In message , at 11:46:39 on Mon, 12 Mar
2012, Stephen Sprunk remarked:

There are hundreds of thousands of possible fares, multiplied by over a
dozen different discount rates. The databases and algorithms "belong to"
the train companies, and they are the sole customer for the equipment
and have also decided that they don't need to buy it.

They've obviously given all that information to the companies that make
their current terminals, so what would be different about telling _those
same companies_ to do the same for terminals with mobile data capability?

Nothing different, it's just that they would need to place an order with
that company, and as far as I can see they have no intention of doing
so. So having the data doesn't help much.

let alone deployed. And the previous generation has at least 10 years
life left in them.

They were obsolete the day they were purchased; how long they're
capable
of meeting obsolete needs before turning into paperweights is not
terribly relevant.

It is, when there's no money to replace them,

... which is why savvy customers look at the ROI: you pay for capital
assets with the cost savings from employing those assets.

They don't see a cost saving, only a cost increase (all those mobile
data bills).

We can debate _how large_ the cost saving will be, and therefore whether
it is worth solving, but it is not zero.

Indeed, and I'm saying the saving might well be less than zero (ie a
greater cost). The recent introduction of card-based terminals to pay
for refreshments on board the trains I catch to London has been
scrapped, and they went back to accepting cash only, citing the cost of
operating (including leasing, probably) the terminals.

You are promoting a classic "solution looking for a problem to solve",
and there isn't one.

I clearly identified the problem to be solved and was told there was no
solution; now that I identify the solution, you claim there is no problem?

The problem you identified exists, but is not serious enough that it
needs a solution.

Unless the train is so full no-one else can squeeze aboard, or the car
park is completely full, the opportunity cost is zero.

Opportunity costs are _never_ zero. They may be small, perhaps not
worth worrying about, but once dismissed such costs have a nasty
tendency to grow and surprise you later.

There's even a possibility that opportunity costs are less than zero.
People dislike travelling on crowded trains, and the availability of a
few "spare" seats is a benefit. The existence of "First Class" fares,
which on many routes provide little more than a guarantee of a seat, are
proof of this.

There is also opportunity cost in not accepting money
from potential paying customers who only have a debit card.

You can use cash as well. Although the chances of (eg) needing paid car
parking and not having plastic is pretty small.

And if someone has only a debit card and no cash, you're going to throw
them off the train? What is the cost of doing that--particularly the
cost in PR?

There's never been a question of not-accepting debit cards. Some now
deprecated *versions* of debit cards are not accepted, but this is well
known and anyone boarding a train with no means to pay (in most cases
having deliberately not bought a ticket beforehand) isn't going to score
any points with the public.

Obviously, one would need some analysis to figure out if these costs
were more or less than the cost of better terminals. If more, you buy;
if not, you don't.

And they aren't (buying).

Lots of folks don't buy things that will save them money because they
either haven't done the analysis or don't think they have the money--but
one of the reasons they "don't have" the money is that they're not
adopting cost-saving technologies and processes, so it's a vicious cycle.

That may apply to some situations, but not this one. It's a sledgehammer
to crack a nut.

Also, since this is a gaping security hole just waiting to be
exploited by the masses,

Clearly it isn't.

There is no debate he offline credit/debit payments _are_ insecure.

They are secure (in as much as anything can be - one day someone will
rob Fort Knox), but there's a very small risk of the payment being
"bounced" if the cardholder has no funds.

(Digital cash systems can be made secure, but that's not what we're
talking about here--and it's shockingly difficult, even in theory.)

It might end up being worth the upgrade just to not have to do the
analysis--or to avoid the risk of making the papers when a few million
teens figure out they can easily beat your system and ride all over the
country for free without getting caught. All it takes these days is one
Facebook post that goes viral.

They would ride for free (avoid stations with barriers, and trains with
ticket inspectors) rather than have their credit rating trashed the
first time they tried this on (their account going into an unauthorised
overdraft that they then walk away from).

Someone (you, I think) said that the current terminals accept _any_
credit card presented. That means I can just print up my own cards with
random numbers and ride for free

No, because you'd have to make a Chip (for the C&P) that validated
correctly. So far, there's been no reported incidence of someone being
able to counterfeit the chips (and I'm quite sure a lot of people have
been trying for years).

If you don't understand that very basic parameter, no wonder the rest of
your posting doesn't make sense to us in the UK

--and the carrier doesn't know until the
terminal uploads the card information later, long after I'm off the train.

Using _my_ credit/debit card for such a fraud would be silly.

So it has to be a stolen one, where you know the PIN.

the regulators have been steadily reducing the cost of roaming.

Our regulators don't care since _customers_ don't pay for roaming;
that's a problem for the carriers to hash out between themselves.

Really, so I can get a refund for that $1/minute I was charged when
roaming in the USA last year?

I think some context got snipped: US customers do not pay for domestic
roaming. You were neither a US customer nor doing domestic roaming;
what you pay is up to your non-US carrier, and US regulators obviously
have no power over that anyway.

s/_customers_/_domestic_customers_/ makes it clearer

I remember the stories, like people being charged vast roaming fees to
call from (eg) Minneapolis to St Paul.

The other end of the call had no effect on roaming charges; what
mattered was the "service area" you subscribed to and from which
carrier. So, if you lived in NYC, traveled to Chicago and made a call
to a "local" number, you would be charged roaming fees for being out of
your service area plus the LD fees from NYC to Chicago.

Indeed, and I should have made it clear that in my example it was
implied that someone's "service area" would only have been one of the
twin cities, and not both. With predictable consequences when they
picked up a tower in the wrong one.
--
Roland Perry