[Alpine-info] Alpine is converting ? to ??

[Benjamin R. Haskell]

On Tue, 10 Mar 2009, Mark Crispin wrote:


> On Tue, 10 Mar 2009, M.RossARR wrote:

> > > The main usage today of ISO-8859-1 is by Americans who don't know 

> > > any better.

> > That would be me! Until now! Thanks for the enlightenment!

> 

> I do my best to try to get the word out.  Death to ISO-8859-1.  Death to 

> all charsets other than UTF-8.

> 

> One charset to rule them all

> One charset to find them;

> One charset to bring them all

> And in interoperability bind them!


The rest of Mark's explanation (which I've omitted from this response) is 
really good.  UTF-8 = wonderful, happy, joy, hooray.  If you don't have 
very specific, very technical reasons to do otherwise (encoding ancient 
Chinese texts, or dealing with legacy systems that absolutely, positively 
can't be converted), you should use UTF-8 everywhere and anywhere it's an 
option.  It's far and away the best encoding for general purpose computing 
today.

If you want some Unicode/UTF-8 advocacy info, especially geared toward 
software developers, everyone should read:
http://www.joelonsoftware.com/articles/Unicode.html
(repeats many of Mark's points, but is a good read)


But, I felt like playing devil's advocate and pointing out the fact that 
it's not perfect.  Two good, relatively-short lists of problems are:

The secret life of Unicode ["ancient" in WWW timescale]:
http://www.ibm.com/developerworks/java/library/u-secret.html

and Wikipedia's list of "Issues" on the Unicode page:
http://en.wikipedia.org/wiki/Unicode#Issues
(many repeats between the two)

Personally, the only thing that ever frustrated me in Unicode was 
discovering that there's no way to specify the "simplified" (vs. 
"traditional") form of a Chinese character.  There are longstanding 
debates about this, and in the end I feel Unicode's Han Unification is a 
good compromise.  [It's similar enough to wanting to be able to specify 
the "italicized" (vs. "regular") form of a letter (i.e. something that 
should be handled "above" the Unicode layer)... But, I digress.]

I'm not so much "UTF-8 forever!" as "UTF-8 for now and the foreseeable 
future!"

Best,
Ben