I'm not in the target market, but https://sendsocial.com/how_it_works doesn't really say explicitly that the sender address cannot be read from the barcode. Obviously, most humans cannot read barcodes, but that's not very good security.
This does require a bit of thought to solve, since you most likely don't want to discard the return address. A bit of crypto can solve this, but that puts quite a bit of work on the delivery service.
(On an unrelated note, what's with forcing every page to https? It doesn't hurt, but it's a bit wasteful.)
Just confirmed this one with one of the founders...
"that puts quite a bit of work on the delivery service."
This is exactly what has been done - the barcode translates to a consignment number that is useless without access to myHermes' database. This is in virtue of them having recently incorporated new barcode scanners that link them into this system.
Comments
I'm not in the target market, but https://sendsocial.com/how_it_works doesn't really say explicitly that the sender address cannot be read from the barcode. Obviously, most humans cannot read barcodes, but that's not very good security.
This does require a bit of thought to solve, since you most likely don't want to discard the return address. A bit of crypto can solve this, but that puts quite a bit of work on the delivery service.
(On an unrelated note, what's with forcing every page to https? It doesn't hurt, but it's a bit wasteful.)
Just confirmed this one with one of the founders...
"that puts quite a bit of work on the delivery service."
This is exactly what has been done - the barcode translates to a consignment number that is useless without access to myHermes' database. This is in virtue of them having recently incorporated new barcode scanners that link them into this system.
[EDIT: dunno about the https overkill, though]
Hmm. That's cool; the numbers would need to be per-package, not per-sender, and not sequential, but I assume those have been taken care of.
Cool.
Hmm. That's cool; the numbers would need to be per-package, not per-sender, and not sequential, but I assume those have been taken care of.
Cool.