Okay, a few more words about
http://www.collation-charts.org
MySQL's team lead for character sets is Mr Alexander Barkov.
On his own time he set up collation-charts.org
to show, graphically, how collations work in DBMSs
(Firebird, SQL Server, MySQL, Oracle 10g, and Sybase ASE)
and operating systems
(i5, HPUX, Solaris, Linux, Windows, FreeBSD).
An example of a page is
http://www.collation-charts.org/mysql60/mysql604.latin1_swedish_ci.html
which shows how the MySQL 6.0 latin1_swedish_ci collation works.
Have a look at the line for the letter 'Y' on this page.
If you hold the keyboard over the box for the letter 'Y',
you will see "[59] LATIN CAPITAL LETTER Y". Here '59' is
the weight, and the rest is the character's Unicode name.
The other characters on the same line are LATIN SMALL LETTER Y,
LATIN CAPITAL LETTER U WITH DIAERESIS, LATIN CAPITAL LETTER Y
WITH ACUTE, LATIN SMALL LETTER U WITH DIAERESIS, and
LATIN SMALL LETTER Y WITH ACUTE. These all have the same weight
as 'Y', but a different colour (this particular colour shows
that the difference from 'Y' is not a level-1 difference).
The equivalence of Y and U WITH DIAERESIS is of course
not what you'd see in the German collations, for example
http://www.collation-charts.org/mysql60/mysql604.latin1_german1_ci.html
Why is the latin1 default latin1_swedish_ci not latin1_german1_ci?
Ask MySQL's three founders, who are all Swedish speakers.