How do you think EMV cards work?
They are significantly more complex than just a record in a database and carry out cryptographic operations on the card! When you add your card to your favourite wallet app, your phone is then emulating the chip.
That's because of a need for extra security to prevent people just copying them. All they in principle do, though, is to transfer the card number and expiry date and a few other bits of data. The processing is done on the back end.
It isn't like Mondex (remember that?) where the balance was stored on the card. Or Oyster, where it still is (as well as in the back end).
I think fundamentally we may disagree here.
I think we do, yes.
This requires every gateline to have a stable internet connection 100% of the time, and massive scaling of centralised infrastructure up to cope with a peak time rush of commuters, and down again in quieter hours. You suddenly need a lot of very expensive IT infrastructure to make this work at all, let alone well.
That's just about doable now. In 5 years it'll be easy. In 10, it's a given.
Northern's rentathugs, for example, are already doing ticket checking using their phones which do an immediate online check.
As I've mentioned before, there is a reason that it takes time for your Oyster history to update, there is a reason that TfL batch charges you for journeys at the end of the day rather than when you tap in on the bus.
The former is because Oyster is 15 years out of date - it's mid-2000s tech, not 2020 tech. The latter is so capping can be applied; putting the transactions on then refunding at the end of the day would be messy and poor customer service, and would leave you out of pocket until the cap kicked in.
By making everything simply a reference to something in a database, you lose the ability to validate the information at the edge and thus a lot of redundancy.
But you gain massive flexibility. Most notably, a lost ticket can be reissued easily - just print off another reference to it.
Even e-tickets, as I understand, are signed blobs of data, as opposed to just a reference - they can be validated offline.
The present e-tickets are, yes, which is why I refer to "true e-ticketing" as the kind airlines do (and Megabus etc) - all you need is your 6-character PNR[1] in any form you like, plus potentially ID. Indeed, before airlines realised they could save money by doing away with check-in desks, you literally just had to walk up to the desk with your passport, and they'd type your name in and see what flight you were booked on.
There is a shift in IoT towards edge computing where although there are records in a database, devices do much of the processing outside of the cloud and report back with aggregate, instead of precise, data.
Client vs. server processing bounces back and forth all the time, roughly every 10-20 years or so. I wouldn't read anything into that.
I would suggest what you propose is a 2010s utopian vision, which the world is moving on from
No, it's something that will be made easier as we move to a situation where near-unlimited bandwidth, always-on Internet connections are available on every centimetre of the UK's land surface - and certainly anywhere there's a gated railway station. It really will not be long.
Now it might be that that PNR is transmitted via NFC from the phone, but that doesn't break the principle.
[1] PNR stands for "Passenger Name Record", largely for historical reasons, in actual fact it's the primary key of the booking record in the airline's booking system.