On Exactitude in (Computer) Science
More thoughts on Sam’s Long Bets, with the emphasis on giving up on write consistency, and settle for incomplete, inconsistent views:
For example: what’s on the internet right... NOW? OK, how about... NOW? How about NOW? Firstly, its an unanswerable question... and even if we could know it, things like proper ACID transactions certainly would never help. By the time the transaction finished, the data’s no longer valid.
The way he shows how futile is the attempt to capture consistency reminded me of Jorge Luis Borges' “On Exactitude in Science” (told by Borges in Spanish). The story is elaborated after Sylvie and Bruno Concluded by Lewis Carroll:
`We very soon got to six yards to the mile. Then we tried a hundred yards to the mile. And then came the grandest idea of all! We actually made a map of the country, on the scale of a mile to the mile!'
`Have you used it much?' I enquired.
`It has never been spread out, yet,' said Mein Herr: `the farmers objected: they said it would cover the whole country, and shut out the sunlight! So we now use the country itself, as its own map, and I assure you it does nearly as well.
I think the metaphor illustrates best that we need to drop the pretense of having a consistent, precise representation of the world and settle for abstractions that are fuzzier but more expressive.
Links - 10.08.2007
force.co$ The Perils of Platform As A Service If you have to pay tolls to a gatekeeper in order to reach your customers then you are not participating in a market. That’s the whole essence of a market: the frictionless intermingling of buyers and...Excerpt from discipline and punish at
Rounded Corners - 159 (As seen on TV)
As seen on TV. Steve Vinoski on the similarities between ESB and the bargain multi-tools they sell you on late-night TV: Frankly, if I were an enterprise architect today, and I were genuinely concerned about development costs, agility, and...Excerpt from Labnotes at
bex: auditors should not be allowed to ask questions like: are you sure that transaction #123456, the withdrawal of $100 by John Doe from his account at Washington, took after before transaction #123457, the transfer from her employer to Elena Martin, in Lima? Couldn’t it have happened 100ms before? Those questions are the sort of questions that I claim are irrelevant in most setups, and the ones that make DBs turn into 1:1 scale maps.
This is for a typical auditable app. The problem is that we are so used to transactions that we can’t avoid thinking in terms of them. In a VoIP company I’m consulting with, they are thinking about writing the CDR record locally, and push the changes to the central DB as possible. This would make the system much more robust, at the expense of having the possibility, when the DB is down, of customers calling beyond their balance. Consistency here is given back to the managers as a business decision, where it belongs.
Those are simple examples of places where some consistency can be relinquished. The web is the paramount one.
Posted by Santiago Gala at
auditors will always be a problem, tho... as long as they can demands all kinds of info gathering (who knew what and when), it may be too costly to be fuzzy.
To borrow from Tufte, its better to be approximately right than exactly wrong.
However, when being audited, the exact opposite is the case.
Posted by bex at