Burn the Books
So, as I predicted when I booted all of my subscribers and converted this blog without copying old posts, my Google PageRank has plummeted (to 3). I have actually brought back most of the old posts, but the URL format is a bit different (although the file at http://www.netcrucible.com/blog/2002/12/22.html exists on the filesystem, WordPress returns a 404). So more than 99% of post permalinks over the past 6 years are broken. Watching the 404s pile up in my error log is heart-wrenching; I feel like Li Si presiding over the burning of the books.
I recently discussed this issue with Jeff Sandquist. In the context of “e-mail retention policies”. Some companies have such policies, which are really “e-mail deletion policies”. Lawyers think it’s a good idea to have clear and consistent processes for deleting e-mail, so that people are not tempted to delete e-mail in ways that arouse suspicion. On the other hand, people like me argue that the “delete” button is obsolete anyway, and deleting things is anti-human and destroys institutional memory. Jeff tends to be rather pragmatic, arguing that “it’s not that big a deal when you get used to it”. And he’s right. But I prefer to argue from a religious standpoint. As I argued in “Renmin Voice“, one of the two fundamental principles of semantic web is that people’s voices are indelible. Or, as this photo of Google’s master plan (see, even lies are preserved!) jokingly states, the real master plan is “don’t erase”. This week, Qwest communications called for mandatory data retention policies at online service providers; and in this case they really do mean retention (not deletion). Qwest’s reasons were exactly the ones I used in defending Google, “When Privacy is Bad“.
The other fundamental law of renmin voice is that voices are audible — that is, no artificial barriers to your voice being heard by someone who has ears and wants to hear what you have to say. Again, Jeff is right when he says that a private blog is still a blog. But speaking religiously, anything that gets in the way of future humans hearing what you want to tell them, is unethical. Depending on the circumstances, such barriers could include DRM, security ACLs, and proprietary closed formats for data like social networks.
August 23rd, 2006 at 11:19 am
Well me thinks your religious edict regarding Indelible needs some qualifications. A world with no delete button is a world full of spam, garbage and inanity. The delete button is the force of evolution on information. Delete it at your own risk.
August 23rd, 2006 at 8:24 pm
Yeah, you should only look at what you want to look at — garbage should not be in your face; or even in any folders that you have ready access to. But getting rid of it entirely, forever, I dunno..
Take spam for example. I agree there should be a way to say, “this is spam; never show me again”. But that’s not the same as deleting it — in fact, if the spam is marked as spam, that makes future spam filters, collaborative filtering, more powerful. Another way to look at it is, “sure, there should be a delete button, but always have the ability to undelete (or retrieve) if you need/want to”.
OTOH, I think it’s fine to delete machine-generated gibberish. But anything any human ever says/states/opines should be preserved for all of history.
August 24th, 2006 at 6:11 am
When I heard Tim Berners-Lee’s early exposition of RDF my first take was that there weren’t enough quarks in the universe to store all this indexing, let alone the underlying content. I think that was a misplaced fear and that we can in some sense retain all this lore in digital form.
I’m for having compression schemes for this storage so long as they can always be re-expanded so we can always look back at how many ways we’ve been able to advertise how to get a hard on.
Love.
August 25th, 2006 at 5:54 pm
For speech/e-mail, though, it is easy. Assuming a human can speak/type at an upper bound of 150 wpm (very generous), and using almost no compression, and assuming that we spend 8 hours per day generating language — everything you speak from age 3 to 83 can fit in less than 4GB.
September 21st, 2006 at 1:20 pm
Oh, on the subject of this…
I was wanting to read your XSLT for OPML article linked from the front page, and it’s not there anymore.