![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
kinzelFriends, Writers, and countryfolk,
lend me your POS terminals....
beware your objects, your comforts, your sales items,
for if they carry an RFID chip, they can really screw things up.
Passports, too, could do this, if they happen to use RFID...
the article I saw shows that someone went looking for problems in the technology... and found them. Imagine if you will that by buying and messing with a chip from a cheap pair of pants at WalMart you could have them selling TVs for $1.99.... or hang the system that communicates sales volume through the stores.
Otherwise today is an important day: it's the day we put away the rest of the Yule/Christmas stuff until the fall. That's the fall of the year and not the fall of merchandising as we know it.
lend me your POS terminals....
beware your objects, your comforts, your sales items,
for if they carry an RFID chip, they can really screw things up.
Passports, too, could do this, if they happen to use RFID...
the article I saw shows that someone went looking for problems in the technology... and found them. Imagine if you will that by buying and messing with a chip from a cheap pair of pants at WalMart you could have them selling TVs for $1.99.... or hang the system that communicates sales volume through the stores.
Otherwise today is an important day: it's the day we put away the rest of the Yule/Christmas stuff until the fall. That's the fall of the year and not the fall of merchandising as we know it.
![[identity profile]](https://www.dreamwidth.org/img/silk/identity/openid.png){kind=small}
GIGO, anyone?
2006-03-16 16:15 (UTC)Anytime there is input to a system, you need to take what precautions you can. Part of the responsibility of systems/software engineers has to be aimed at checking data at the edges, and limiting the effects when something isn't quite right. What they showed, as far as I can tell from the reports I've seen, is that an RFID tag could supply a system with data that it didn't handle correctly due to overflows and such. No surprise there, although I like to think that someone who has been through a design review with me would put more checking into the interface (e.g. buffer overflows are a well-known error, and should never get past design, code, and test into operational code).
Admittedly, I suspect the RFID protocols (which I have not reviewed) probably lack basics such as error-checking code, length, and so forth. We know how to design good protocols, but somehow every new one seems to have to reinvent the mistakes we've seen before.
Okay, you caught a hot button, and I'll quit huffing now. It's a concern, but I have to admit, there are quite a few inputs to software now - and plenty of them leave holes for the virii to get in. I'm waiting for whole-house integrated networks with distributed cooperative processing, and finding out that bringing a Panatonic watch into a Sani house sets off a war of the bootkits! And that one will be intentional.
RFID just makes another convenient delivery medium.