[Foundation-l] Long-term archiving of Wikimedia content

[Thomas Dalton]

2009/5/5 Aryeh Gregor <Simetrical+wikilist at gmail.com>:
> On Tue, May 5, 2009 at 10:32 AM, Thomas Dalton <thomas.dalton at gmail.com> wrote:
>> But rebuilding civilisation is probably not the most likely use such
>> archives would be put to (it's just the most exciting, so the one I
>> mentioned). The historical and cultural value 1000 years from now of
>> knowing what people 1000 years ago knew and thought would be immense.
>
> But if you don't postulate a catastrophic event that we can't plan
> for, like civilization ending due to an overnight thermonuclear war,
> then we don't need to plan in advance.  If Wikipedia ceases to exist
> at some point, and if at that point it looks like it would be useful
> to preserve its contents, we could preserve it more durably at that
> point.  We don't need to preserve it now.  In fact, it would be
> counterproductive: better to preserve it at the last possible moment,
> when it will contain as much data as possible.

You make a good point, but that point applies just as well to any
other time capsule plan and people still consider them worthwhile.
It's not about having as much information as possible, it's about
preserving the information as it is now. Obviously, Wikipedia has
revision histories going back years and, as long as they are
maintained, I suppose you could just make your own "archive" when you
wanted to read it. So, I guess all we really need to do is ensure we
have reliable full history dumps backed up. This kind of plan would
only be needed if we stopped storing histories (which, before anyone
says otherwise, we are not required to do by either GFDL or CC-BY-SA).

However, most information isn't lost because of disaster, it is lost
because people don't think they need it any more and delete/destroy
it. Can we trust whoever is around in the future to continue to
preserve the history dumps they've backed up?