Telephone +44(0)1524 64544
Email: info@shadowcat.co.uk

Sherlock Holmes debugging

Or, mst returns to Iron Man with a complete non sequitur

Mon Jan 10 19:00:00 2011

Digg! submit to reddit Delicious RSS Feed Readers

"Once you have eliminated the impossible, whatever remains, however improbable, must be the truth."

I'm on a flight from LAX to New York JFK[0] happily working my way through The Memoirs Of Sherlock Holmes when the young lady in the window seat next to me turns around and asks if I can hear anything through her headset, which is plugged into her seat's entertainment system.

I can't.

Time for a little exercise in deduction.

The headset was one I watched her buy and remove from the packet, so we have no idea of its prior/current state - it could be a dud.

Both her screen and mine are displaying a picture (the same channel, in fact) so the entertainment systems are not entirely broken.

First question: is it a problem with the headset?

I try the headset in my seat's entertainment system as well. Nothing.

What can we deduce from this? My first thought is "the headset's broken" - a reasonable theory, except that all we actually know is:

(1) Headset X does not work for the seat 13F entertainment system.

(2) Headset X does not work for the seat 13E entertainment system.

Therefore we in fact have two possibilities:

(1) there is a fault with headset X

(2) there is a fault with the entertainment systems of both 13E and 13F

I discard a potential third theory, that headset X is simply incompatible with the entertainment systems, on the basis that the cabin crew just sold her the headset and therefore for the purposes of the current problem incompatibility can be considered to be equivalent to a fault.

My JFK to LAX plane outbound on friday was also an airbus - although not the same model, the outbound had only one pair of escape doors by the wings whereas this one has two (can you guess why I found today's layout more intelligent?[1]) - and the seat arm controls appear similar.

I remember that the brightness control showed the current brightness on screen as you adjusted it (all the way down to "off" in my case), so I try the brightness control for my screen.

No response.

No response from the volume control either, or the channel change buttons. Or from any of the buttons on the young lady's panel either, when she tries hers.

We can now conclude:

The 13E and 13F entertainment system control panels are not operating.

At this point is seems extremely likely that the sound for both is also broken, but this is not actually proven. I think for a moment, and realise I have an iPod with me with a separate headset, Y, which I do know operates. A small amount of checking yields the following additional data:

(1) Headset Y works in my iPod

(2) Headset X also works in my iPod

(3) Headset Y does not work in the 13E or 13F speaker sockets.

So it is now fairly safe to assume that:

The 13E and 13F entertainment system speaker sockets are not operating.

Of course, this says nothing about the locality of the problem. Looking across the aisle, nobody is attempting to use their systems, which means either they are working fine or that those passengers have tried, failed, and not bothered to make mention of it.

I move my drink onto the young lady's tray table, and stand up - just in time to spot a man a few seats forwards put his headset on. I count to ten and observe nothing from him to indicate confusion or distress, so conclude that likely his system is functioning. Looking further forwards, there appears to be one other person successfully using their system, so the fault is presumably localised.

A localised fault is probably not going to be fixable by rebooting the entire entertainment system, which is usually the only means of repair available to the flight's service crew. This means that any further investigation is almost certainly meaningless in terms of the young lady's ability to distract herself, but I'm curious now.

I decide that it's now time to disturb the quiet Jewish lady[2] next to me, to determine the state of 13D before I consider 13A-C, 12 and 14, and ask her if I can check if her system works; she smiles, nods and goes back to her book. I have high hopes of something interesting since her screen is showing a different channel to ours.

The 13D control panel is also as dead as a doornail[3].

I fear my hopes are dashed, but unwrap headset Y again for completeness' sake and attempt to plug it in.

It takes two attempts - the feeling is somehow spongy - suggesting that the socket is either not quite the right shape (but D is unlikely to be shaped differently to E and F) or ... I prod with my fingers. Loose!

I lift the chair arm so I can actually see the thing, and it appears to be entirely adrift from its mounting in the arm, although otherwise undamaged.

Finally, I believe I have enough facts to produce a theory that explains the entire situation:

(1) Some form of point impact knocked the 13D socket loose

(2) This impact also damaged the wiring within in some unknown way

(3) The last signal from the 13D control panel changed the channel

(4) The wiring runs through the section of three seats in such a way that damage to one causes all three to fail

Noting, however, that correlation does not imply causation, one could also propose an alternative root cause:

(1) A fault developed in the 13D control panel or socket

(2) An attempt by a maintenance person to fix it resulted in the socket being left slightly adrift

However, I consider this less likely since the air hostess was unaware of the nature of the problem until informed of it [4] - and while point (4) can continue to stand, (3) seems noticeably less likely in this scenario.

Ergo: Working hypothesis successfully developed.

Smiling, I settle back to read Holmes' next case; on screen, a man in a full racing driver's outfit is playing a grand piano while on fire, but I am unable to hear why.

-- mst, back to ironman (wheeee ...)

[0] Crucial differences between Los Angeles and New York from what I'm able to tell: the urinals are mounted much lower on the walls of the gent's toilets, and the primary danger of death as a pedestrian is being run over by a police car rather than a taxi.

[1] Given everybody heading to the nearest exit in the event of an emergency, two exit door pairs in the centre will lead to an equivalent amount of traffic per exit door, whereas a single centre exit door pair requires only the middle third of passengers to head for it. Such level headed mathematical thinking from the average human seems highly unlikely in a crisis, whereas "which line to not dying is longer?" seems a question many more people would think to ask.

[2] Modest (dare I say Dwane Dibley?) haircut, practical but unlovely glasses, reading what looks to me like musical notation that that's been put through a mangle but is presumably Hebrew script. Theory all but proven by 13B and 13C being occupied by two young men in black trousers, white shirt and the appropriate sort of black cloth cap who are reading a much larger book of the same script.

[3] I know not why doornails are phraseworthily moribund; presumably something along the same lines as "a cat's chance in hell" making a lot more sense once one learns the original pre-truncation version ended "without claws".

[4] Since the fault led to a $15 discount card being issued to each passenger in 13D-F her not admitting to knowing about it is potentially a null data point, but her going and fetching another headset (too quickly for me to point out the futility of such an action) as a first attempt to solve the problem makes the odds good. For a person to be happy to be dishonest but not to use that information to save a trip to fetch another headset seems relatively unlikely.