Due to a logic-baffling stroke of luck and having met some awesome people a few months ago, I was given the opportunity to work with Microsoft Research for a little more than a week. Don’t get this confused with regular Microsoft development centers, these guys work on technology that you won’t even hear about for five years.

I have a clinical phobia of equations and theory in general. I have a fair amount of knowledge on development but absolute zero when it comes to computer science, leading me to feel very apprehensive about working there.

But the people I worked with were absolutely incredible, and were some of the nicest people I’d ever met, and made me feel that there is a place and role for me in their ecosystem. Rest assured I spent a productive few days listening, throwing an idea in when it seemed fit, and generally having a whale of a time.

While I can’t talk about what exactly I did there (partially because of the NDA but mostly because it’s not really that interesting), I can assure you that the final incarnation of the tool will be quite useful for everyone.

It’s been two week since the attack on Mumbai, and all but one terrorists were killed, but I’m starting to feel that they actually won.

As far as I’ve heard, malls and grand restaurants, the kind where the attacks took place, are now almost empty and are forced to drop their prices and ruin their profit margins. I even got a couple of messages on my phone from people who haven’t spoken to me for years saying

“Do not go to malls or theaters in december due to terrorist threat. This is from reliable sources from the military and intelligence”.

I could recognize this as a hoax as any dickhead can send a message and the sheeple(hi reddit) would be scared nutless.

The trouble with predicting these kind of attacks is that you can’t. These are more or less random and useless to predict, and it certainly won’t be repeated again. Nevertheless, the Indian public is now in uproar, with the usual bullshit statements like we must take action, politicians are at fault, etc, etc.

People fail to see that the issue was largely due to poor and late response. If such a threat was in the US or Europe, fully armed response would be there within a few minutes. The crisis took several hours to resolve in Mumbai, and their guns locked up and weren’t powerful enough.

Good domestic intelligence, and fast response is, in my opinion, the best we can do.

Whenever I go through security checks at the airport (and nowadays even at temples), I make it a point to ask how many people have been caught trying to sneak in with weapons, the answer to which is always none. Political policies, cooperation from religious heads, and a stricter United Nations would do more good than a good frisking ever will.

But now nobody trusts anyone else. Everybody are always a little concerned to step outside, and fear reigns supreme.

Some bloke on slashdot linked to his solution to the knight’s tour problem.

I had a different approach with dynamic programming and a seemingly simpler solution. I used python tuples for coordinates, and the magical “in” keyword (if target in visited: continue

Knight's tour

def get_legal_moves(coord):
  # Legal move means that the knight won't leave the board.
  # Doesen't take visited into consideration to save memory
  x, y = coord
  legal_moves = []
  for delx, dely in ((-2,1),(-1,2),(1,2),(2,1),(2,-1),(1,-2),(-1,-2),(-2,-1)):
    if (x + delx < 0 or x + delx > 7) or (y + dely < 0 or y + dely > 7):
      continue
    legal_moves.append((x+delx,y+dely))
  return legal_moves

def solve_tour(start, visited):
  legal_moves = get_legal_moves(start)
  visited.append(start)
  if(len(visited) == 64):
    print "Finished"
    print visited
  for move in legal_moves:
    if move not in visited:
      solve_tour(move,visited)

for x in xrange(0,7):
  for y in xrange(0,7):
    solve_tour((x,y),[])

This gives almost 15 solutions and is more or less brute force, so I don’t see how it’s worse off than the fancy solution which the slashdotted article proposes. And does it in 22 lines, compared to the 60 on the slashdot article.

either that or i’m doing something really wrong here.