Comment 23 for bug 206884

The whole thing seems very clear to me.
But, of course, I didn't write the program.
Nor am I writing the documentation.
Or am I?
Is this the patch?

View|Character Encoding mentions the encoding that has been used to
display the page currently displayed. If you change that selection, that
page is redisplayed using the new encoding you select. Of course, the
selection you make applies to the current page only. The list of
encodings is made of user selected entries plus actually used encodings.

View|Character Encoding|More Encodings is used to reach encodings that
are not found in the list (and that, by being used, are temporarily
added to the list).

View|Character Encoding|Customize List... does that for the user
selected entries.

View|Character Encoding|Auto-Detect needs more explanation (and author
validation).
Normally, a HTML page indicates the encoding it uses.
By default (no indication) the encoding is ISO 8859-1.
But Bug #228988 <https://bugs.launchpad.net/bugs/228988>, reports that
Firefox sometimes fails to use the default.
So, why is an Auto-Detect needed if it's clear what encoding to use?
Well, because Web page authors sometimes forget to indicate the encoding.
In that case, Firefox may help you auto-detecting the pages in your
language.
Auto-Detect is a global option that applies in advance to any page that
does not indicate an encoding : it will try to best guess if it should
display such pages with your language's encoding instead of the normal
default (ISO 8859-1).
But other users will continue to display those defective pages badly and
Auto-Detect may get yourself into problems by not displaying ISO 8859-1
correctly.
So, using this option must be done with caution and does not dispense of
asking the Web author to put his pages right. Auto-Detect should be
returned to off after a problem is solved. Otherwise, additional
problems will not be detected.
Auto-Detect Universal : best guess of the best guess is that this means
auto-detect any UTF code.
BTW, all the problems will be over when UTF-8 will be used exclusively
(AP, 1992).