🚀 go-pugleaf

todd@uvmark.uucp (Todd Cooper) writes:
> adamg@world.std.com (Adam M Gaffin) writes:
>> richb@kronos.com (Rich Braun) writes:
>>>
>>>light's timing circuit has been adjusted.  Such events are *rare*.
>>>
>>
>>Say what you will about New York, at least they know how to synchronize
>>their lights.  At least on the avenues in Manhattan, they're timed to
>>promote driving around 30 m.p.h. -- go too slow or too fast, and you get
>>stuck at a light...  I realize the cowpath nature of streets
>>around here might make such things more difficult around here, but it is
>>astounding how out of sync lights are with each other.  

The two-way nature of most streets makes it impossible to synchronize long
stretches in both directions, but it seems that few are even synced in
one direction (Berkeley and Clarendon being notable exceptions).  It's my 
impression that sometimes lights are deliberately timed to keep speeds
down, but more often the timing is random because the lights have been 
installed one intersection at a time, with each corresponding traffic
study applying only to that intersection.  Of course, if the signals aren't
even electronically connected, then they wouldn't stay synced for long
(unless they have extremely precise timebases).

>Wasn't there a BIG press story that went on for weeks about Boston 
>computerizing all the traffic lights 3-4 years ago?  Have they done it?
>Has anyone seen improvement? Do they adjust them from a central location?

Yes; don't know; no; that was the idea.
New York was/is also undergoing a project to replace some ancient IBM
mainframe controllers with (I think) VAXes, and at the same time extend
coverage to the other boroughs (the old system served Manhattan only).