>I'd vote for switching to ISO Latin-1 because it is one of the
>best-documented encodings (it's almost impossible to find a Mac encoding
>chart on the web). Besides, codes from 128 to 159 are "free" so we can put
>special characters there.

I'm not married to the status quo, but this reasoning seems to me 
problematic.  It was trivial for me to find the MacRoman chart on 
web, which is available from many sources, not the least of which is, 
ahem, the Apple web site (look under Technical publications: Text):

http://developer.apple.com/techpubs/mac/Text/graphics/EX_L-04.jpg

Nor do I find the use of a "standard" particularly compelling, since 
we will likely be moving a few glyphs around for programming purposes 
(caret to return glyph, underbar to assignment glyph).  To the 
contrary, I found it disconcerting to discover both the Tex and ISO 
Latin-1 fonts "break" existing displays that depended upon the Mac 
encoding.  If we do change encodings, some machine-independent 
facility will need to be implemented across platforms to assure that 
keycodes/screen glyphs conform to expectations uniformly, or we need 
to be willing to adopt a different approach.

In short, perhaps we should worry about this another day, and simply 
modify these fonts to conform to the present "modified" MacRoman 
format.  They can later be changed once a "purer" encoding is found. 
The repertoires are not identical (MacOS supports several 
diacriticals not in ISO Latin-1).  On the other hand, do we really 
want to continue supporting what may become a dionsaur?  (Does anyone 
know what encodings are primarily used in OS X?)

At any rate, some decisions should be made as to a Squeak standard 
repertoire and encoding, so that new fonts can be used seamlessly and 
without compromising existing code (or identifying where changes 
might need to be made).  Where shall we put the "replaced" glyphs? 
Shouldn't we place a Euro symbol in all of our fonts?