Thursday, June 27, 2013

The origin of evil genius

I guess I always assumed that the term "evil genius" was a Batman or James Bond kind of term, a reflection of the deranged excesses of 20th Century greed and technology.  Perhaps in the cartoon usage it is.  But I ran across a much older use of this phrase.

I am currently reading (make that slogging through) Camilla by Fanny Burney.  In it, the sweet young heroine, Camilla, in love with the wealthy, but apparently indifferent Edgar Mandelbert, is trying to fend off the advances of the wealthy and somewhat annoying Sir Sedley Clarendel, who is being egged on in his suit by their mutual friend Mrs. Arlbery.
'O, Mrs. Arlbery!' she cried, 'lend me, I beseech you, some aid, and spare me, in pity, your raillery! Sir Sedley, I fear, greatly mistakes me; set him right, I conjure you....' 
'Me, my dear? and do you think if some happy fatality is at work at this moment to force you to your good, I will come forth, like your evil genius, to counteract its operations?'
Wiktionary provides this sense as its first definition of evil genius: The spirit each person is believed to have in attendance, according to certain religious or mythological traditions, which tries to negatively influence him, and is opposed by one's good genius; loosely, someone who is a bad influence.

Tuesday, June 11, 2013

Snowden's next step?

Wanna know how the Snowden story can get more interesting than it already is?

Mr. Snowden has enumerated the vast breadth of secret intelligence data that he had access to as a systems administrator.  No surprises there.  Systems administrators are inherently the most trusted of insiders, and in general, take their trust extremely seriously.  A sysadmin is kind of like your gynecologist...you want your GYN to check you thoroughly for the purposes of keeping your privates healthy, but you kind of have to take it on faith that they don't have hidden cameras in the examination room taking pictures of your privates to post on Women-I-Have-Seen.com.

But that's where the analogy ends.

Because a sysadmin also has access to a second kind of information that may be even more interesting in this case.  That would be the forensic metadata.

Yup.  Just as the NSA appears to have been collecting metadata about you (who you called, when you called, and possibly where you were physically calling from), so any system worth its salt will log information about who accessed it, what they accessed, when they accessed it and where they accessed it from.  This is an important fraud and abuse prevention feature, especially in systems that maintain highly sensitive data. For example, if your medical insurance records got posted on a public website, you'd want some way to go back to your provider and demand accountability about how your records got there.  You'd want them to be able to track down the person who accessed your records and fire them.  Along with giving you a $27 million settlement.

Normal system users aren't interested in the logging information.  They just want to get at the main information that the system manages (in Snowden's case, the actual surveillance information).  An intelligence analyst normally doesn't care what subject the guy in the next cubicle is tracking, just get me information about my subject.

Snowden, however, as part of his sysadmin job, would have been expected to assist with forensic examinations of system logs if fraud or misuse were ever suspected (or if they wanted to nail some analyst for any reason).  So he would almost certainly have had access to all sorts of log metadata (and systems for analyzing the metadata) about which analysts accessed what when.

Now if things at NSA are anything like they were at State during Bradley Manning's time, I think we can reasonably expect that WAY, WAY more people have access to more sensitive data than they should.  Which means that it's quite likely that Snowden's not the only one leaking information.  He's just the only one to come out and talk about it in the public interest.

Now, if he really wanted to muddy the waters and confuse the people who are inclined to call him a traitor, here's what he'd do.

Before he left, he would examine the log files and identify a few bad actors who had clearly accessed information they could have no work-related reason for accessing.  Things like analyst Jimmy Jones, (assigned to the Africa desk) downloading a report on the calling history of the US ambassador to Russia (which happens to include numerous calls to his Chinese mistress, who lives in Thailand).  And Jimmy's access of reports on several prominent US businessmen who have been trying to set up businesses in Russia.  And Jimmy's access of data on members of Pussy Riot.  Suddenly, it's starting to look like maybe Jimmy's helping out some Russian contacts, even though he's supposed to be focusing on the Congo.

If Snowden is smart (and he seems to be a bright young man, capable of thinking more than one move ahead), he will have assembled a certain number of cases like this.  He will hand the documentation about these cases (log files, analysis of them, etc.) to Glenn Greenwald & Co. The journalists will then publish a sanitized version of them (removing names of targets, etc.), clearly demonstrating that this data is being actively used to harm American interests, while not showing any demonstrable benefit in hunting terrorists.  They will also send the unsanitized analysis to trusted members of the NSA/intelligence community, so they can take action on the rogue elements.  If there is anyone un-corrupt enough to do so.

If it's potentially so easy for random analysts to access stuff they shouldn't, you might wonder what is to stop any analyst from accessing the records of Kim Kardashian or Sergey Brin or Obama or the annoying neighbor next door.  Since I am not privy to the technical details of the NSA systems, I cannot answer that with any certainty, but I can guess at some of the controls.

First of all, even though the analysts are not interested in viewing the log metadata, they are almost certainly aware of its existence.  They know, better than anyone, that Big Brother is watching, and if they are caught with their hands in the cookie jar, they will get in trouble, possibly to the extent of jail time.  There may even be the equivalent of a burglar alarm built in to the systems, such that if an analyst does something naughty, an alarm goes off somewhere, and somebody investigates what's going on and fires/prosecutes the analyst who's peeking at stuff they shouldn't be.

HOWEVER, there are a couple of problems with relying on this.  Think about getting an alarm system installed in your house.  ADT charges a fair amount for their alarm monitoring services.  And they don't respond to every alarm.  If your alarm goes off and it looks like you opened your door without turning off the system, they don't bother.  An alarm system has two choices:  make sure to react to all the naughty accesses, but in the process also react to many, many things that look like they might be naughty but are actually legitimate OR don't bother unless you're fairly sure that the access is naughty, but potentially miss a bunch that look like they might be legitimate but are actually naughty.  Similarly, in NSA's system, the vast majority of accesses would be legitimate.  Sometimes an analyst may raise an alert as they legitimately access data for an unusual connection to their case.  Less often (we hope) they may raise an alert as they access data about their annoying neighbor.  The question is...who's going to respond to all the alerts (mostly false alarms) generated by hundreds or thousands of analysts working for the NSA?  System logs work much better AFTER an abuse has been discovered, and you need to track down/document whodunit.  They don't prevent the abuse (except by fear), and they don't work at all if nobody looks at them much.  Which a rogue analyst would also know.

This country is frankly not in the mood to address this issue.  Americans want to catch all the bad guys (terrorists and people abusing the intelligence systems), but don't want to pay for the effort.

We are told there are more bombshells to come.  I predict that revelations of data misuse or that sort of thing will be included.

Friday, June 7, 2013

The first person to march to a different drummer

I've always been a fan of Frost's poem "The Road Not Taken."  It suggests an approach to life that I aspire to:

Two roads diverged in a wood, and I,
I took the one less traveled by,
And that has made all the difference.
A traveler down that road would undoubtedly be marching to her own drumbeat.  And as I recently discovered, that image comes to us, courtesy of the Transcendentalist Henry David Thoreau.  I'm no authority on Thoreau, but from what I know of him, I'd have to guess that this quote from Walden probably expresses his quintessence:
"If a man does not keep pace with his companions, perhaps it is because he hears a different drummer. Let him step to the music which he hears, however measured or far away."
So Sidharta!  I like an extended version of that passage, as well:
"Why should we be in such desperate haste to succeed and in such desperate enterprises? If a man does not keep pace with his companions, perhaps it is because he hears a different drummer. Let him step to the music which he hears, however measured or far away. It is not important that he should mature as soon as an apple tree or an oak. Shall he turn his spring into summer? If the condition of things which we were made for is not yet, what were any reality which we can substitute?"
Food for thought in our Race to Nowhere culture.

Saturday, March 23, 2013

Review of "Broken Ballots: Will Your Vote Count"

Broken Ballots: Will Your Vote Count?,” by Douglas W. Jones and Barbara Simons, is a tour de force review of the history and current state of voting technology security. The authors (who were key players in security reviews of a number of electronic voting systems, as well as in voting technology policy discussions at the local and national level) provide us with unique insights into the technical, procedural, policy, and even political difficulties of assuring election integrity.

The central challenge in voting systems, from a security perspective, is the absolute requirement for ballot secrecy (to prevent coercion or vote-buying) while ensuring that all eligible voters are allowed to vote, but no more than once per election. The notion of an eligible voter implies some sort of authentication system, while secrecy demands that in spite of authentication, ballots not be linkable to individual voters. Requirements for massive scalability, efficient vote tabulation, usability, accessibility for voters with a range of disabilities, ballots containing multiple races, and cost-effectiveness impose additional complexity on voting systems. This book explores how technology has attempted to achieve these conflicting goals, and how the complexity has often created vulnerabilities that threaten election integrity, which in turn, has required technology to evolve.

“Broken Ballots” documents a large number of case studies of security concerns on a variety of electronic voting technologies, including Direct Recording Electronic (DRE) and Internet-based systems. An entire chapter is devoted to the missteps of Diebold, “the poster child of much that is wrong with DREs”, including overt partisanship suggesting vote rigging, hiding poor coding and deployment practices behind a screen of “trade secrecy”, circumventing the voting software Independent Testing Authority and certification process, use of programmers who had previously been convicted of computer-based fraud, and harassment of independent researchers who disclosed the existence of vulnerabilities. Another chapter is devoted to risks associated with Internet voting, quite similar to those characteristic of e-commerce, including server-based attacks, client-side malware, phishing, counterfeit sites, man-in-the-middle attacks, DDoS, the loss of ballot secrecy (a risk specific to voting).

Lest we be tempted to return to simpler times before touch screens and the Internet, the authors provide a fascinating history of voting technologies and attacks on them, from voice vote to ballot boxes to punch cards. (Who knew that lever voting machines were susceptible to jammed gears?)

The authors describe the challenges associated with developing meaningful voting standards and critique a number of failed attempts (such as the 2002 Help America Vote Act and several of its revisions.) They leave us with a number of concrete recommendations for improving the integrity and transparency of elections, including:
• Development of uniform election standards (technological as well as procedural requirements)
• Technological support for audits and mandatory post-election audits
• Greater vendor accountability
• Revamping voter enfranchisement laws
• Explicitly forbidding Internet (as well as fax and phone) voting until significant security breakthroughs have been achieved

“Broken Ballots” should be largely accessible to a non-technical audience, but those with IT experience will respond more viscerally to the cringe-worthy practices it documents. The CISSP will appreciate the thoroughness of the analysis, as it touches on practically every domain of the CBK, from physical security to secure coding practices to governance. The case studies it cites are primarily (though not exclusively) US-based, but the principles these illustrate are universal. It is not a quick read, and at times feels a bit shopping-listy, but it is well-worth working through it, and should be of interest to every adult citizen of any country that conducts elections.


This review appears on the ISC2 book review section.

Monday, March 11, 2013

Inquiring minds want to know

So don't ask me why...really don't...I have no idea...but I was struck this morning by the similarity between the words "require" and "quiero" (Spanish for "I want"). Perhaps I've been working on requirements too much lately. In any case, what is a requirement, after all, but something I want?

In such intriguing cases, I generally resort to my handy dandy Random House College Dictionary (acquired when I was in Grad School) which often provides helpful etymological information. Paydirt!

Unsurprisingly, Require, Acquire, Inquire, Query, Quest, Question all originate from the Latin quaerere, to seek, ask.

That's as distinguished from Quarry, which, according to Merriam Webster Online (now that I've put away my dictionary) is derived from the French cuir (skin, hide), which, itself, is derived from the Latin corium (which also gives us Excoriate). Which is weird because if you'd asked me, I'd have said that a lot of hunters seek quarry. In fact one of the definitions that Merriam Webster offers for quarry is "one that is sought or pursued." And then there's the stone quarry, which is totally unrelated and comes from the Latin quadrus meaning square (as in a square block of stone).

Who gets to decide the official etymology, anyway?

Friday, January 18, 2013

Look before you click - or Phish Spotting

As a public service, I thought I'd provide a little forensic analysis of an incident I experienced yesterday.  The punchline is:  when in doubt about the legitimacy of a web site, try a whois search.

Annapurna was very excited yesterday evening at receiving an email that purported to be from the school of her choice, encouraging her to request admission info from them.

Now, let me say from the start, that I think that said school would be very wise to encourage her to apply,  and I think it not unlikely that it will in time be brought to its senses and will ultimately send her a supplication to grace it with an application and a $50 application fee.

But that may not be what she got yesterday.

No.  This was a college admissions phishing email.  It said,

Based on your achievements thus far, I invite you now to request our viewbook [with a link to...well...you'll hear more about that later],  Within its pages you will discover the countless ways our dynamic campus in [city of School of Choice] can educate, invigorate and inspire you.
Sincerely,
[Real name of Dean]  
Dean of Undergraduate Admissions and Financial Aid 
P.S. We also offer one of the most generous need-based financial aid programs in the country. We received your contact information from the Student Search Service of the College Board. 
Flattering, no?  Who could resist clicking on that magic link?  And if you did, you'd see a page that looked like something the School-of-Choice might put together, complete with school colors, logos, links to legitimate pages at School-of-Choice, etc.

Here's how I got in this story.

I had heard Annapurna going on for a while about the invitation, but I didn't pay much attention.  She's gotten enough legitimate solicitations for applications that in my books this one was no biggie.  It was when she started pestering me about "Should I put your email or Dad's?" that I started to tune in.  I was like, "What?  What do they want MY email for???"  My spam detector was buzzing.
  • Dawn's Helpful Hint (DHH) #1:  when a site that already has YOUR email and has already taken the liberty of establishing a relationship with you, they probably don't have any legitimate reason to be asking for anyone else's email address or contact info UNLESS you are in some very, very official process (like an actual college application or providing death benefits information).  Whenever you get a request for someone else's info, your scam detector should go off.  Before you supply the information ALWAYS think very carefully...does this site REALLY need this?  Really?  Why??? And would that person get upset if I gave out their information?
Further investigation showed that the link went to something that looked like it might be related to School-of-Choice, but wasn't quite right.  Let's pretend that the URL for School-of-Choice is http://www.soc.edu.  This thing went to http://www.soc-admin.org.
  • DHH #2:  universities almost always use a .edu If you see something purporting to be from a university but doesn't end in .edu...be suspicious. 
  • DHH #3:  when in doubt about whether a web page is associated with a large organization like a university or a bank, try to get there from the home page of the organization.  In this case, we went to http://www.soc.edu, and of course the normal admissions pages we got to from there had nothing to do with http://www.soc-admin.org.
And here's the piece de resistance:
  • DHH #4: You can always get some pretty interesting information about who really owns the site by performing a WHOIS search.  Network Solutions provides a reasonable one. For example, if you do a whois search on mtsd.us, you will find
  • Registrant Name:                             Thomas DeSisto
    Registrant Organization:                     Montgomery Township Board of Education
    Registrant Address1:                         1014 RT. 601
    Registrant City:                             Skillman
    Registrant State/Province:                   NJ
    Registrant Postal Code:                      08558
    Registrant Country:                          United States
    Registrant Country Code:                     US
    Registrant Phone Number:                     +1.6094667182
Looks pretty legit for Montgomery, no?  But when we did that whois search on that soc-admin.org, we found it was registered by one Royall and Company in Virginia. Nowhere near School-of-Choice.  Not surprisingly, when we combed through Annapurna's email for the assorted solicitations she has received over the past few months, many, many of them contained links to websites registered by that very same Royall and Company.  

So why does Royall and Company want to pretend to be assorted gourmet universities?  I don't know.  I'm not Royall and Company.  Their web site indicates that they do direct marketing for student recruitment.  So maybe I'm being too cynical and these universities really have hired Royall to reach out to students who fit some profile or other.  Or maybe they want to sell mom and dad's email address to banks looking to make student loans.  Or both.  I can't tell.  

Saturday, January 12, 2013

US Statistics

There are some really interesting statistics at Ranking America.  This site provides a number of graphics illustrating how the US compares with a number of other countries on assorted categories.  The US does rather poorly on many indices of health and wellbeing, compared to other developed countries.

Some of the more telling: US ranks:

  • 33rd out of 34 selected countries (Europe + Japan + US) in acceptance of evolution
  • 2nd out of 35 selected countries in child poverty behind only Romania
  • 12th out of 12 countries in introduction of sex education 
  • Abominable in arts funding:  $0.44 per capita, as compared with, say, middle-of-the-pack Northern Ireland, which spent $13.62 in 2005
  • And one I am certain you have often wondered about:  2nd in imports of asses, with 28.1% of the world's ass total imports (second only to Yemen).  
I really like two things about this site
1)  It provides simple graphics (generally bar charts) to show relative rankings
2)  It provides sources for the statistics it is using.

Some of the graphics are a tiny bit confusing or misleading, where they show the US relationship to the top 10 (even though the US isn't actually in the top 10, and they don't tell you what position the US actually has.)

Tuesday, January 8, 2013

Total Information Awareness...Take 2

Oy...

Check out this Wall Street Journal (of all sources!) article on how Total Information Awareness is being resuscitated:

The rules now allow the little-known National Counterterrorism Center to examine the government files of U.S. citizens for possible criminal behavior, even if there is no reason to suspect them.
And
The agency's best-known product is a database called TIDE, which stands for the Terrorist Identities Datamart Environment. TIDE contains more than 500,000 identities suspected of terror links. Some names are known or suspected terrorists; others are terrorists' friends and families; still more are people with some loose affiliation to a terrorist.  
Yeah, that means you, you threat to national security peace activists

Facebook and Pinterest posts not included.  

Yet.

Wednesday, January 2, 2013

2012 books

As I did, for 2011, I'm posting a list of books read/listened to in 2012.  Reasonably good year.  Couple of groaners that I felt obligated to finish, in spite of my better judgment.
* = Recommended
X = Stay away! - No absolute musts, this year, so I've indicated groaners with an "Oy"
  • Mike and Psmith: P. G. Wodehouse (I'm tempted to put this as a *, but it's such fluff :-))
  • An Occurrence at Owl Creek Bridge: Ambrose Bierce
  • Death of Ivan Ilych: Leo Tolstoy
  • * The Greatest Show on Earth (audiobook, narrated by the man, himself): Richard Dawkins (I'm putting this as a * -- it's pretty accessible, but at times it gets a little technical -- stick with it, though.) I think I still prefer Sapolsky for writing style, and coverage of a lot of the same kinds of concepts.
  • Persuasion: Jane Austen (what would a reading year be without an Austen?)
  • The Life of Pi (audiobook): Yann Martel - didn't get it - can someone please explain it to me?
  • Mansfield Park: Jane Austen (Mostly because I saw the 1999 movie and was like...really??? She didn't write that...did she?)
  • * Last Man in Tower (audiobook): Arvind Adiga
  • * Plugged (audiobook): Eoin Colfer - may not be for everyone...a bit heavy on the violence, but darkly charming
  • Nation (audiobook): Terry Pratchett
  • The Thousand Autumns of Jacob de Zoet (audiobook): David Mitchell
  • The Complete Stories of Dorothy Parker (audiobook):  Dorothy Parker
  • Wives and Daughters:  Elizabeth Gaskell (This was a re-read, mostly because I discovered the lovely BBC miniseries based on it, by Andrew Davies, the same screenwriter as did my very favorite Pride and Prejudice production - it's almost as good, though nothing can ever compare, really.)
  • Between the Assassinations (audiobook): Aravind Adiga
  • Next (audiobook):  Michael Crichton - Oy...
  • Fairest (audiobook):  Gail Carson Levine - think I liked Ella Enchanted better, but this was good, too.
  • Mrs. Dalloway (audiobook): Virginia Woolf - very strange, though I enjoyed it - never really saw a woman writing stream of consciousness like that
  • Return of the Native (audiobook):  Thomas Hardy
  • Daniel Deronda:  Mary Ann Evans (aka George Eliot) - kind of Oy
  • Pillars of the Earth (audiobook):  Ken Follett - Oy...but now I'm really, really comfortable with the difference between a transept and a chancel and a nave.  Also clerestory.
  • Psmith in the City:  P. G. Wodehouse
  • Hard Times (audiobook):  Charles Dickens - Oy
  • The Girl Who Played With Fire (audiobook):  Stieg Larsson - bit of a disappointment...I loved the other two, and it was a mistake to read them out of order, but honestly, I think this one could probably be dropped completely or have at least 30% excised.  Still hoping to meet this Blomqvist guy.
  • * The Kitchen God's Wife:  Amy Tan (inspired by Annapurna, who read it for an English project) It is a testament to how much I enjoyed the story that I was willing to read it on paper.  Helped that I had an upstairs copy and a downstairs one, courtesy of a donation of a spare copy from my mother :-)
  • * The Time Traveler's Wife (audiobook)Audrey Niffenegger - I'm putting this as a *, but it's kind of chick lit, though writers will find it an interesting study, as well - lot of structural and point of view technical challenges handled nicely here.  Actually, I'm not done with it yet, but I'm close enough that I'll claim victory.
Started:
  • Fixing my Gaze: Susan R. Barry
  • Moby Dick: Herman Melville
  • Little Brother: Cory Doctorow
  • Assorted essays from The Best American Science and Nature Writing (2011 edition) -- some quite good.
  • Interpreter of Maladies:  Jhumpa Lahiri  (inspired by Annapurna, who read it for English class)
  • Broken Ballots:  Will Your Vote Count:  Douglas W. Jones and Barbara Simons
  • Harry Potter 4 - with Annapurna and Sidharta - parts
  • The Great Gatsby: F. Scott Fitzgerald (inspired by Annapurna, who read it for English class)
  • Moll Flanders:  Daniel Defoe - still working on it
  • Rakshasa's Ring:  Visakhadatta - still working on it - I read it years ago and loved it...talk about royal skullduggery...it's like Philippa Gregory on steroids
And with Sidharta:
  • Heidi: Johanna Spyri -- surprisingly, he really enjoyed it, though he got bored and abandoned it after 80%
  • Lots of Calvin and Hobbes, as always
  • Lots of his 1000 page tome on Mammals (aka The Princeton Encyclopedia of Mammals which I got a couple of years ago in a lucky purchase on sale for $20 on one of the outdoors tables at Labyrinth.  It normally goes for about $40...ask me why I know that :-)
  • Lots of Asterix
  • Children's abridged version of Time Machine: H.G. Wells -- he wanted to read it a second time after we finished it - he had read parts of it himself, and wanted to hear it again
  • Some Grimm's Fairy Tales and Yellow Fairy Book
  • Son of Neptune:  Rick Riordan (parts - he read most, himself)
  • The Hobbit:  J.R.R. Tolkien (part - I think the language got a little highfalutin for him)
  • * Robin Hood: (As retold by Roger Lancelyn Green) - one of my all time favorite children's books
And a good chunk of Harry Potter 4 with Annapurna and Sidharta.

And I feel I also deserve credit for having read a couple of additional 1000 page novels by virtue of having watched the Forsyte Saga (seasons 1 and 2), Downton Abbey (seasons 1 and 2) and the Wives and Daughters series :-)  Also a couple of excellent Coursera classes (especially Alex Halderman's Securing Digital Democracy)

Monday, May 14, 2012

Like a Bat Out of Hell

No, I am not a Meat Loaf fan. To tell the truth (shhh...don't tell anyone)...I don't think I've ever heard anything by Meat Loaf. Well, I take it back...I did watch Rocky Horror Picture Show...once...many years ago...many, many years ago. Maybe I would be a Meat Loaf fan if I tried.

But I digress.

I had always assumed that the expression " bat out of hell" originated with Meat Loaf's song of the same name. I was, therefore, surprised to run across it in one of Dorothy Parker's short stories from a 1932 edition of Harper's Bazaar, "Dusk Before Fireworks". In it, a young woman, jealous of another woman, says to her putative boyfriend, "She'll shoot out of Greenwich like a bat out of hell, if she thinks there's a chance of seeing you."

The Word Detective tells us that this expression first appeared in print in 1921, but does not provide a source.  Wordwizard says it was used in the novel The Three Soldiers by John Dos Passos, and supposedly it started up in WW I.  The Wiki page on Meat Loaf's song mentions a reference to a bat coming out of hell in Aristophanes' The Birds, so that takes it back a little further.  Kind of like that Homer reference to biting dust.

Just thought you ought to know.