The Factual Error exists because this scene is happening in 1944 has defined moments earlier by an in picture subtitle and the picture above was taken in 1947. Time travel maybe?

While trying to update a bogus BIOS on a computer I bought a couple years back I was sent into a yak shaving trip that resembles that of a shaman looking the right set of herbs for a voodoo session that will heal the wounded.

The objective was to update the BIOS of an ACER Aspire One 110. These models came with a faulty BIOS that does not recognize the battery and will not allow it to charge, rendering the netbook a very small desktop computer has it needs to be connected to a power source to work.

The machine I bought came preloaded with some flavor of Linux that I got rid off many moons ago and ACER only provides BIOS updates that can be loaded from a Windows environment, thus not supporting users after the sale is made. So the first step on my yak shaving trip was to get Windows to run on this machine. That meant getting some flavor of Windows into an USB drive.

Windows on a USB is not an easy task, specially if you do not have any Windows machine and for me it was actually impossible. I tried everything from dd to unetbootin and some more. Has it stands one cannot get an USB drive to boot an ISO 9660 image of Windows using OS X.

I had to load VirtualBox with XP, install FlashBoot and load the USB drive with XP because Windows 7 would not fit.

With the USB drive ready I had to convince my Aspire One to take this evil medice, it would not be an easy task. First FlashBoot gives you a two step install process, you start with a free version of fdisk and then proceed with the install. On my first run of the installation the setup was complaining about drive size, Google told me I had to downgrade the system from FAT32 to either FAT16 or NTFS. What is this, the XIX century? Ok, start fdisk without FAT32 support and create a primary partition.

Second time with setup and it complains about not having a “place for a swap file”. And I’m thinking is this really worth? Thankfully my first attempt at solving this was actually the solution, entered fdisk once again and created another extended partition besides the primary partition.

After that Windows XP is copied to the extended partition and once the machine reboots you are presented with the actual installation of XP. Of course this was not without it’s bumps, but I won’t put it on Windows XP for this copy came from devious origins. A few files skipped and reboot.

Updating the BIOS itself did not come without it’s own hops and the network along with other devices were unavailable under XP. USB drive to transfer the required files did the work and thankfully the ACER provided programs to update the BIOS did work on the first run.

Now I can go reinstall Linux again.

Many thanks to haochen for the tumblr export tool[2]. Neat dude.

This is pretty neat because if you have someone or something abusing your app you can add their ip or subnet to the file and forget about them.

The name of the service, DoS, stands for Denial of Service and has the Google Team put it “It is designed for quantitative abuse prevention … only.”

The way it works now you have to identify who’s attacking your app, add them to the dos.yaml file, update the blacklist and wait for the changes to reach the app.

Owners of an app being targeted by a DoS only find out about the attack when the app starts to fail or after a good chunk of resources have already been consumed.

In the apps dashboard the admins can see a list of the top 25 ips responsible for the most requests since the beginning of the current hour.

If an app is left overnight without supervision and an attack occurs it’s in the hands of the developers and admin to mitigate the abuse. Developers can implement methods of analyzing the requests and firing alarms that can prevent major abuse and gain some time to alert admins and have the blacklist updated.

The App Engine team does seem to keep a look at how the apps are doing and will put blocks in place if they notice anything out of the ordinary. How this works is a bit of a mystery has would be expected but it’s a fact that they won’t even contact the owners of the app about the blocks.

The DoS service seems to be pretty low-level and the rules are probably loaded for in each app instance at creation time. This makes it somewhat hard to play with the configuration and results in a text-based yaml file.

If a developer is given the task to implement a system that would take note of each users request and fire alarms and systems that prevent major abuse it would be really nice if these developers could also programmatically access the options of the DoS Protection service to add and remove rules.

The apps being able to protect themselves and able to adapt to the attacks would be a good addiction to the toolbox of GAE.

This was copied from Manu.

With the help of Wireshark I managed to see the information is pooled from an HTTP GET request to mirror.toolbar.netcraft.com.

curl http://mirror.toolbar.netcraft.com/check_url/http://medecau.com

Not the case for Superfeedr. As of today the only thing in Superfeedr’s blog front page pointing for a feed is a sidebar link to an Atom feed and feed discovery is non-existent. Has this wasn’t enough you really have to be willing to read the sidebar to see this link. And the feed is not even valid XML.

I mean, Superfeedr works with feeds for a living, they should at least make an effort.

This should happen.

I have been playing a few games on OMGPOP and most of the time I have a second tab for Grooveshark.

Problem is music is played like a radio and there is no connection to what is happening in the game.

It would be nice if the guys at iminlikewithyou would hook with Groveshark and open “radio” channels for each game.

Even better would be to change the music when bosses came into the level or a poll game came to the end.

After some time thinking about it I finally decided to move away from WordPress to tumblr.

Looking around the web I could not find anything in Python that would do the job, so I wrote a little something to help me on that.

Enter totumblr.py, a python script to easily transfer your old posts from WordPress to tumblr.

totumblr.py makes use of pyblog.py and tumblr.py to move the posts in your old blog to tumblr. totumblr.py is available has a gist in github. I invite you to use and modify it, and if you find a bug please let me know.