The exception that proves the rule that was the exception that proves another rule (aka On the variability of the Invariant)

by Michael S. Kaplan, published on 2007/05/12 03:31 -04:00, original URI:

It was over two years ago that Shawn talked about how Culture data shouldn't be considered stable (except for Invariant).

In that post he was explaining how cultural and regional data was all subject to change at pretty much any time, with the one exception (which only proves the rule) being InvariantCulture.

Anyway, I was at an interesting Windows Developer talk yesterday, and before the talk I was having a conversation with somebody that reminded me of the exception to this rule. So I thought I'd blog about it.

This is nothing new, mind you. I first pointed it out in Where is the locale? "Its Invariant." In where? and followd up the next day with Comparison confusion: INVARIANT vs. ORDINAL.

The one thing that will change between versions/platforms? The collation.

It was in fact the only feature that was originally functional for the locale, and it was only later after bugs were entered that it became a full locale (and subsequently a culture).

Since its sorting behavior is based on the DEFAULT collation table, which can (and has) changed in different versions, collation represents the one part of INVARIANT that can be reasonably be expected to VARY....



This post brought to you by (U+3007, a.k.a. IDEOGRAPHIC NUMBER ZERO, the first character with no weight that was ever reported to me as a bug!)

Blake Handler on 12 May 2007 11:06 AM:

I always thought that the "exception" DISproves a rule. You must be able to reproduce an experiment to prove its validity.  An “exception” would require us to rethink the rule.

OK, it’s just an expression . . . (^_^)

Blake Handler

Microsoft MVP

Michael S. Kaplan on 12 May 2007 6:41 PM:

No, I understood the idiomatic usage correctly. :-)

go to newer or older post, or back to index or month or day