Author: adamo

I am a Greek DevOps Engineer ranting about everything, from email to sports; to living in Athens, Greece.

Bureaucracies and information flow

Interesting quote from a security presentation that I attended recently:

“Bureaucracies depend on information flow to maintain function. Change the information flow and you can predict function”

I miss the days when hacking cracking was about ownage and defacement. It has now evolved to a strategic capability, which makes things far more difficult for the defenders.

wash your hands

Bear with me, this is actually a computer security post. In parentonomics Joshua Gans cites an Australian study according to which doctors believed they washed their hands after going to the toilet 73% of the time. Close monitoring however revealed that this happened only 9% of the time. This in a pediatric intensive care unit!

This is a simple requirement: Wash your hands when leaving the toilet! One would expect that medical professionals, of all people, would follow it and not believe that they follow it.

So if the most simple measure, and one instructed to them from a very young age, cannot be followed through, how on earth are we supposed to make people read, understand and actually follow any security policy? How much simpler than “wash your hands“ does it have to be?

Recently I heard the argument that “I do not mind using cracks and pirated software*, since I trust the source”. Oh really? I am sure they [the source] wash their hands every time too…

In the case of hospitals the problem was solved using a kind of public embarrassment (screen savers with the names of doctors with no clean hands). Or as Gans puts it “Data plus shame equals trust”. However, I am sure that no legal framework can allow for the public embarrassment of any computer user. Nor any administrator wishes to make more enemies among their users than they already have.


[*] – Using cracked versions of software when the price is not right is not the way to go. If you want to punish the vendor quit using their product and stop advertising it by using it.

parentonomics

Unlike what one might expect from the title, parentonomics is not Jo Frost disguised as an economist, nor an economist playing Jo Frost. Joshua Gans is a father of three, applying his scientific discipline into parenting and documenting the results. And he does so in an instructive and humorous way. I wish I could write my experiences with my three children in a similar way. Maybe someone else can document parenting while viewing it though an algorithmic or engineering prism.

In my opinion this is a book for fathers. Other books on parenting that I have checked have a more motherly approach, so this is a refreshing change. Because unlike the “no two kids are the same” principle, significant others’ reactions seem to follow a pattern* regardless of the number of children. Soon to become fathers prepare yourselves.

I really liked the fact that this book discusses parenting of three. Most of the literature that I have browsed seems to address the issue of the first (or single) child in the family. One would expect that after the first child one is prepared to deal with the second (and third), but hey you are not: Family management complicates exponentially. And in my case (child 0 first+, twins next) it complicates even faster than Gans’.

One interesting observation that occurred to me while reading the book is that all parents seem to be non systematically trained game theorists (game practitioners maybe?). Which is basically the reason why many strategies we employ as parents are flawed or simply not working. All in all this is a good book that has advice to offer and data to back up the opinions it carries. I really enjoyed reading it.


[*] – Either there is a pattern, or our wives would definitely be friends (or both).

[+] – Gans enumerates his children as Child 1, 2 and 3. I prefer the K&R approach :)

pseudoscience

Όχι ρε σεις! Εάν δεχτώ τις πίπες που σερβίρετε έχω κλειστό το μυαλό μου. Το μυαλό μου είναι ανοιχτό ακριβώς επειδή δεν τις δέχομαι.

Αθλητικό ΠΠΣΚ

  • Παρασκευή: Ο αγώνας κρίθηκε στο καλάθι. Χάσαμε, αλλά πήγε καλύτερα από ότι περίμενα. Διάφοροι ειδήμονες που γράφανε και λέγανε “έπρεπε να πάει για τρίποντο” με πιθανότητα P → 1 δεν έχουν βρεθεί ποτέ στη θέση να πάρουν και να εκτελέσουν μια σημαντική απόφαση σε 5 sec.
  • Σάββατο: Αφαιρώντας τους αγώνες της Εθνικής στο Euro 2004, πρέπει να είναι ο πιο ενδιαφέρον αγώνας που έχω δει · για το Ελληνικό Πρωτάθλημα και Κύπελλο σίγουρα. Παρολαυτά, όλοι όσοι αναφέρονται στον “τελικό των τελικών” βολικά ξεχνάνε τα επεισόδια. Το σχόλιο του Stazybo Horn επί αυτής της βολής είναι χαρακτηριστικό.
  • Κυριακή: Πάλι στο καλάθι κρίθηκε το ματς, αν και το πρώτο ημίχρονο με έκανε να σκέφτομαι την πιθανότητα τελικού “σούπας”. Όμως αυτό το σφύριγμα για τα 8 sec ήταν όντως από τα “strange things happened in the second half”.

Και γιατί ΠΠΣΚ και όχι ΠΣΚ; Γιατί την Πέμπτη στον παιδίατρο:

– Τι ομάδα είσαι Θ ;
– Ο-λυμπιακός! (δυνατά σε δυο ανάσες)
– Ποια είναι καλύτερη ομάδα, ο Ολυμπιακός ή ο Παναθηναϊκός;
– Παθακός
– @!#!&^%

Οι όροι εισόδου στο σπίτι πρέπει να αλλάξουν και το πρόγραμμα κόκκινης επιμόρφωσης να εντατικοποιηθεί.

12/13

Και τυπικά πλέον. Αλλά μια από τα ίδια. Ή μάλλον χειρότερα: Κύριε Σάββα Θεοδωρίδη κάνε την μέγιστη προσφορά που σου μένει στον Ολυμπιακό. Είναι καιρός να αποσυρθείς. Δεν προστατεύεις κανένα · εκθέτεις όχι μόνο εμάς, αλλά και τον εαυτό σου. Και δεν είναι η πρώτη φορά. Ο παραγοντισμός έχει αλλάξει από το 1950 και πρέπει να το καταλάβεις.

Με σεβασμό.

twitter for file sharing?

The thought occurred to me while reading Matt Welsh’s criticism on how Project Graffiti conducted their experiment. I think it is possible to share files over twitter and I will sketch the concept bellow:

  1. Create N accounts on twitter.
  2. uuencode the file in question so that we have to deal only with printable characters. Optionally encrypt it first.
  3. Pick randomly the account that will tweet first.
  4. tweet the first line of the file.
  5. Use a hash function (like SuperFastHash) and hash the line. That way you will get the next account to tweet the file contents.
  6. The next account replies to the previously tweeted line. The reply contains the next line.
  7. Repeat hashing and replies until the whole file is tweeted.

When the process is finished you end up with a string of replies that if put together contain the (uuencoded) contents of the original file.

Yes, this is neither practical nor distributed file sharing, but in a way it is hiding obscuring information in plain sight.