Last month, while I was away, Don Wiss wrote:
>> I was looking into taking Amtrak from NYC to Hartford. So I did a search at
>> their site. One of their suggestions is rather strange. For a lot more
>> money it suggests I first go to Philadelphia, then switch to a train to
>> New Haven. And then switch to one to Hartford. Some programmer must have
>> been asleep.
>>
>> See it at (the last alternative): http://donwiss.com/amtrak.jpg
And "Tim" responded:
> It seems fine to me. It meets the requirements of getting
> you there: after the previous 'good' alternative and before
> the next 'good' alternative. It would be wrong to program
> this any other way, as for some routes such a journey
> may be more reasonable and you would lose them from
> the results if you tried to code them out. ...
Actually, Don was right -- this was a bizarre itinerary. What Tim missed
was that train 94, the one to be taken for the second leg, from Philadelphia
to New Haven, *goes through New York*, the starting point.
July 19 was a Wednesday. I just tried the same query for next Wednesday
and it offered a 5:40 pm departure from New York *on train 94*, with one
change at New Haven and the same 8:05 pm arrival in Hartford on train 494.
This is 4 hours faster and $73 cheaper, with the same arrival time.
So whatever the problem was, either (a) they've fixed it or (b) it depended
in some weird way on what other options were selected for the query. But
the anomaly was real.
One more reason to use a proper paper timetable instead of those newfangled
computer thingies, that's what it is.
--
Mark Brader "Inventions reached their limit long ago,
Toronto and I see no hope for further development."
msb@vex.net -- Julius Frontinus, 1st century A.D.
My text in this article is in the public domain. |