Proving the Obviously Untrue

This Friday, the time from the start of the Unix Epoch will be 1234567890, as noticed by Linux Pro Magazine, and no doubt others.

Mathematics, being in many ways really the study of patterns is always interested in these kind of things. We are now several years on from 19-11-1999 where all the digits in the date are odd. That was the last such date for over a thousand years, the next being 1-1-3111, but then we are enjoying an era of lots of dates where all the digits are even right now, which started on the 2-2-2000 after the last one being 28-8-888 I guess.

Posted by Colin Turner

Every year I attempt to visit my old PhD supervisor Brian McMaster (old in the sense that my PhD is now a thing of the past, I am making no reference to the man in question!) at Christmas time to have a quick natter and exchange gifts. I was squeezed for time this year since I also had to hire a gown from Queen's to attend an event for the University of Ulster (long and boring story - getting an award for our work on OPUS). Anyway, just before I left, I asked him about my previous Crashing Cars problem. He wasn't by any means the first PhD I'd asked about this, and I'd even asked a few physicists. I was hoping of course, that he would immediately say I was being stupid and had missed something obvious, but he found the problem as bothersome as I did.

I continued to mull it over a bit, and even found the problem had some more unsettling properties to do with the masses of the car, but didn't make any progress unraveling the mystery. Well, a few days later Brian CC'd me on an email to someone with whom he had clearly been discussing the problem with a possible solution. Having read it a good few times and thought it over it makes sense to me and it doesn't come as a surprise that Brian was the one who cracked the central nub of the problem. I'm a lot more cheerful about it now, but it goes to show that dark nasties can lurk in surprisingly simple problems.

I was brooding some more about the frames of reference thing and maybe beginning to see where the paradox lies. Thing is [perhaps] that the observer in the car is non-accelerated only up to the moment of impact: we can't use him to assess KE after that moment with the same cavalier abandon that prevailed beforehand. [Especially since he'll have a headache.] Thus it is not legitimate to say: "from car 1's POV, KE before = 2mv2, KE after = 0 + 0, therefore KE dissipated into crunch = 2mv2". Which would be very troubling since it would appear to let us distinguish between states of rest/uniform motion by Physics.

What we can say instead is that from the point of view of a non-accelerated observer *travelling initially with car 1*, the KE before = 2mv2 and the KE after = 1/2 2m (-v)2 = mv2 so the KE going into the impact process = 2mv2 - mv2 = mv2. Which agrees with the observer on the roadside! So maybe the old geezer with the mop of white hair and the century's most iconic formula was right after all. I'm sufficiently encouraged to copy this email to my tormentor in Jordanstown and see if it allays his apprehensions. Hi Colin!

Posted by Colin Turner

Consider two identical cars, each travelling at a constant velocity v towards each other. Suppose that they crash, and to simplify things, let's not consider any rebound. Then the energy of each is kinetic in nature and so the energy of the crash is given by



Ok. That's a nice way to show the energy of the collision goes up with the square of velocity. Now, according to the principle of relativity this should be an identical situation if you consider it from the reference frame of one of the cars. In that case, the car you are looking at has a velocity of 0, while the other car has a velocity of 2v. Now the total energy of the crash looks like



So. Why the discrepancy?

Posted by Colin Turner

On Saturday night, quite by accident, I sat down to watch "Einstein and Eddington" on the BBC. Einstein has been an inspiration to me from an early age, and I'd known a lot about his life and work, and for many years a picture of him has adorned my room or office, though not at the moment as it happens. I suppose in compensation einstein is the hostname of my work computer at the moment.

Anyway I was also for this reason aware of Arthur Eddington, and his famous experiment to verify that gravity bent starlight. I was also aware of his famous quote regarding being one of three men who understood General Relativity. Nevertheless, it was fascinating to see the historical context of the interaction between these men, and a bit more of Eddington's life. I would recommend the programme highly, so if you're in the UK you have a few more days to catch it on iPlayer.

Great performances from Tennant and Serkis.

Posted by Colin Turner

Today my sister-in-law, Siobhan, asked me to help her debug her brand new Sky ADSL set-up. She told me her PC, with a wireless adapter, was connecting OK, but she couldn't get onto the internet at large. I was relieved I had decided to throw my laptop in my bag when I came down to visit as that would make the whole thing a lot easier. I went up to the house, opened my laptop and in a few seconds it was ready from hibernation (into Debian). Very quickly we identified the wireless network, put in the passphrase, and got connected straight away.

So the wireless was working OK as it appeared, and all attempts to surf elsewhere failed. I surfed to 192.168.0.1 and was rewarded by the router login screen, I recognised it as a Netgear device, but the usual netgear "admin" and "password" didn't cut it. Google, via my N95 supplied that "admin" and "sky" was the answer. Result, I was in and could see that there was an IP allocated to the router. I did some pings from my laptop, and found that the packet loss was almost, but not always, complete, and that sometimes DNS worked, sometimes it didn't. All my nice tools in Debian made the diagnosis easier. At this point I was assuming a line problem and that the speed would need to be restrained.

So, I hooked the laptop to the router directly with a cable, just to rule anything else out, and because I knew the tech people would fret about it needlessly, and I saw the router kept losing it's connection periodically - it would stay up for minutes, but nothing useful could be done with it then. So, it was time to ring the dreaded helpline. Siobhan had to go through the usual Q&A, including some quizzing about how her name was spelt, and then I was placed on the phone to speak to the tech-girl as it happened to be in this case. I tried, politely but fairly deperately to short-cut the endless questions by outlining what I had already discovered. She asked me if I was using Vista, I said "No, Linux", so she asked again if it was Vista or XP, and I said it was neither, it was Linux, but that the problem was between the router and their operation somewhere. So we got put on hold for about 5 minutes, and already I had the feeling of dread - I should have just lied. She came back on the phone and told me that her colleagues said Linux wasn't supported and I'd have to use another OS so she could talk me through settings.

With now rapidly mounting exasperation I explained I was already connected to the router, both on the network and with a browser, could clearly see the outbound connection going up and down like a yoyo, and that rebooting into another OS would make absolutely no difference to the problem, but of course, she was adamant, so with great irritation and reluctance I rebooted into Vista, waited for it to finally sort itself out, logged in and ... the phone line went dead.

So now, punching the floor with irritation, I ask Siobhan to phone back, get through the maze of keypresses and introductions and we get the second person. She wants to start at the beginning, but we try hard to get her to understand what we have already found. She at least seems to understand it's not an OS problem. She tells me to try another microfilter, I say fine, but I obviously can't do it just as we're on the phone, she says "That's correct, you would be disconnected". Well... yes, I got that. Anyway, we go through a few other gems and I anticipate her asking me to unscrew the BT master faceplate and plug things in there.

So we do that, with little hope, it doesn't work. I curse Vista as I try to navigate IE (I don't even dare use firefox in case they moan about that). Again, we call, the same procedure, the third person, who talks to Siobhan only, and is happy without further technobabble, to finally redirect us to the "Solution Centre" or some such. Siobhan begins to talk to operative number four, and passes me over.

Joy... She understands the problem is at their end, she limits the speed on the connection, just like that it all starts to work. I thank her profusely and say it's a pleasure to finally get to someone who doesn't ask me to check if I have an electricity supply. She comes from Northern Ireland too, we share a quick laugh about it all, she helpfully suggests that she'll leave the ticket open while we check over the weekend the new "fix" is stable. Yes, I know I can be a grumpy old [whatever] about it, and that the tech people often have to speak to people who don't have a clue what they are doing, but it is exasperating when they can't go a little beyond the script to realise you might actually know how to help them fix your problem.

Way back in the early days of ISPs in Northern Ireland I found some problem, carefully excluded local causes and called the ISP to report it to them. Foolishly I told them at the time I was using WAN and LAN at the same time (Netware), to which they promptly told me that was impossible and refused to discuss the problem further. I pointed out that not only was it possible, it was all working nicely in the morning. So I called back and just lied that time.

The moral of the story seems to be, when someone asks if you are running Vista, just say "yes".

Posted by Colin Turner

Like lots of people who are too hooked into technology for their own good, I tend to receive loads of emails a day, even after the huge quantity of spam is removed. Sometimes it's easy, and fast to respond to a given email. Once in a while I get an email from someone that I really want to respond to in some detail; usually an email from a good friend I don't often see or haven't seen in a while.

Since I'm usually up to my neck in too much stuff, it's precisely this email that I leave to the side to deal with later, "when I have more time". Yes, the people who know me well already understand the problems with that. It would help if icedove stored its tags correctly in my dovecot based IMAP server. but for some reason it often doesn't. The end result is that "later", I'm still struggling to complete a dozen tasks on a list, and the email is anonymously buried in my inbox under many more recent arrivals.

That's really not good, and I'll have to make more of an effort to keep in touch with all these friends.

Posted by Colin Turner

Today I spent the day at the chemistry department in Edinburgh University demonstrating (at their invitation) OPUS with Ron Laird. I frequently receive these sorts of invitations, and that's nice, but it's a whole day to go there and back again.

I got home and checked my email, and saw dozens of emails about a conference in Japan. It seems the organiser thought it would be a good idea to set-up a mailing list, but hadn't configured it to be read-only. Ok, a pain, and basically spam. Now the real stupidity sets in, when several people start to mail the list to ask to be removed. Ok, more annoyance. Now we get the real fun, computer professionals sending "unsubsribe foo@bar.com" to the list (not the list server) so we all get to enjoy that.

I've noticed this cascade effect before, when people start doing this and everyone else gets fed up and starts repeating the mistake, so want off the list; a bit like a run on a bank. But, please people, email the originator of the spam, not the list and/or look in the list control lines to find out how to unsubscribe.

Posted by Colin Turner

I have to say, I've enjoyed the "re-imagining" of Battlestar Galactica immensely so far. When it first aired, I thought it was a pretty corny idea, and didn't watch it, but I watched an odd episode here and there, and I began to realise that this was probably the most slick, dramatic, special effects laden sci-fi show yet. It's very addictive, and one of the very few TV shows I feel I don't want to miss an episode of.

Tonight I just sat down to watch season 4, episode 3 (no spoilers, don't worry!), and was very gratified to see Cylon base ships moving in a background that contained Orion, absolutely no doubt about it, a nice touch to show Earth is near. Now Orion is the easiest constellation to see, and I admit I never paid attention to the background before, but it's just that kind of nice detail I like about the show. I'll keep looking for more constellations now!

Posted by Colin Turner