4 ms·
This isn't so much a criqitue of the theory of "updateable content" as it is about the way in which it's updated. Nick lists the benefits, but, losing the "perm
by philjr 18y ago
This isn't so much a criqitue of the theory of "updateable content" as it is about the way in which it's updated. Nick lists the benefits, but, losing the "permanence" of a hardcopy of a book works both ways.
In the printed forms, books can go out of print and become extremely rare heightening their value and making it extremely difficult for them to be acquired.
If someone recognises the problem and introduces a revision system to the system in which books are updated, this is effectively sorted. Anyone could potentially rollback.
Whether the updateable book or a revision history of changes is on the cards is anyones guess however. I'd imagine the publishers would prefer this wasn't the way it was implemented.
Also, I believe the example given is more censorship than anything wrong with some sort of auto-update protocol. If a book is censored, for example, two weeks after it's published and re-published thereafter, the number of uncensored copies in distribution is probably extremely small. I can only imagine in this scenario using a technology like the Kindle, that someone would simply copy the original manuscript and redistribute it themselves.
Again, open formats make things like this easy to solve. Closed DRM-like formats are going to make this hard. Guess which one we hackers want? :)