Similarly, I have noticed that a difference between American subways and European subways is that when you stop at a station in Europe, the interlocks allow the automatic doors to open as soon as the train is slower than maybe 1–2 km/h. Whereas in America they don't open until a couple seconds after the train has come to a dead stop.
I keep wondering how many wasted man-hours this adds up to over the course of a day — especially on a line that's running at capacity, so if one train is 4 seconds later then the next train will be 8 seconds later and so on.
The release of door interlock used to be as you described, but now in Europe train stop is controlled prior to releasing it. The delay you note is generally tied to the reaction time of the ATP computer in charge of authorizing doors opening. Some of these could be slow (sometimes 5s). Today it’s less than 0.5s.
Almost every delay in computerized systems that is greater than the physical/logical sensor debounce interval (ie. on the order of milliseconds) is due to programmer error.
It plagues software literally everywhere. We need to stop!
Every gas pump, payment terminal, elevator door is a testament to poor implementations of event driven state machinery...
I am with you in this general rant. However, in this particular case, the delay was a consequence of the introduction of computers and asynchronous telecommunications. I was not around at that time but it seems nobody anticipated that putting all the logic in a computer would result in worst performances as opposed to older designs were computations were essentially carried out by relays wired together. Relay logics worked in parallel, when CPU have to do things sequentially.
Fun fact: some operators around the world refuse to use computer-based systems for some applications because of that, and they are getting a strong support now because of cyber security threats. A relay is hard to hack :D
And every new tech is shinier and slower than the old generation. Nobody pays any attention to latency as a part of UX (except people making pro-photo and music gear, it seems - even gaming these days accepts pretty large latency as a part of the experience).
>Similarly, I have noticed that a difference between American subways and European subways is that when you stop at a station in Europe, the interlocks allow the automatic doors to open as soon as the train is slower than maybe 1–2 km/h
No, this must only happen with older rolling stock. Newer rolling stock both from European and Asian manufacturers operates like your "American" example
That definitely doesn't happen in Chicago. Nothing drives me nuts like pushing on the back door of a bus when it's completely stopped only to jam it cause I jumped the gun by like .2 seconds because the light hadn't come on yet.
That and also driven so many European subways (most called METRO) you never seen cars as overcrowded as in USA and barely people leaning over the door! In USA its normal to lean against the door because people are exactly gotten used to the fact the train will stop first before the door gets open.
> That and also driven so many European subways (most called METRO) you never seen cars as overcrowded as in USA and barely people leaning over the door! In USA its normal to lean against the door because people are exactly gotten used to the fact the train will stop first before the door gets open.
The only places in the EU that have public transportation on a scale comparable to NYC are London and Paris. I haven't ridden in Paris in a while, but in London, you absolutely do see cars that are overcrowded with people leaning against the doors.
I think the train complains if the doors are being leaned on - you certainly get an announcement from the driver, so I reckon the train won't go if that's happening.
Ehh, if you changed it I bet people would learn really quickly not to lean on doors. Most people leaning on doors are going to be young people with decent reaction times and balance so I wouldn’t even expect injury more like surprise. It would probally be reasonable to run a publicity campaign to reduce surprise as people don’t like being supprised.
Further to your point, missing a few seconds due to this will reduce capacity but will not create delays, since this time will be factored in the timetables with a margin.
I keep wondering how many wasted man-hours this adds up to over the course of a day — especially on a line that's running at capacity, so if one train is 4 seconds later then the next train will be 8 seconds later and so on.
(The same seems to be true of elevators.)