[MPlayer-cvslog] CVS: main AUTHORS,1.176,1.177

[The Wanderer]

Rich Felker wrote:

> On Sun, Apr 09, 2006 at 11:11:32AM +0300, Ivan Kalvachev wrote:
> 
>> We may be one big (flaming;) family but I doubt there are enough
> 
> ROTFL! :))))
> 
>> family relationships in the people of the list so family_name_first
>> to bring something else than confusion.
>> 
>> I really hate to see my name backwards.
> 
> Me too. I vote to correct it. Modern usage is sorting in ordinary
> order by how the name is written (look what cell phones do for
> instance..). I hate legacy collation rules.

I vote to keep it, surname first. (Not all cell phones sort the way you
note, although I don't have a counterexample handy to cite.)

As I understand matters, the primary rationale behind this 'rule' is
that it is far more likely that two people will have the same personal
name/first name than it is that they will have the same surname, and so
sorting by surname produces a more immediately heterogenous list - i.e.,
less duplication at the beginning of the sort field. If you're looking
for someone named (to pick an example at random) "John Walker", looking
in a personal-name-sorted list is going to find you dozens of "John"s to
wade through before you find the right one; looking in a surname-sorted
list is going to find you proportionally very few "Walker"s, and finding
the right one will be easier.

"Family relationships" as such have little or nothing to do with it. It
is largely, perhaps even exclusively, a matter of duplication of
*personal* name - not of surname.

Admittedly the problem of duplication of personal name is not likely to
be significant in a population as small as the one on that list, but the
principle behind the rule is the same in any case.

-- 
       The Wanderer

Warning: Simply because I argue an issue does not mean I agree with any
side of it.

Secrecy is the beginning of tyranny.