Those of you who want a shorter introduction: the presentation is available on issuu.com, as a pdf or as a quicktime video with transitions.

In the paper I describe what HTTP Live Streaming is, fairly technically, and why it is a great step forward. I compare the new technology with the industry incumbent RTP/RTSP protocol and Adobe’s proprietary Flash Media Streaming, popularized by YouTube.com.

As a teaser: here’s a graph of the client CPU Load on a MacBook when viewing the same video clip:

But I have to warn you, the paper is probably a bit too technical for the casual reader.

The final case he makes proves it all.

If you wonder how all of this could have happened, watch the video below. It might go a bit fast sometimes, but I found it really interesting and a nice summary.

The Crisis of Credit Visualized from Jonathan Jarvis on Vimeo.

Best viewed in full screen in HD (on Vimeo).

Today I booted my bloody windows box and 3 partitions were missing, and I had a pretty bad Virus infection. Booting windows in safe mode i could remove the entries made to my system from the trojan and scan the systems folders for more trojans with ClamWin open source anti-virus software.

Problem 2 was a bit easier as I had already dealt with ‘deleted’ partitions before. Booting from the floppy and searching for the missing partitions was not too hard.

After two hours of misery, the machine is working again. I swear this is the last Windows box I’ll ever buy. I’ll probably replace it with a MacMini when the new version comes out.

The Wikipedia entry explains why this “Student Syndrome” is a good thing :

The student syndrome is defended by a layman’s understanding of the functioning of the human memory, most notably the concept that a person’s short-term memory fades over time, and thus studying at the last possible moment leads to better results than studying early because more of the material will be remembered during the exam.
Wikipedia

I wrote a small tool to extract laptime data from the race reports so it can be easily analyzed in MS Excel. I wrote it as a Sinatra webapp. Sinatra is a Ruby DSL for writing small webapps with a minimum amount of code. This tool was written in 35 lines of code.

It’s open source, so you can view (or fork) the sourcecode on github.

F1Time managers: You can simply copy & paste the html code of your race report HERE, and it will generate a table that you can copy/paste into Excel. (to get the html code from the race report: right click and click “View page source”)

But even Delhaize attracted quite a lot of interest from fellow students who are looking for more traditional management work. Apparently you can apply for the position of store manager after an internship of only 18 months. This job is way to narrow minded for me though.

I found it a bit disappointing that there were so many tax & auditing firms present, yet so little innovation & technology driven businesses. The companies I would love to work for would be Apple, IDEO, Intel and I guess Accenture too. If I haven’t founded my own start-up by then .

At any rate, I’ve put my resume online for anyone who is interested.

(click on the full screen button on the lower right of the movie player)

If this is how future user interfaces will look like, i want one!

Only the IDE drive crashed so my important files were safe, but I needed to remount the raid array. It took me quite a while before I got that right, so I’ll share my knowledge here for you. Here are the commands needed to reconstruct / remount a raid array with a Gentoo live cd:

load the raid module and lvm driver if you need to:

# modprobe raid5
# modprobe dm-mod

Find the UUID’s of the devices

( you can skip this step if you have a backup of your /etc/mdadm.conf file )

# mdadm --examine --scan /dev/sda1

Where /dev/sda1 points to one of the partitions used for the raid array. The output looks like this on my box:

ARRAY /dev/md1 level=raid5 num-devices=4 UUID=94366f38:ccdadda0:06f85cb1:f1263d52

copy it to /etc/mdadm.conf

# mdadm --examine --scan /dev/sda1 >> /etc/mdadm.conf

Now we need to add a list of devices to mdadm.conf of where to look for member disks:

(manually add this to mdadm.conf, the devices i had to add where /dev/sda1 to /dev/sdd1)

DEVICE /dev/sd[abcd]1

At the end that leaves us with this in our mdadm.conf file:

DEVICE /dev/sd[abcd]1
ARRAY /dev/md1 level=raid5 num-devices=4 UUID=94366f38:ccdadda0:06f85cb1:f1263d52

Assemble the raid array

We make the mirror device node:

# mknod /dev/md1 b 9 1

Assemble the array:

# mdadm --assemble --scan

You can check /proc/mdstat to see if the array is online:

# cat /proc/mdstat
Personalities : [raid0] [raid1] [raid6] [raid5] [raid4] [raid10]
md1 : active raid5 sda1[0] sdd1[3] sdc1[2] sdb1[1]
937705728 blocks level 5, 64k chunk, algorithm 2 [4/4] [UUUU]

Activate the logical volumes

# vgscan

Scans your disk and should show you the Volume Group that was on the raid array and add it to the lvm database

# vgchange -ay

This should activate the volumes and auto create the nodes in /dev

Mount the filesystems

I just needed to mount the filesystems just like I would mount a regular one:

mount --type ext3 /dev/vg2/var /mnt/gentoo/var

Hope this helps someone!