On 2016-08-01 a law came into force, allowing customers in Germany to use any internet router they want, as long as it meets the providers requirements. On that day I decided to buy a Fritz!Box 6490, for three reasons: 1. I wanted to replace the crappy router my provider gave me (which didn't even have WiFi) 2. besides its routing function, it is able to stream up to 4 concurrent TV programs into the Network 3. it was the only device that fulfilled the above-mentioned requirements. I wanted to setup [Tvheadend](https://tvheadend.org/) to act as an intermediate between clients and the router, because it would allow me to have an EPG, stream the same program to 2 clients, while only occupying a single stream of the Fritz!Box, and allow me to record TV shows. As you can see from the date of this post, it took me one year to figure out how to do it.


This blag post describes how to use a [Raspberry Pi](http://amzn.to/2wjICvo) to remotely "press" and potentially "hold" the power button on a PC. This is my first non-trivial (still pretty-trivial) hardware-related project. So don't expect anything too fancy. Would you like to know more?


I have a [problem with my Nokia 3](https://android.stackexchange.com/q/180669/197410) and decided to try the [Nokia customer support](https://www.nokia.com/en_int/phones/support) while doing some [recreational math](https://projecteuler.net/), that's what a great Saturday is to me. Don't judge. During the conversation, I grew very suspicious that I was talking to a chat bot. This was the first time for me hearing about chat robots being deployed in support by a [respectable company like Nokia](https://en.wikipedia.org/wiki/Nokia#Controversies), so I decided to make sure I wasn't being paranoid. Time for a Turing test! I guess you can never really know, but I am quite convinced they lie to me and are actually robots. You can try it yourself, just go to [their support site](https://www.nokia.com/en_int/phones/support) and start a chat. Would you like to know more?


My lamenting will be about the overall way in which device encryption is implemented in Android. This is mostly a collection of links where you can find out more about how it ~~worked across the recent versions~~ all went south. This article strictly expresses my own, badly informed opinion and you should check all the provided references carefully before forming your own. read();


We decided to use the javascript markdown engine [showdown](https://github.com/showdownjs/showdown) for the blawg, and [$\KaTeX$](https://khan.github.io/KaTeX/) for rendering latex. If you think that's a good way to go: Here is how to do it with wordpress.


I like to have all "stuff that I need to do something with" collected at one place. For the virtual world, this place is my email inbox. To remind me of things, I send myself an email and can be sure that I'll process it at a later point in time for example. Having another place with the result of RSS feeds always annoyed me a bit. As soon as I realized that I find this annoying, a solution was easy to implement of course:
aptitude install rss2email
r2e new your@email.address
r2e add blag https://blag.nullteilerfrei.de/feed/
and a cronjob, executed every few minutes:
*/10 * * * * r2e run
will lead to all our fine blag posts ending up as email notifications.


Github has a history of not giving a frack what their users want12. For example, a few developer friends of mine were reluctant to click any links in their notifications-page, since after they clicked the link, the notification was marked as read and you might lose track of it, if you just close the browser tab3. So David Badura4 and I decided to fix this problem by writing a browser extension. The result can be found on Github: https://github.com/larsborn/GithubToDos. After installation (also possible in Opera btw, the best browser there is), the extension injects an "Add ToDo" button on every issue page and pull request. When clicking, it, the URL gets saved to the local storage of the browser5. The list of all URLs added like this can then be access through a new button in the header toolbar of github. You can clone the project from github and add it as an "Unpacked extension" or just head to the Github ToDos on the Chrome Store and just install it from there. Pull requests are welcome, open an issue, if you find a bug, open source yadayada.
  1. https://github.com/dear-github/dear-github []
  2. https://github.com/isaacs/github/issues []
  3. on a side note, tellmewhenitcloses.com is pretty handy to avoid too many notifications in the first place []
  4. https://github.com/DavidBadura []
  5. Using local storage is handy for people that are not very concerned about privacy and just use the cloud synchronization feature of their browser: the content of their ToDo list will then also just be synced to all their devices. []


I wanted to crop out a specific rectangle in a few dozen scanned documents with ImageMagick like this:
convert -crop 1600x1880+100+420 image.nrm.png cropped.png
Scanning often results in a tiny skew, which would lead to a slightly different rectangle location on every image. It seems that a common preprocessing step when doing OCR is to automatically correct this skew. The Python toolset ocropy1 for example contains a tool to do this: ./ocropus-nlbin image.jpg creates the file image.nrm.png which is optimized for OTR and has corrected skew.
  1. https://github.com/tmbdev/ocropy []


Obviously, my Bank does not provide a REST API to download the transactions happening on my accounts. After I asked for "machine parseable" data, they told me that I can download CSV files. Awesome! So I wrote a parser and lived happily ever after. Except that they change their CSV format without notice every few months and at some point they started to mix different encodings in the same file. So I lived unhapply, regularly fixing the script reading the CSV file. What did not change for around 10 years now are their banking statements. And this also holds for the PDFs you have to download, if you want to avoid getting them via snail mail (and paying for the postage of course). I decided to parse the PDFs instead and this went pretty well for may years... up to recently, when something changed (it may have been a software update of the parser I use or something on their side). Do you want to know more?


We have updated the layout, also the posts are now markdown-rendered on the client side. Oh, and we switched to [$\KaTeX$](https://khan.github.io/KaTeX/) for math rendering. Since this breaks compatibility with the old hairball of plugins we had, it will take some time before all the old entries are fixed, but we are working on it.


My Windows desktop computer is set to turn the screen off after one minute of inactivity. It's usually a good idea. Except when I am reading something on screen. Then it is terribly annoying. Most media players can prevent Windows from turning off the screen by calling [SetThreadExecutionState](https://msdn.microsoft.com/en-us/library/aa373208(VS.85).aspx). Well. So can I:
import ctypes
ctypes.windll.kernel32.SetThreadExecutionState(0x80000002)
input('screen stays on. press any key to exit.')
ctypes.windll.kernel32.SetThreadExecutionState(0x80000000)
Pretty neat, huh? Python really can do all things.


I finally decided to set up a real router in front of my router. The main use cases, I wanted to cover were the following: * nice local domain names: Since I run a FritzBox (which is a pretty common plastic router in Germany), all local host names (sometimes!) get suffixed by .fritz.box. This domain cannot be configured to .something.awesome and the whole setup is quite in-transparent, that I could not figure out, in which cases, the suffix is mandatory when resolving host names. * play around with [snort](https://www.snort.org/). * bandwidth monitoring: to replace the current workflow of randomly killing machines, when the internet connection is slow and one needs bandwidth to do something important™. In this blog post I will cover nothing of this. Instead, I'll explain, how I set up a RaspberryPi to enable me to do all the above in the future. Do you want to know more?


Es [gibt einige Probleme, im WiFi der Deutschen Bahn im ICE ein VPN zu verwenden](https://community.bahn.de/questions/1197692-nutzung-wifionice-ssl-vpn), und das ist natürlich ein Problem, weil es sich dabei um ein offenes und unverschlüsseltes Netz handelt. Die Bahn erklärt das Problem dadurch, dass Mobilfunk aktuell nur Pakete durch lässt, die maximal 1500 Byte groß sind. Da noch Platz für deren Header bleiben muss, sollte man die MTU auf 1440 setzen. Unter Windows macht man das wie folgt: Click!