[sc34wg3] Re: Unicode normalization

[Lars Marius Garshol]

* Patrick Durusau
| 
| Unicode normalization is not required.

Yep. Note that the subject of this statement is omitted, but is
implicitly the TM implementation. (It can't be anything else.)
 
| But use of logic to determine "equivalent strings" is encouraged.

Correct. TM implementations are encouraged to do this.
 
| But reliance on normalization is discouraged.

Correct. Note that here the subject is, again implicitly, different.
This time the subject is TM authors.
 
| So, what is being encouraged is use of logic to determine equivalent
| strings that does not rely on normalization?

Not really. :)

If you as a TM author am passing your TM to someone else you are
discouraged from relying on the someone else to have a TM
implementation that performs normalization.

| OK, I'll buy that but the mixing of encourage and discourage seems
| confusing.

Maybe. I'd be curious to hear what others think of this.
 
| Suggest:
| 
| "Application of logic to determine the equivalency of strings is
| encouraged. However, since Unicode normalization of strings is not
| required and therefore cannot be guaranteed, reliance on
| normalization in the application of such logic is strongly
| discouraged."

The implementation is what applies the logic, so this doesn't work.

-- 
Lars Marius Garshol, Ontopian         <URL: http://www.ontopia.net >
GSM: +47 98 21 55 50                  <URL: http://www.garshol.priv.no >