and with that few lines of code I get a black resized thumbnail.

The message pointed out above gets printed out when calling imagecreatefromjpeg(). A random thought was to rewrite a function from scratch in PHP that creates a valid image resource [1] in spite of the corrupt JPEG format. But as we have 23p.m. now, I am just too tired now ;) So I am stuck. A corrupt image can be found here [2].


Update: Problem-walk-around, as I got imagemagick support on the server ;)

Comments (2) Permalink

01. November 2004

Fundamental mistake to let people pay for Fahrenheit 9/11?
@ 22:45:46

Today I read that one can watch Michael Moore's Fahrenheit 9/11 for $9.95 on a pay-per-view basis:
Voters in the US will be able to log on to the net tonight and download Michael Moore's controversial documentary, Fahrenheit 9/11, for some topical election-time viewing. Net movie provider CinemaNow is making the film available on a pay-per-view basis as of this evening, The BBC reports.
I have seen the film today on ORF and would assume, that it is a mistake to let people pay for that film. My arguments are towards providing that piece of master-work under a license like the Attribution-NoDerivs-NonCommercial of Creative Commons to bring as many as possible before the movie players ;) Finally, those having seen that film, would possibly play an important role in the upcoming elections; especially those in the swing states. Just an idea.


Update: I obviously did not have all facts while writing the above as you can read in a post on DRM News [2].


Comments (1) Permalink

26. August 2004

Book on Demand
@ 10:33:17

I am dropping this link here as a reminder for myself: Book on Demand [1]. If this is the way of straighter bookmaking for authors, why not. There is another service in Germany as far as I know, called [2].


Comments (0) Permalink

24. August 2004

Ventilwächter (Tire Deflator) für volle Staatskassen
@ 23:39:58

Da les' ich doch letztlich von offenen Rechnungen von Parksündern, insbesondere solchen, die nicht hier wohnen. Dabei ist mir vor kurzem ein Produkt unter die Augen gekommen, das offenbar erfolgreich in Deutschland im Falle von säumigen Zahlern eingesetzt wird: der Ventilwächter [1,2,3].

Tire Deflator

Das Teil wird von den Behörden - oder derem verlängertem Arm - an das Ventil eines Autorads montiert und mit einem Schlüssel abgeschlossen. Dabei wird der Autolenker mittels einer Hinweistafel auf die pneumatische Wegfahrsperre hingewiesen. Denn, sollte das Auto ungeachtet dessen in Bewegung gesetzt werden, so geht besagtem Reifen innert ca. 500m die Luft aus. In der Schweiz wurde das Teil bisher noch nicht gesichtet.

Für Jungunternehmer und solche, die es werden wollen, wäre da noch zu reservieren ("The domain name is available!"), der Import zu regeln und ein Businessplan zu erstellen ;-)

Nun, eigentlich müsste man diese pneumatische Wegfahrsperre anders konzipieren. Es sollte möglich sein, bspw. eine Parkbusse direkt via Mobile Phone bezahlen zu können. Da der Betrag einigermassen gross sein wird, so ca. im Bereich von CHF 40.-, könnte allenfalls eine Kombination mit EasyPay [4] eine Lösung bieten. So im Stile von
Bezahlen Sie die Parkbusse mittels SMS bspw.: "123456789 an Nr. 999"
123456789 würde dem EasyPay Code entsprechen. Damit wäre zwar die Zahlung gelöst, nicht aber das Lösen der Wegfahrsperre. Da braucht es bekanntlich einen Schlüssel. Da dieser kaum vom Himmel fallen wird, braucht es eine andere Lösung. Entweder die Securitas kommt in den nächsten 15 Minuten dahergeeilt, oder die Wegfahrsperre sei bspw. durch ein elektronisches Schloss zu sichern, das sich vom Zahlenden mittels der durch die Antwort-SMS erhaltene Entsperrinformation öffnen lasse: RFID, Infrarot, etc. etc. Da besteht noch einiges an Innovationsspielraum ;-)


QR Code for your mobile phone.
Comments (0) Permalink

29. May 2004

Swiss Post buys search portal
@ 01:53:17

My favourite search engine for domestic searches,, has been bought by the Swiss Post [1,2]. This is an interesting deal of a company dealing with letters and packets mainly. The Swiss Post itself has its own online portals providing services like e-banking, ticketing, mail and sms, e-payment etc. etc. To integrate a search portal seems like logical if you look at it now, but who would have guessed that ;-)

I am quite interested in how the Swiss Post intends to enhance the services provided by You might argue, that a typical Post company should not burn its hands with e-business, then you should keep the following in mind; even if the citizens tend to travel or move around more than they did ten or twenty years ago, they mostly are at residents somewhere. Traditional services as delivering snail-mail can be combined with online services. The Post company has probably the most actual data regarding addresses. To give an example; the first thing one does after having moved to a new location, is letting the snail-mail get forwarded to the new address. This can be done by signing a contract at the Postoffice. So, the Swiss Post does know beforehand of that change; probably earlier than the Bank, the assurance company or the telecom company. Where does a search engine or search technology come in? Well, you better ask Swiss Post ;-)

But already combined search technologies with the lookup of telephone numbers resp. addresses. One of their latest feature is the "related websites" ("Websites zum Eintrag"):

Since I am not an economist, I can not say exactly, how valuable "early bird" false-free address-information in combination with online services is. But I would guess, the earlier you get them, the better. Froogle "only" does allow smart shopping - in other words - seeking products on demand. But if the Swiss Post creates a site, where one can set up his or her own profile, the Swiss Post should be able to create a "portfolio" of possibly interesting products based on some search magic matching to that profile. Finally, Postfinance offers a comfortable way to pay the bill. Just a random thought...

Oh, and congratulations to Berhard.


Comments (0) Permalink

16. March 2004

Real-time preview of Blog Comment
@ 00:38:15

Chris J. Davis [1] ported an MT feature to Wordpress; while you type your comment into the textarea, you get a real-time preview. I like that feature, since that simplifies the form by dropping the preview input button.

But instantly I had that other random thought; a javascirpt transforms the entered text into the preview. As long as you do not copy-paste any prewritten text, writing comments usually takes quite a while.
Now, the time argument could eventually be used to prevent comment spam in some way. If the spam is posted automatically, the time used, is probably around zero or marginal compared to handwritten text. Maybe this could be just an additional method. What do you think?


Comments (1) Permalink

08. March 2004

FOAF - Trust and Reputation
@ 01:03:22

Today I found the Trust and Reputation project homepage for FOAF [1]. You can now upload your FOAF file and rate the people you know resp. the ones you have listed in your file. There is a perl script to convert your addressbook to a textfile for an easier conversion to add to the FOAF file [1]. Give it a go and build your own Orkut; at least I could imagine some nice applications. I just had a random thought about a single sign-on management for online services. According to the trust given, my friends could be able or not to post to a certain blog, a calendar, a newsletter or what so ever... Just a thought to reawaken my crosspost idea ;)


Comments (2) Permalink

27. February 2004

[RT] Handling SQLite remote files with PHP
@ 10:38:28

First I have to clarify the [RT]; Bertrand Delacrétaz [1] mentioned at LOTS, that he indicates "random thoughts" in mailinglists as "[RT]". I like that idea, because the title to this blog post does not mean to be a new feature in the PECL sqlite extension but generally a random thought of mine. I started a new project on sourceforge called wikita [2]. It is a common playground for me to set up a simple wiki based on PEAR classes where ever possible. Since I have started the PEAR sqlite DB wrapper earlier, it needed some more testing. Wikita from that point of view is also a testbed for this wrapper.

Just the other day I checked in some code into the CVS of sourceforge. The code is not much more than a hack for now. Anyway. To come to the point of this post: SF does not yet support the sqlite extension. So the title here suggests that there could/should be a way to handle sqlite remote database files. PHP does allow to handle remote files but IMHO the sqlite extension does not. If this is secure or not is to discuss afterwards. PHP5 will generally offer sqlite support. Nevertheless I ask myself if there are any applications that would make sense to have sqlite databases handled remotely. Maybe for portable devices such as mobile phones or handhelds, if they ever offer PHP support (some actually _do_ offer already Python [3]). Since they offer limited storage space, it could be a solution. Contradictions, comments?


Comments (1) Permalink


Beitrge von Dritten:

Nachfolgende Titel verweisen auf von mir gelesene Weblogs.


Blog Content
Blog Comments



Good question, but...
Hi, thank you very...
Unter http://www.s...
Ich weiss mir nich...
ThanQ matthias. Th...
in case you just w...
ich liebe dir, urs...
hi there, sorry i...
Hoi Leo. I haven'...
Do you know the si...


Blog stack:

Bill Humphries
Wendy M. Seltzer
Christian Stocker
Roger Fischer
Sandro Zic
Wez Furlong
Ben Hammersley
George Schlossnagle
Joichi Ito
Lawrence Lessig
Derek Slater
Karl-Friedrich Lenz
John Palfrey
Bernhard A.M. Seefeld
Gregor J. Rothfuss
Rainer Langenhan
Elke Engel
Sebastian Bergmann
Simon Willison
Jeremy Zwaodny
Udo Vetter
Axel A. Horns
Miguel de Icaza
Andreas Halter
Silvan Zurbrügg
Hannes Gassert
Markus Koller

$Date: 2005/11/05 11:14:30 $