RE: proposal: change some confusing method names (getCharStart() / getCharLimit())

[Yves A. <yv...@re...>]

> Well, the names are supposed to be parallel to 
> UTF_SET_CHAR_START() and UTF_SET_CHAR_LIMIT(), as much as 
> possible (it's "get" here because the UnicodeString functions 
> do not _change_ the input index).
> 
> I would personally like to leave it as it is.
> 
> Other opinions?

Mine again :-)

I think it's worse to have an inconsistent C++ API because of a couple of C
macros, than to have different names between the two languages.

In C macros, you consistently have CHAR when manipulating a character.

In the C++ code, you consitently use Char for a code *unit*, and Char32 for
a code point, except for these two methods. I do not think that the typical
C++ programmer is going to be aware of the C CHAR macros, and I have real
people here getting confused by all of this. Why make the learning curve
steeper than needed? API consistency within a language should come before
API consistency across languages (which would be great, but isn't there in
ICU).

YA