Centroid.EU Blog

This is the 1171st article of this blog

June 12th, 2017

I didn't realise I passed 1000 articles here. Even with 10 articles per month for the last 8 years I've surpassed 960 which is what I figured with 8 times 12 times 10. I'm very proud to be blogging still.

The keyword search engine is very valuable tool as well. I use it often to find things I wrote about. One day I'd like to change this to an advanced search where you can use AND, OR or boolean search. But that'll have to hold for now.

It's also worth noting that this blog does not use an SQL backend. I use plain files to store the content. I use a small script in UNIX to add content (articles), and the editor is vi. With all the injection attacks on SQL I felt that what I did was much safer, but granted it doesn't scale on large blogs with 10,000's of articles. For my operation this is still OK.

0 comments

---

Schizophrenia does not 'burden' society

June 8th, 2017

A friend sent me this article. I read the Objective in the abstract and stopped reading at the sentence "Given the immense cost to society"... What costs? I have schizophrenia and I work. I pay my own health insurance, rent and I take my meds. There is no costs on society. The sentence itself is self-justifying, in other words selfish. In fact I'll tell you right now I do not have any "voices from god" telling me to stab someone 3 doors down. That's not what my illness is about. In fact if I were such a danger to society then how come I can do everything anyone else can.

Except, it's harder to find work. Since my diagnosis of schizophrenia I held three jobs. Two of them I was fired from, the third I'm on right now, and I'm hoping to be working in the future at some other job. I spent 7 years on welfare and this was because I could not even convince anyone for a job interview.

Put yourself in my shoes, first everyone ignores you, and then they say you're a burden because you don't have work. Let's get rid of the stigmas, because we're very capable people, and not all of us hear voices. In fact I don't. Even without medication I do not hear voices, sorry to spoil your fanatical view of schizophrenics. The problem is people who freak out when they see someone they identify as a schizophrenic. And perhaps that's a burden to society.

0 comments

---

Donation Time: OpenBSD gets 50 EUR

June 8th, 2017

I have donated 50 EUR to Theo de Raadt. OpenBSD is an operating system for all sorts of computers. Keep up the great work, OpenBSD!

0 comments

---

Registered freifunk-schweinfurt.de

June 8th, 2017

I'm an avid Freifunker and Schweinfurt just got a hood, which means we have a few gateways (tunneling gateways) for our city and the immediate region. Anyone from the Schweinfurt Hood can get a hostname in this domain, just email me, or find me on IRC. The new domain forwards http to the Freifunk Franken Schweinfurt section. Yay!

0 comments

---

Why I think America is greedy, under Trump's presidency

June 1st, 2017

I don't know why Donald Trump thinks America is getting the short end of the stick in trade. Let's look at some facts. America has a monopoly in data. You know..google, microsoft, apple, facebook, twitter, amazon, what have you, those are all american companies. There is not too much competition for these that I know other than perhaps baidu in China and vzkontakte in Russia. So while the rest of the world is locked-in to american data brokers, they are saying they are getting the bad deals. This is greedy! As you may know, data isn't profitable by itself. But massage it with analysis it is very valuable and it is powerful. The trick is to run powerful algorithms over it after it's being collected.

Now, the NSA has access to some of that data and they have their own data. But the NSA isn't for-profit per se. But they run some of the same algorithms that for example google might. It's in their interest and again they have a monopoly of this data in the world. You may remember the spy plane that was damaged in April 2001 as it was collecting data from cell phones in China. That is what's called aggressive data collection.

So while perhaps other areas in trade are lacking overall, the "data" cannot be overlooked. And that's why I think America is greedy.

0 comments

---

Launched first page of http://goldflipper.de

May 30th, 2017

I have launched the first page, more to follow on goldflipper.de. I think it looks great so far. Also check out the reload the colour border is randomly created.

0 comments

---

Played with my wacom a few days ago...

May 26th, 2017

Or... from a different perspective:

Go flipper go!

0 comments

---

Built an aggregator for big files

May 25th, 2017

I had the problem of getting 8 Mbps downloads across the Big Pond (tm) from my VPS psi.virgostar.net. So I constructed a little system with scripts and ssh to aggregate several locations together. Now I get ~30 Mbps.

As you can see this is how it works, each arrow is an ssh connection. BPI is the Big Pond Internet, the area where I cross the ocean. I checked I get routes from Tata, Telia and Cogent, so everything is nicely dispersed and they can't put 2 and 2 together and throttle all my hosts. At uranus I assemble the temporary files into 1 big md5 summed file. If the md5 sum is fine then it deletes the temp files. Below is the script from Uranus which does most the work of chopping up the workload for each host.

#!/bin/sh

if [ $# -ne 1 ] ; then
        exit 0
fi

RET=`ssh psi /home/pipe/pipe.sh $1 || echo NSF`

if [ x$RET == x"NSF" ] ; then
        echo no such file
        exit 1
fi

SIZE=`echo $RET | awk -F: '{print $2}'`
MDSUM=`echo $RET | awk -F: '{print $NF}'`

CHIP=`expr $SIZE / 4096`

date
echo getting 4 chunks of size $CHIP
TMP1=`mktemp 1.XXXXXXX`
TMP2=`mktemp 2.XXXXXXX`
TMP3=`mktemp 3.XXXXXXX`
TMP4=`mktemp 4.XXXXXXX`
BASENAME=`basename $1`

ssh omega /home/pipe/pipe.sh $1 0 $CHIP > $TMP1 &
OFFSET=$CHIP
ssh supercluster /home/pipe/pipe.sh $1 $OFFSET $CHIP > $TMP2 &
OFFSET=`expr $CHIP \* 2`
#OFFSET=`expr $OFFSET + 1`
ssh chi /home/pipe/pipe.sh $1 $OFFSET $CHIP > $TMP3 &
OFFSET=`expr $CHIP \* 3`
#OFFSET=`expr $OFFSET + 1`
ssh psi /home/pipe/pipe.sh $1 $OFFSET 0 > $TMP4

wait %1
wait %2
wait %3

cat $TMP1 $TMP2 $TMP3 $TMP4 > $BASENAME

SUM=`md5 $BASENAME | awk '{print $NF}'`
if [ $SUM != $MDSUM ] ; then
        echo sums do not match!
        exit 1
fi

echo file is $BASENAME
date
rm -f $TMP1 $TMP2 $TMP3 $TMP4

exit 0

There is intermediate scripts on omega, chi and supercluster and another big script on psi to make this all work. But I'm not gonna share it all here.

0 comments

---

Purchased 2 WR-1043ND Routers for Freifunk

May 22nd, 2017

I purchased two more routers for Freifunk. I need these to accellerate my network at home a little. In other news I'm trying to give out a few WR-841 routers to stores near here in order to grow the Freifunk in my neighbourhood. We recently got a new router for Schweinfurt and the Freifunk Net is gonna rock!

0 comments

---

Purchased a 4TB external USB3 harddrive

May 18th, 2017

In fact I did this last week even on the 12th. The new drive is amazing. It can be written to at 70 MB/s, I find that extremely high. And that's only from copying from the internal drive on my computer "earth" (mac mini) to it, so it may just be limited to the read speed of earth's internal hd.

First I wanted to encrypt this drive but noticed that I had too many sectors in order to use softraid's crypt. I guess softraid by default is limited to 2TB unless you fiddle with the sector size which I did not want to do. So I left it unencrypted, on FFS2.

I already put some backups on this drive:

Filesystem     Size    Used   Avail Capacity  Mounted on
/dev/sd1a      3.6T    367G    3.3T    10%    /backup

Now you may wonder why i's 3.6 TB here and I said 4TB. Well WD sold this with the label "1 TB == 1 trillion bytes", so let's do some math.

$ dc
4 1000 * p
4000
1000 * p
4000000
1000 * p
4000000000
1000 * p
4000000000000
1024 / p
3906250000
1024 / p
3814697
1024 / p
3725

It comes very close to 3.6 in fact when I do the last step with a calculator I get 3.6376953. (dc doesn't give me decimal places in the way I'm using it). I'm overall pretty happy with this purchase. I've never had this much capacity before.

0 comments

---