Dun and Bradstreet spam

I recently got mail from Dun and Bradstreed, a Wall Street firm that (as far as I can tell) sells financial information on people and businesses.    I've never done any business with them, so their mail was (as far as I was concerned) spam.   And, typically of spam, it was not sent by them, but by another firm that they've hired to do this for them.    There were a number of links in the mail that purported to go to Dun and Bradstreet, but the URLs went to exct.net (and not to Dun and Bradstreet) and looked like http://cl.exct.net/?qs=long-hex-string-here.  They even had "unsubscribe" links.   I'm always reluctant to click on such links as who knows what is on the other end, and, of course, because clicking on unsubscribe links usually just confirms the email as valid to the spammers.  Eventually, I did click on one from a sandboxed browser and it took me to Dun and Bradstreet. 

I was still a bit suspicious though and sent Dun and Bradstreet an email pointing out that their mail was spam (at least unsolicited commercial email), that it looked like a phishing attack at the mail level, and that this was not considered good behavior.   I usually try to let companies know when they're being phished or are (perhaps inadvertently) being used in spam.    I got a response back from someone who had clearly not understood my points and fired off another email to them trying to explain just why their mail was spam, why it was suspicious and why I was bothered.  

A while later I got another response, and some of that is worth quoting.    I'll leave out the part where they say I'm misguided (which may be the case, honestly enough).       I'd also note that I now seem to be on their email list (I've received two emails from them asking me to do a survey in the last day) and will be marking their mail as spam in the future.    This response confirms that they are harvesting emails with a view to spamming ("sent to millions of recipients") and the part about "html to most and text to others" is also nonsense as I got the HTML version and the text version in the same email and it had the same links (that is, not to Dun and Bradstreet).  

The rest of this post is excerpted from their email :

As an FYI, the campaign was sent to millions
of recipients whose e-mail addresses we've collected through our Jigsaw
partnership.  Due to the large number of recipients involved, we're
bound to get a certain number of complaints from people who don't
understand the purpose of the campaign (though we tried to explain in
the message) and others who simply like complaining.

The only other comment I'd make is that the message was transmitted in
html to most and in text to others.  Our e-mail vendor informed us that
this is standard, as it depends on the formatting of the various ISP's
through which the messages are transmitted.  It looks like the recipient
below received the message in text format, which is why the links look
weird and unofficial.  I believe that only the html version shows the
graphics, D&B brand, etc.  Unfortunately, neither we nor our vendor have
any control over the format through which the given ISP's transmit the
message.

Some think new name.

A while back in the New York Times, I ran across a printed black rectangle with the words "Some think new name." printed (in white) in the top left hand corner. I had no idea what it meant, but posted it on my door because, well, it was odd and intriguing. Today I was looking at it and decided to do a Google search on the phrase. Not a hit to be found, nor was there one on Bing, nor on the NY Times web site.

Anyone have a clue as to what this is/was?

Cracker

BBC America today showed reruns of "Cracker" featuring Robbie Coltrane (probably better known as Hagrid in the Harry Potter movies). This is one of the best TV dramas I've seen, funny, tragic, and oddly haunting. Even if you don't like the usual TV crime drama stuff, this one is worth a watch and a second watch. Coltrane handles his part beautifully and almost effortlessly and the writing (especially in the episodes shown today) is almost perfect.

Pictures

I was in the Peace Corps from 1973-1977 - Zaire - now Congo (again) Shaba province (now Katanga again, I believe) in the towns of Luabo, Chibambo (on the Luapula river) and Kalemie (on Lake Tanganika). I carried a camera around and took pictures pretty often while I was there. This resulted in a box of fading photographs and another of negatives that I didn't really look at often. I've tried to scan the negatives a couple of times.

Once was in the local lab at EWU. They had a negative scanner attached to a nice big mac, but they'd disabled Terminal on the Mac so the only way to get images off was email or to burn a CD - instead of being able to scp them to my local desktop machine (sigh).

I also tried a negative scanner which scanned one image at a time, slowly, and required manual positioning on the negative. It was essentially unusable for my boxload of negatives. I returned it.

Then recently I was reminded that I wanted to do this and found a recommended scanner on Amazon - an Epson V300. And it worked very nicely, nicely enough that just two weeks later I now have 1500+ scanned negatives. I've put them all, pretty much without looking at them up on my work server (this will be going away at some point in the next year or so, I'll try to find another place to put them). I'll be going through these and removing the ones out of focus and that don't show anything, duplicates and the like and I'll also generate some thumbnails and labels. If you happen to be in one of these pictures and want it to be un-posted, let me know and I'll do that.

I may put some of them on Panoramio or something so I can add google map links to them.

Fractran in Haskell

One of the blogs I tend to follow is Good Math, Bad Math which had a post on one of those oddball things that can be lots of fun to play with, in this case Fractran, a Turing Complete language by John Horton Conway. I implemented this myself in Haskell back in (hmmm, not at all sure) 1997(?) in
Haskell. The quality of the Haskell novice I was then (and remain) shows though my Haskell has improved since then. Read that post for better information on what Fractran is and how it works.

This implementation factors the numbers used and just keeps track of the factors in the numerator and denominator of the fractions involved. I probably reinvented the wheel to get the primes and factors and all. I tested it in the primes program (named primegame) and
the addition program (named addergame), but not on much else as actually writing Fractran programs was not something I tried very hard to master.

import Ratio
--
-- run takes an integer i
-- and a list of fractions (the program)
-- it returns a result list of the integers generated
--

run :: Integer -> [Rational] -> [Integer]
run i p =      takeWhile (>0) res
        where
      res = i:(map runstep res)
      runstep j        = runl j p

runl :: Integer -> [Rational] -> Integer
runl    j []     = 0
runl    j (f:fs) = let
                     ifv = (fromInteger j)*f
                  in  if   isInteger1 ifv
                      then (numerator ifv)  -- essentially a toInteger
                      else runl j fs

isInteger i = (i == (truncate i))
isInteger1 i = ((denominator i) == 1) -- works in this context

--
-- the numbers themselves dont show much, so write them as products
-- of powers of primes
--
primes      :: Integral a => [a]
primes       = map head (iterate sieve [2..])
sieve (p:xs) = [ x | x<-xs, x `rem` p /= 0 ] 
powersOfTwo = 2:(map (2*) powersOfTwo)  
--
-- returns 0 if not a power of two
-- else returns the power 
-- 
whichPowerOfTwo x =     l                     
 where                  
       (a,b) = span (< x) powersOfTwo
       l = x == (head b) then 1 + (length a) else 0

--
-- brute force factorization
--

factor  x   = factor1 x primes 
factor1 1 _      = []
factor1 x (p:ps) = let (c,q) = fp x p
                  in  if c > 0
                      then (c,p):(factor1 q ps)
                      else factor1 q ps

--
-- multiply two lists of prime, power pairs
--

mult [] l = l
mult l [] = l
mult l@((p,pow):ps) l'@((p',pow'):ps')
 |  p == p'              = (p,pow+pow'):(mult ps ps')
 |  p < p'             = (p,pow):(mult ps l')
 |  p >  p'              = (p',pow):(mult l ps')

--
-- divide one list of prime,power pairs by another
--

divvy []  l = []
divvy l  [] = l
divvy l@((p,pow):ps) l'@((p',pow'):ps')
| p == p' && pow == pow'       = (divvy ps ps')
| p == p'                      = (p, pow - pow'):(divvy ps ps')
| p <  p'                      = (p,pow):(divvy ps l')
| p >  p'                      = (p, -pow'):(divvy ps ps')

--
-- isIntegerL returns true if the list represents an integer
-- when all the powers are >= 0
--

isIntegerL l = and (map ((>0).snd) l)

--
-- fp takes two integers and returns the largest
-- power of the second that evenly divides the first
-- and the quotient thus determined
--

fp :: Integer -> Integer -> (Integer, Integer)
fp x p     = if   x `mod` p == 0
               then let (c,q) = fp (x `div` p) p
                    in  (c+1,q)
               else (0, x)

primegame = [17%91, 78%85, 19%51, 23%38, 29%33, 77%29, 95%23, 77%19, 1%17, 11%13, 13%11, 15%2, 1%7,5
5%1 ]

--
-- start with 2^a * 3^b
-- ends with 2^a+b
-- so started with (8=2^3) * (243 = 3^5) = 1944 should result in 256 = 2^8
-- that is, run 1944 addergame => 256
--
addergame = [2%3]

p1 = filter (/= 0) (map whichPowerOfTwo (run 2 primegame))

-- main = putStr (show (take 20 p1))
main = putStr (unlines (map show (map factor (take 10000 (run 2 primegame)))))