On 06/23/2015 02:23 AM, Joe Touch wrote:
On 6/22/2015 4:38 PM, John C Klensin wrote:
Given the way the whole procedure works,
repeated use of one country's statistic(s) to get some digits
seems no more political than a choice of lotteries or horse
races.
If that is true, then why not make it so? Or use other numbers that are
similarly random and independently verifiable? (e.g.,, the low-order
numbers of the individual stocks on the Dow Jones or FTSE).
I made a simple request that is trivial to address for a convention that
is NOT required by the cited RFCs.
The push back I've seen so far indicates to me that there are clear
political motivations for using the US Debt numbers instead. So be it.
Speaking as the person who actually picked these lotteries and numbers:
I picked them because it was the minimum change possible to what was
done last year (the only effort required was to update an URL that had
changed over the course of the year).
The next nomcom chair will have additional input since this debate has
occured.
Speaking as another geek:
I see the value of having multiple types of sources as being that no
single organization is likely to be able to manipulate *all* of these
without being detected - at least for an input that is commensurate with
the value of controlling the pick of nominees for next year's Nomcom.
The important properties of input are:
- They should be hard for anyone to fix before the fact
- They should be easy for anyone to check after the fact
I thought of adding a Norwegian lottery to the mix (THAT would have been
strictly as a political statement), but decided not to bother; finding
the correct URL and verifying that it had remained
stable for a reasonably long time was more work than I found worth doing.
I'll be glad to pick equally offensive values for entropy in the next
specs I author.
Feel free, but don't expect me to care.
Joe