Oyster
 

[x-posted to uk.transport.london]

On Jan 25, 12:07*pm, "John Clausen" wrote:
On Saturday I travelled around London on tubes and National Rail using my
Oyster card instead of a paper travelcard. My advice to others thinking of
doing the same is: don't bother. Instead of capping at £7.00 it kept
charging me until there was no credit left, over £17 in total. Some credit
has been put back but I have still been charged £14.60.

There were three of us travelling together and we all got print outs at
Victoria in the afternoon. Various parts of the journey were ignored
including a trip to Harrow on the Hill where we touched out and back in
again at Marylebone and the same at Harrow. So the system thought we had
made a journey from Paddington to Vauxhall taking 95 minutes which is over
the time limit. The bloke at Victoria suggested that we hadn't been touching
in and out properly but then how would the gates have opened? The fact that
three of us had the same problem suggests that it is system failure rather
than user error. Has anyone else had similar problems? I will not be using
Oyster as a travelcard again in the near future.


If you were willing to provide the details of your journeys, then we
might be able to untangle what happened.

I think the thing that's at the root of such Oyster problems is the
out-of-station interchange (OSI) issue (which we're currently
discussing in another current utl thread).

The kinda ironic thing is that it's supposed to be beneficial for
passengers - in essence what happens is that Oyster automatically
combines journeys together, so say a Clapham Junction to Victoria
journey on NR would be combined with the subsequent Tube journey from
Victoria to Kings Cross, which could then be combined with a Kings
Cross to Alexandra Palace NR journey - in other words the whole thing
would be treated as one through journey for charging purposes.

That's great, however the problems surface because the system assumes
someone is making a through journey when they touch-in at an OSI
location (e.g. at Waterloo Underground station) within a certain time
period from touching-out (e.g. at Waterloo NR station). If they're
making a straightforward journey then that's fine. If however that
passenger spends a period of time doing something else - leaving the
station, getting some food, whatever - then the problem can be that
the overall journey 'times out', i.e. it the maximum journey time that
the system allows for the journey (which varies) is exceeded. It's at
this point that things go skew-whiff, and the system applies the
'maximum charge' which exists to discourage misuse (the exact
mechanism is a bit more complex, but that's a summary).

I've a feeling that since Oyster's coverage expanded onto NR in London
we're going to be hearing a lot more of this here (on these groups)
and also elsewhere. I dare suggest that it's more likely to hit rail
enthusiasts making atypical journeys, but it can snag others too.

Oyster
 

On Mon, 25 Jan 2010 04:32:20 -0800 (PST)
Mizter T wrote:
this point that things go skew-whiff, and the system applies the
'maximum charge' which exists to discourage misuse (the exact

I'd love to know what they think is misuse. Why should there be a time limit
on completing a journey anyway? If someone wants to sit at a station for a
while, perhaps they don't feel well or whatever, then why shouldn't they?

No doubt its just a cynical way to extort more money out of passengers.

"Train delayed? Got lost? Don't feel well? Tough, we don't care. And we'll
overcharge you for your journey while we're at it. Have a nice day."

*******s.

B2003

Oyster
 

"MIG" wrote in message

On 25 Jan, 13:03, "Recliner" wrote:
wrote in message







On Mon, 25 Jan 2010 04:32:20 -0800 (PST)
Mizter T wrote:
this point that things go skew-whiff, and the system applies the
'maximum charge' which exists to discourage misuse (the exact

I'd love to know what they think is misuse. Why should there be a
time limit on completing a journey anyway? If someone wants to sit
at a station for a while, perhaps they don't feel well or whatever,
then why shouldn't they?

No doubt its just a cynical way to extort more money out of
passengers.

"Train delayed? Got lost? Don't feel well? Tough, we don't care. And
we'll overcharge you for your journey while we're at it. Have a nice
day."

*******s.

I think the 'misuse' penalty is meant to deter dishonest people from
avoiding touching in/out, perhaps because the barriers were open or
weren't being monitored.

Presumably they think people would attempt to combine two longish,
separate journeys, which were separated by a lengthy interval, into
one short (cheap) journey (ie, by not touching in/out at the
intermediate station).

That's the justification for the timeout existing at all, which seems
fair enough if you haven't touched anywhere.

It's not a justification for applying the original timeout when you
have been touching at OSIs in the meantime and the system knows
exactly where you've been.

But their solution to this possible problem is so complex and
convoluted that it's clearly penalising many perfectly innocent pax,
and confusing almost everyone.-

The problem being solved by not resetting the timeout at OSIs is the
risk of someone travelling around all day doing brief business only at
OSIs and getting charged for one journey instead of several. This
seems negligible.

Yes, I agree -- their algorithm is at risk of spawning a nation of
raving Boltars, and swamping the Oyster helpline.

Oyster
 

wrote in message

On Mon, 25 Jan 2010 04:32:20 -0800 (PST)
Mizter T wrote:
this point that things go skew-whiff, and the system applies the
'maximum charge' which exists to discourage misuse (the exact

I'd love to know what they think is misuse. Why should there be a
time limit on completing a journey anyway? If someone wants to sit at
a station for a while, perhaps they don't feel well or whatever, then
why shouldn't they?

No doubt its just a cynical way to extort more money out of
passengers.

"Train delayed? Got lost? Don't feel well? Tough, we don't care. And
we'll overcharge you for your journey while we're at it. Have a nice
day."

*******s.

I think the 'misuse' penalty is meant to deter dishonest people from
avoiding touching in/out, perhaps because the barriers were open or
weren't being monitored.

Presumably they think people would attempt to combine two longish,
separate journeys, which were separated by a lengthy interval, into one
short (cheap) journey (ie, by not touching in/out at the intermediate
station). But their solution to this possible problem is so complex and
convoluted that it's clearly penalising many perfectly innocent pax, and
confusing almost everyone.

Oyster
 

On 25 Jan, 13:03, "Recliner" wrote:
wrote in message







On Mon, 25 Jan 2010 04:32:20 -0800 (PST)
Mizter T wrote:
this point that things go skew-whiff, and the system applies the
'maximum charge' which exists to discourage misuse (the exact

I'd love to know what they think is misuse. Why should there be a
time limit on completing a journey anyway? If someone wants to sit at
a station for a while, perhaps they don't feel well or whatever, then
why shouldn't they?

No doubt its just a cynical way to extort more money out of
passengers.

"Train delayed? Got lost? Don't feel well? Tough, we don't care. And
we'll overcharge you for your journey while we're at it. Have a nice
day."

*******s.

I think the 'misuse' penalty is meant to deter dishonest people from
avoiding touching in/out, perhaps because the barriers were open or
weren't being monitored.

Presumably they think people would attempt to combine two longish,
separate journeys, which were separated by a lengthy interval, into one
short (cheap) journey (ie, by not touching in/out at the intermediate
station).

That's the justification for the timeout existing at all, which seems
fair enough if you haven't touched anywhere.

It's not a justification for applying the original timeout when you
have been touching at OSIs in the meantime and the system knows
exactly where you've been.

But their solution to this possible problem is so complex and
convoluted that it's clearly penalising many perfectly innocent pax, and
confusing almost everyone.-

The problem being solved by not resetting the timeout at OSIs is the
risk of someone travelling around all day doing brief business only at
OSIs and getting charged for one journey instead of several. This
seems negligible.

Oyster
 

On Mon, 25 Jan 2010 05:37:26 -0800 (PST)
MIG wrote:
Presumably they think people would attempt to combine two longish,
separate journeys, which were separated by a lengthy interval, into one
short (cheap) journey (ie, by not touching in/out at the intermediate
station).

That's the justification for the timeout existing at all, which seems
fair enough if you haven't touched anywhere.

Well if you don't touch anywhere it can't apply any charge.

It seems to me its a way of fixing the ****up they've created by having
validators that are both touch in and touch out at the same time so the
system can't tell if you're finishing a very delayed first journey or
starting a 2nd journey but forgot to touch out on the 1st. If they had
seperate validators for in and out like they do with the gates then it
wouldn't be an issue. But to save themselves a bit of cash they decided to
stiff the public instead.

B2003

Oyster
 

On Jan 25, 2:10*pm, wrote:
On Mon, 25 Jan 2010 05:37:26 -0800 (PST)

MIG wrote:
Presumably they think people would attempt to combine two longish,
separate journeys, which were separated by a lengthy interval, into one
short (cheap) journey (ie, by not touching in/out at the intermediate
station).

That's the justification for the timeout existing at all, which seems
fair enough if you haven't touched anywhere.

Well if you don't touch anywhere it can't apply any charge.

It seems to me its a way of fixing the ****up they've created by having
validators that are both touch in and touch out at the same time so the
system can't tell if you're finishing a very delayed first journey or
starting a 2nd journey but forgot to touch out on the 1st. If they had
seperate validators for in and out like they do with the gates then it
wouldn't be an issue. But to save themselves a bit of cash they decided to
stiff the public instead.

B2003

It's all a conspiracy to stiff you, isn't it?
Tim

Oyster
 

On Mon, 25 Jan 2010 07:28:16 -0800 (PST)
TimB wrote:
wouldn't be an issue. But to save themselves a bit of cash they decided t=
o
stiff the public instead.

B2003

It's all a conspiracy to stiff you, isn't it?

No , its standard practice. Do everything as cheaply as possible. I was
involved, albeit on some peripheral back-end systems, in the original paper
travelcard system rolled out to newsagents back in the mid 90s (large blue
boxes if anyone remembers them). And believe me, if a cost could be cut and
got away with it would be. I see no reason to believe that the decisions
behind oyster were any different. So if they could cut the number of
validators in half and put some software lash up in its place then thats
exactly what they'd do. And have done.

B2003

Oyster
 

On 25 Jan, 19:09, Paul Corfield wrote:
On Mon, 25 Jan 2010 17:13:52 +0000 (UTC), wrote:
On Mon, 25 Jan 2010 07:28:16 -0800 (PST)
TimB wrote:
wouldn't be an issue. But to save themselves a bit of cash they decided t=
o
stiff the public instead.

B2003

It's all a conspiracy to stiff you, isn't it?

No , its standard practice. Do everything as cheaply as possible. I was
involved, albeit on some peripheral back-end systems, in the original paper
travelcard system rolled out to newsagents back in the mid 90s (large blue
boxes if anyone remembers them). And believe me, if a cost could be cut and
got away with it would be.

Oh you were involved in the PASS Agent Terminals. chortle Lucky you.

And yes they were built down to a price because the people running PASS
didn't want to incur the costs from a supplier like Cubic. To be fair
the terminal needed to be basic and simple as the transactions were only
supposed to be simple. *The scope of service assumed for the PASS
network is now far greater than back in the late 90s because there is
such a determined shift to get sales off stations.

I see no reason to believe that the decisions
behind oyster were any different. So if they could cut the number of
validators in half and put some software lash up in its place then thats
exactly what they'd do. And have done.

Well you would be wrong wouldn't you? *I identified the need for
validators at the interface points with the LUL system. When I was
involved there was no agreement about NR involvement on the scale that
is now in place. The concept for validators was to permit easy but small
scale validation for the limited numbers of people who may need to
register an entry or an exit for SVT (now PAYG) travel. *The basic logic
is "enter" or "exit" and assuming there was a valid entry it is entirely
logical to assume someone is exiting the system. *Similarly if the last
exit was a fair time in the past it is logical to assume that the card
holder is "entering" the system.

There was also the practical issue that space is at a premium at some of
the interchange points and it was feasible to install ranks of entry and
exit validators. *The point about Oyster is that you should be able to
"touch and pass" (old concept from the old days) and not worry what the
system is doing. *I'll grant you we've ended up some distance from that
admirable goal but then Oyster is being asked to do far, far more than
the original design.

Er, hang on ...

Now the reason why Oyster PAYG can't replace the travelcard is because
Oyster is being asked to do too much.

Previously the reason why Oyster PAYG couldn't replace the travelcard
was because NR wouldn't cooperate and the system couldn't be used
where it had been envisaged.

I think Oyster could cope either way; it's the accompanying decisions
that have caused the problems.

Oyster
 

On Mon, 25 Jan 2010 19:09:22 +0000
Paul Corfield wrote:
Oh you were involved in the PASS Agent Terminals. chortle Lucky you.

It was a job. I actually met the guy who did most of the programming for
them. IIRC he said they had something like 4 or 5 seperate CPUs inside them
and each had to be programmed with a mixture of some cut down version of C
and assembler. He certainly earned his money.

Well you would be wrong wouldn't you? I identified the need for

Would I?

validators at the interface points with the LUL system. When I was
involved there was no agreement about NR involvement on the scale that
is now in place. The concept for validators was to permit easy but small

Well you should have thought ahead shouldn't you. Even without NR involvement
there are still places like finsbury park where seperate in and out
validators would make things a lot less confusing and obviate the need for
this silly time out penalty charge.

register an entry or an exit for SVT (now PAYG) travel. The basic logic
is "enter" or "exit" and assuming there was a valid entry it is entirely
logical to assume someone is exiting the system. Similarly if the last
exit was a fair time in the past it is logical to assume that the card
holder is "entering" the system.

Well that logic has been shown not to work hasn't it.

There was also the practical issue that space is at a premium at some of
the interchange points and it was feasible to install ranks of entry and
exit validators. The point about Oyster is that you should be able to

Oh come off it. They're not that big and you could easily have 1 validator
with 2 seperate touch pads marked with IN and OUT in big bold letters.

system is doing. I'll grant you we've ended up some distance from that
admirable goal but then Oyster is being asked to do far, far more than
the original design.

In other words its not up to the job.

B2003