As is probably obvious, I’m no fan of heavy-handed solutions, but if a solution’s got to be heavy-handed (as some, sadly, must), it should at least have the ghost of a chance of working.

The addition of an expiry date to our files makes excellent conceptual sense, much the same as every calculus professor I ever had who used the accursed phrase “…and therefore it follows…”.

The problem is that it’s far, far more than a small technical change. I’ll spare you the details, because they’re at best tertiary to the discussion, but the file systems in use today cannot be changed to allow data to simply disappear as of a date.

Now just watch – I’m going to read your full paper, and I’m giving even odds that you’ll have addressed that by conceptualizing the solution far enough into the future, when most of us will be using file systems as yet unimagined, that what’s impossible today will be trivial then.

If so, please consider me to be abashed in advance.

The technical details are moot, of course, because what’s important is your actual point: We have to confront the question of intelligent and reasonable data preservation. I couldn’t agree more. Advances in the way we handle data can only be made if we’ve confronted the issue, and I thank you for having caused me to do so.

More seriously, in my paper (which is available online) I evaluate other more legal proposals and suggest – you may like this – that they are all too hevay handed. What I suggest, namely that basically we add expiry dates to ou files seems to me a very small technical change, but one that may subtly forc eus to confront the question of whether we want to preserve something and for how long mnore so than we do now. As I argue in the paper I try to keep the law as limited as possible.

Hope that addresses some of your concerns…and yes, LSE taught me that…among other things,

VMS