It is not an isolated issue unless you are the core maintainer who has absolute authority and can give your word that you'll match Sun's interface stability with ICU going forward.
And for the reasons I mentioned in the previous email, I assert ICU 4.0.1 should be Volatile. Ienup Nicolas Williams wrote at 02/24/09 12:00: > On Tue, Feb 24, 2009 at 11:47:30AM -0800, Ienup Sung wrote: >>> Can you point to specific instances where they've done this? >> Yes, between ICU 3.8 and 3.8.1, date formatting and parsing has changed in >> ways that may require re-coding of user programs. Please see Known Issues, >> Bugs and Patches section of the following: >> >> http://icu-project.org/download/3.8.html#ICU4C > > Thanks. That seems like a very isolated issue. I don't see any other > such issues noted in the releases of ICU4C versions 3.2.1, 3.4.1, 3.6 or > 4.0.1. The actual readmes don't mention any other backwards > incompatible changes either. > > As far as I can tell there's no reason that we couldn't ship multiple > versions of ICU if we had to as the result of an unfortunate backwards- > incompatible change. Therefore I don't see why we couldn't ship ICU > 4.0.1 as Uncommitted for C and Java (and Volatile for C++). > > Nico
