From bob@inforamp.net Sat Jan 6 15:25:37 1996 Xref: elastic tor.general:23289 Path: elastic!exorcist!lethe!abyss!news2.compulink.com!news.compulink.com!news1.toronto.fonorola.net!news1.toronto.istar.net!news.toronto.istar.net!inforamp.net!ts7-07.tor.inforamp.net!user From: bob@inforamp.net (Bob Richards) Newsgroups: tor.general Subject: TTC Tragedy:Blame Everyone Else! Date: Fri, 05 Jan 1996 00:23:54 +0000 Organization: Cosmic Abuse Lines: 31 Message-ID: NNTP-Posting-Host: ts7-07.tor.inforamp.net OK,the facts are in... 1- The driver ran at least 3 red lights!.There is a lot of hot air from his lawyer and the union about why it wasnt entirely his fault,but the facts are that he ran 3 red lights. 2- There was a malfunction at a signal downstream that caused the train he hit to be stopped in the tunnel when it was hit. However,the driver ran 3 red lights to hit him. 3- My source,Deep Tunnel,sez that even with a malfunctioning trip arm,there is no way any SANE driver could have gone so fast as he did at the time of the collision.No sane driver uses full throttle in the face of red lights ,even if you think the "lunar" will kick in a split-second before you hit it.(the "lunar" or white light,notifies the driver that he is in an area where the lights only change from red to yellow if you stay within A certain speed limit.A timed zone,so to speak) In some circumstances,the best thing for the driver to do is to just slow down a little.Not floor it. 4-There are certain areas within the subway system that have chronically malfunctioning signals(between Sheppard and Finch northbound,for example). 5-The bottom line is,there are some weak areas in the Subway signal system,though not many.The driver fucked up ,big time,aided by a trip arm that did not do what it's desingned to do.However,the guilt is with the driver.Deep Tunnel and the rest of the gang want to make sure you know that. "Alas,the most difficult thing to defend is the true asshole's right to freedom of expression"- Bob Allisat From msb@sq.com Tue Jan 16 08:02:09 1996 Xref: elastic tor.general:24679 Newsgroups: tor.general Path: elastic!exorcist!lethe!gts!geac!sq!msb From: msb@sq.com (Mark Brader) Subject: Re: TTC Tragedy: Blame Everyone Else! Message-ID: <1996Jan13.095406.29456@sq.com> Summary: Clarification about trip arms and red signals. Organization: SoftQuad Inc., Toronto, Canada References: <726_9601112206@tor250.org> Date: Sat, 13 Jan 1996 09:54:06 GMT Lines: 63 > > He ran three red lights, but only two stop signals. One set of > > lights was a repeater because of the sharp curve of the tunnel. > > He ran 3 red lights...2 signals...both of which had malfunctioning > trip-arms, which should have stopped the train if they were properly > maintained. The second one would not have done so in any case. The reason is that the signal system only detects which blocks have trains in them. If two consecutive blocks are occupied, the logic assumes that they are occupied by the same train, which is at this moment crossing from one to the other. Accordingly the trip arm between the two blocks is kept down, so that it won't trip the train itself. So long as no red signals are passed, there is always an unoccupied block between two trains; this block provides enough distance for a train moving at full speed to come to a stop. It's the red at the entrance to this block that is critical; the next red is probably too close to the train ahead to provide sufficient stopping distance anyway. Here is a train moving along: #1 #2 #3 #4 #5 #6 1 |----------|----------|----------|--######--|----------| G Y R R Y R 2 |----------|----------|----------|-------###|###-------| G Y R R R R 3 |----------|----------|----------|----------|--######--| G G Y R R R See how the block to the right of the leftmost red is always an unoccupied block. In diagram 2, signal #5 is red but *cannot* have its trip arm raised. If it did, it would trip the train passing over it. All the other reds in each diagram can and do have their trip arms raised. Okay, now note that I've made signal #6 red, so this train is now stopping and waiting. Along comes another train, which has to stop at signal #4: 4 |----------|----------|---######-|----------|--######--| Y R R R R R Only, for some reason, it doesn't, and proceeds into the next block: 5 |----------|----------|----------|-######---|--######--| G Y R R R R The system *cannot* distinguish diagrams 2 and 5, so the trip arm at signal #5 goes down in diagram 5. Whether it malfunctions or not, it does not protect the forward train in this situation. The red light is still there to be seen by the driver, though. And, of course, protection by tripping at this point shouldn't be needed, because a runaway train passing #4 should have been tripped there. -- Mark Brader, msb@sq.com | "In the affairs of this world men are saved, SoftQuad Inc., Toronto | not by faith, but by the want of it." -- Franklin My text in this article is in the public domain. From msb@sq.com Tue Jan 16 08:03:11 1996 Xref: elastic tor.general:24680 Newsgroups: tor.general Path: elastic!exorcist!lethe!gts!geac!sq!msb From: msb@sq.com (Mark Brader) Subject: Re: TTC Tragedy: Blame Everyone Else! Message-ID: <1996Jan13.110311.29842@sq.com> Organization: SoftQuad Inc., Toronto, Canada References: <4cr2lp$4sb@ionews.io.org> Date: Sat, 13 Jan 1996 11:03:11 GMT Lines: 112 David Collier-Brown (davecb@cs.yorku.ca) writes: > You should understand that these aren't the red lights that we're > all familiar with. They're more like the rows of red lights that > ``count down'' to the start of a drag race! > > You're ``on time'' if you pass the lights just as the reds all go **out**. This is correct, and needs to be understood (especially by people both in the press and in this newsgroup who are screaming about "running red lights"). But it's only part of the story. The basic meaning of a red signal is, of course, "do not pass this signal now". It can be red in normal operation for any of several reasons: 1. This is the end of the line. 2. There is a train ahead. 3. You will have to wait for a switch to change. 4. Your train's scheduled time to leave this station has not yet arrived. (Applies at the start of each trip and at certain intermediate stations.) 5. Control wants your train to stop. and 6. There is a speed limit on this section, due to a curve and/or downgrade, and if you're moving at the proper speed, the signal will change just before you reach it. Now, the thing is that a signal which is red for reason 6 fits David's description quoted above (and elucidated in the rest of his message). Such a signal *can and should* be approached at the normal operating speed for this section of route -- which for the section in question was 30 mph (48 km/h). *But* if the signal is red for some other reason, then the driver must expect to *stop* at the signal and, of course, should not approach it at speed. So it is important for the driver to know when the red is merely a timed one (reason 6) and when it's red for some other reason. This information is given by a white light under the *preceding* signal. If the white light is present, the next signal can be expected to change. If the timed sequence extends over multiple blocks, then each signal before the last one has the white light. If the line is clear, the driver will see a yellow and white at the start of the timed section, then one or more red and white signals, which each change to yellow and white as they are approached; then a final red, which changes to green as it is approached. If you ride at the front of a train over any downhill section of the system -- such as the descent to Old Mill, or to York Mills, from either direction, or most of the southbound parts along Yonge, University, and Spadina -- you can see this happening. Passing any of the red and whites before it changes will trip the train, as, of course, will the final red. The entire section from St. Clair West to Dupont is timed in this way. The signal on leaving St. Clair West southbound is never green, but is yellow and white if the line is clear. Suppose that a driver sees a yellow at this location (signal SP77) and *mistakes* it for a yellow and white. Then the next signal (SP71) will be red, but he'll approach it at normal speed (30 mph). And SP71 is the one whose trip arm failed to do its job. The article in the Star that ran under the sensationalistic headline about "running 3 red lights" actually contained a reasonable attempt to convey what I have been writing about here -- but, I think, not a good enough attempt to be easily understood by anyone who didn't already understand it. (It didn't help that they reversed red and green at one point!) And it also gave a strong implication that the driver in question *did* make exactly the mistake at SP77 that I described in the previous paragraph. As far as I know, this has not been authoritatively confirmed, but it certainly makes sense to me. Suppose that this happened. Now, an experienced driver would know that SP71 is not the last signal in the timed series, and therefore would be red and white, not red, if it was red merely because of the speed limit. The appearance of a plain red would set off mental alarms. But a driver two days on the job probably wouldn't yet have that detailed knowledge of the individual signals, or not enough to be confident in, so he wouldn't react that way. The appearance of a *further* red at the next signal, SP65 (and its repeater), should again have set off mental alarms. But by this time it was really too late anyway; the stopped train was just ahead. This explanation is NOT meant to excuse the driver if he did in fact mistake a signal through carelessness or inattention. Not making such mistakes is an essential part of the job, after all. But even if that is what happened, let us not confuse the picture with nonsensical suggestions of idiotic driving. In particular, I want to slap down the remarks about drivers trying to break speed records and charging at full power downhill. Even the least experienced driver would know enough not to do that. You can watch them -- on downhill sections, coasting is normal practice. If they applied power, they'd get tripped! The trick is to switch from power to coasting at the right place when the downhill section is reached, and there are markers in the tunnel to advise on that. I have been simplifying in one respect both here and in the other article I just posted (about signals and train stops). What I said is correct for signals with a single color-light head; but there are also signals with two heads. These show G/G instead of G, Y/G instead of Y, and R/R instead of R. SP77 is actually of this type, so the alleged confusion was really of Y/G with Y/G/W. The differences between the two types are irrelevant to this posting. -- Mark Brader "How can we believe that?" msb@sq.com "Because this time it's true!" SoftQuad Inc., Toronto -- Lynn & Jay: YES, PRIME MINISTER My text in this article is in the public domain.