67 |
is to be downloaded from. The URI of a page must never change, |
is to be downloaded from. The URI of a page must never change, |
68 |
even if it moves to a different server. |
even if it moves to a different server. |
69 |
|
|
70 |
|
Instead, identifiers must be *self-verifying*: after downloading |
71 |
|
an alleged version of a page from some server, it must be |
72 |
|
possible to *check* that it is really a version of that page. |
73 |
|
The simplest way to do this is to use a cryptographic hash |
74 |
|
to identify the content. After downloading the page, the client |
75 |
|
will check it against the cryptographic hash embedded in |
76 |
|
the page's identifier. |
77 |
|
|
78 |
In 1996, Tim Berners-Lee [name-myth]_ argued that |
In 1996, Tim Berners-Lee [name-myth]_ argued that |
79 |
using location-independent identifiers |
using location-independent identifiers |
80 |
is not viable on a global scale: |
is not viable on a global scale: |
91 |
|
|
92 |
.. Proposal: A location-independent Web (closest thing is Freenet (ref)) |
.. Proposal: A location-independent Web (closest thing is Freenet (ref)) |
93 |
|
|
94 |
|
We propose, then, to build a location-independent Web |
95 |
|
based on self-verifying identifiers. |
96 |
|
|
97 |
|
The project that is currently closest to this goal is Freenet, |
98 |
|
a XXX |
99 |
|
|
100 |
.. Benefits of hash-based addressing: |
.. Benefits of hash-based addressing: |
101 |
- Pages easily movable between servers |
- Pages easily movable between servers |
102 |
- Data accessible as long as anyone keeps a copy |
- Data accessible as long as anyone keeps a copy |