John B wrote:
On 24 Aug, 11:31, Michael Hoffman wrote:
The poor design element is that you can't validate on a bus, thereby
making validation actually an issue instead of the non-issue it ought
to be (yes, I know the excuses for this; however, it'd be trivially
easy to make bus-based readers connect to the base by GPRS every five
minutes to exchange relevant data with the central system).
I think we have different definitions of "trivial."

TfL has a secure private network in place to link the fixed Oyster
readers to their central server. Companies have been able for many
years to provide their employees and their employees' devices with VPN
access over GPRS to their secure private networks. Integrating mobile
devices with electronic peripherals is more or less a matter of plug-
and-play.

I'm missing the 'non-trivial' element, here...

Yes, you are.

Even writing a specification for such a system detailed enough to be
implemented would not be trivial. I'd say that anything that involves
the deployment of new hardware and software across thousands of buses
will not be trivial. Almost-real-time wireless communication only adds
to the complexity.
--
Michael Hoffman