LU end-to-end journey data
 

On Tue, 1 May 2007, Paul Corfield wrote:

On Tue, 01 May 2007 18:24:50 +0100, James Farrar
wrote:

On Tue, 1 May 2007 15:05:40 +0100, Tom Anderson
wrote:

On Mon, 30 Apr 2007, wrote:

On 23 Apr, 20:47, Paul Corfield wrote:
On Mon, 23 Apr 2007 20:09:41 +0100, Tom Anderson
wrote:

Also, am i right in thinking paper tickets either don't have a unique
ID on them, or that this isn't recorded by gates? If not, LU should
already have had this data.

Some magnetic tickets did have unique numbers but they were a very
small part of the overall population. The vast majority did not and
although they were counted by type at each gate you could not follow
"ticket 123456" through the system.

Presumably there is some way to identify a ticket within a station - so
that things like gate zig-zag can be identified?

A way to do that would be for the gate to write on the ticket that it's
just been used for exit at that station, and refuse tickets that have been
so marked. This is probably actually simpler, as it avoids having to have
the gates share knowledge of which tickets they've seen.

I believe magnetic tickets hold the details of the last three uses.

No they do not. They do not have sufficient capacity to do so. If a
ticket is valid and is accepted then certain key fields are updated. It
is this revised data that allows things like passback and zig zag to be
detected.

I'm a genius!

Hang on, what's zig-zag? I thought it was a synonym for passback, but i
guess not.

tom

--
Through the darkness of Future Past the magician longs to see.

LU end-to-end journey data
 

On Wed, 2 May 2007 15:54:07 +0100, Tom Anderson
wrote:

On Tue, 1 May 2007, Paul Corfield wrote:

On Tue, 01 May 2007 18:24:50 +0100, James Farrar
wrote:

On Tue, 1 May 2007 15:05:40 +0100, Tom Anderson
wrote:

A way to do that would be for the gate to write on the ticket that it's
just been used for exit at that station, and refuse tickets that have been
so marked. This is probably actually simpler, as it avoids having to have
the gates share knowledge of which tickets they've seen.

I believe magnetic tickets hold the details of the last three uses.

No they do not. They do not have sufficient capacity to do so. If a
ticket is valid and is accepted then certain key fields are updated. It
is this revised data that allows things like passback and zig zag to be
detected.

I'm a genius!

I think we knew that already Tom. Thou art truly a man of many talents.

Hang on, what's zig-zag? I thought it was a synonym for passback, but i
guess not.

And *as* you are a genius I am sure you'll be able to work out what a
zig zag is ;-)
--
Paul C


Admits to working for London Underground!

LU end-to-end journey data
 

On Wed, 02 May 2007 18:38:56 +0100, Paul Corfield
wrote:

And *as* you are a genius I am sure you'll be able to work out what a
zig zag is ;-)

Zig-zag is three uses (in-out-in or out-in-out or purchase-in-out) in
quick succession, right?

LU end-to-end journey data
 

On May 2, 7:05 pm, James Farrar wrote:
On Wed, 02 May 2007 18:38:56 +0100, Paul Corfield

wrote:
And *as* you are a genius I am sure you'll be able to work out what a
zig zag is ;-)

Zig-zag is three uses (in-out-in or out-in-out or purchase-in-out) in
quick succession, right?



What use would that be to anyone? I mean, what scam would be pulled
by that means?

When I go through the barriers at Charing Cross platforms 1 - 4, just
miss my train and have to head for platform 6 for the next one, I've
never understood why the gates won't let me out. What scam is being
prevented by it?

LU end-to-end journey data
 

On Wed, 2 May 2007, MIG wrote:

On May 2, 7:05 pm, James Farrar wrote:
On Wed, 02 May 2007 18:38:56 +0100, Paul Corfield

wrote:

And *as* you are a genius I am sure you'll be able to work out what a
zig zag is ;-)

Zig-zag is three uses (in-out-in or out-in-out or purchase-in-out) in
quick succession, right?

What use would that be to anyone? I mean, what scam would be pulled by
that means?

Alice and Bob buy a ticket and go to a station. Alice uses the ticket to
pass an entry gate, puts the ticket into an exit gate, retrieves it,
passes it over the gateline to Bob, who then uses it to pass through an
entry gate himself.

It's passback, but slightly more subtle. A system which defeated passback
might not defeat zig-zag.

When I go through the barriers at Charing Cross platforms 1 - 4, just
miss my train and have to head for platform 6 for the next one, I've
never understood why the gates won't let me out. What scam is being
prevented by it?

As above. You might have punched out with the ticket, stayed on the
fare-paid side of the gates, given it to one of your filthy lowlife
accomplices, boarded a train, and let the ruffian wander off to use it
again himself.

tom

--
Oh no - I've just turned my arse into a hand grenade!

LU end-to-end journey data
 

On 2 May 2007 13:25:17 -0700, MIG wrote:

On May 2, 7:05 pm, James Farrar wrote:
On Wed, 02 May 2007 18:38:56 +0100, Paul Corfield

wrote:
And *as* you are a genius I am sure you'll be able to work out what a
zig zag is ;-)

Zig-zag is three uses (in-out-in or out-in-out or purchase-in-out) in
quick succession, right?



What use would that be to anyone? I mean, what scam would be pulled
by that means?

When I go through the barriers at Charing Cross platforms 1 - 4, just
miss my train and have to head for platform 6 for the next one, I've
never understood why the gates won't let me out. What scam is being
prevented by it?

If zig-zag is allowed, it's tantamount to allowing passback.

Example: Passengers A and B are travelling together. B has a valid
ticket, A does not. B goes through the "out" gate using the ticket,
then puts the ticket back through an "in" gate, then passing the
ticket to A who uses it to exit. Allowing out-in-out (or in-out-in,
the logic runs exactly the same) is tantamount to allowing out-out (or
in-in) in quick succession, i.e. passback.

Example: Passenger A (who has travelled by rail from long distance) is
meeing friend B (who has arrived on foot) at a gated station. A
travels without a ticket, knowing that if the gate is closed, B can
buy the cheapest single ticket from the machine, use it in an "in"
gate and then pass it to A to use it to exit. Elimination of
purchase-in-out is necessary to prevent this.

LU end-to-end journey data
 

James Farrar:
Zig-zag is three uses (in-out-in or out-in-out or purchase-in-out) in
quick succession, right?

M.I.G.:
What use would that be to anyone? I mean, what scam would be pulled by
that means?

Tom Anderson writes:
Alice and Bob buy a ticket and go to a station. Alice uses the ticket to
pass an entry gate, puts the ticket into an exit gate, retrieves it,
passes it over the gateline to Bob, who then uses it to pass through an
entry gate himself.

Right, got that.

It's passback, but slightly more subtle. A system which defeated passback
might not defeat zig-zag.

Why not? There are still two "in"s in rapid succession there, which
could be trapped.
--
Mark Brader | "Design an idiot-proof system, and the universe
Toronto | will spontaneously evolve a higher grade of idiot
| that is able to circumvent it."

LU end-to-end journey data
 

On May 3, 1:24 am, James Farrar wrote:
On 2 May 2007 13:25:17 -0700, MIG wrote:





On May 2, 7:05 pm, James Farrar wrote:
On Wed, 02 May 2007 18:38:56 +0100, Paul Corfield

wrote:
And *as* you are a genius I am sure you'll be able to work out what a
zig zag is ;-)

Zig-zag is three uses (in-out-in or out-in-out or purchase-in-out) in
quick succession, right?

What use would that be to anyone? I mean, what scam would be pulled
by that means?

When I go through the barriers at Charing Cross platforms 1 - 4, just
miss my train and have to head for platform 6 for the next one, I've
never understood why the gates won't let me out. What scam is being
prevented by it?

If zig-zag is allowed, it's tantamount to allowing passback.

Example: Passengers A and B are travelling together. B has a valid
ticket, A does not. B goes through the "out" gate using the ticket,
then puts the ticket back through an "in" gate, then passing the
ticket to A who uses it to exit. Allowing out-in-out (or in-out-in,
the logic runs exactly the same) is tantamount to allowing out-out (or
in-in) in quick succession, i.e. passback.

Example: Passenger A (who has travelled by rail from long distance) is
meeing friend B (who has arrived on foot) at a gated station. A
travels without a ticket, knowing that if the gate is closed, B can
buy the cheapest single ticket from the machine, use it in an "in"
gate and then pass it to A to use it to exit. Elimination of
purchase-in-out is necessary to prevent this


My Charing Cross example involves a travelcard. If I was meeting
someone who had travelled without a ticket, I could just pass it over
the exit gate to them, far less ostentatiously than by backing out of
the entry gate and passing it to them with someone trying to enter
behind me.

If there were no restrictions, people would just pass the ticket back,
so there would be nothing to be gained by putting the ticket through
an extra time.

Given that there are restrictions that prevent any kind of successive
passes at the same gateline, there is still nothing to be gained by
passing the ticket through an extra time, and no need to program
against it, because it's already ruled out by simpler restrictions.

I must be missing something.

LU end-to-end journey data
 

MIG wrote:

Given that there are restrictions that prevent any kind of successive
passes at the same gateline, there is still nothing to be gained by
passing the ticket through an extra time, and no need to program
against it, because it's already ruled out by simpler restrictions.

That depends on the exact restriction.

If the restriction is that a single card cannot be used twice in rapid
succession at the same gateline, then you are correct. But that is
/not/ the restriction that is actually enforced, since the system /will/
permit an entry followed immediately by an exit.

If the restriction is that a card that has most recently used to enter a
station cannot shortly thereafter be used to enter that same station,
then it prevents simple passbacks but not zig-zags, since with a
zig-zag, the most recent usage was an exit, not an entry.

Incidentally, I have been told that WMATA (the Metro in the Washington,
D.C. area) does /not/ have zig-zag restrictions on its unlimited cards.
--
David of Broadway
New York, NY, USA

LU end-to-end journey data
 

James Farrar wrote:
On Wed, 02 May 2007 18:38:56 +0100, Paul Corfield
wrote:

And *as* you are a genius I am sure you'll be able to work out what a
zig zag is ;-)

Zig-zag is three uses (in-out-in or out-in-out or purchase-in-out) in
quick succession, right?

Even the New Yorker figured that out! (Well, not the purchase-in-out bit.)
--
David of Broadway
New York, NY, USA