[pyicu-dev] towards release 1.0

[Andi Vajda]

On Wed, 31 Mar 2010, Andi Vajda wrote:

>
> On Wed, 31 Mar 2010, Christoph Burgmer wrote:
>
>> Am Mittwoch, 31. März 2010 schrieben Sie:
>>> On Mar 31, 2010, at 9:57, Christoph Burgmer <cburgmer at ira.uka.de> wrote:
>>> > Hi
>>> >
>>> > Python has been converting all upper case library names to lower
>>> > case for
>>> > Python3. From what I see PyICU doesn't support Python3 yet, so this
>>> > might be a
>>> > non-issue for now, still I'd like to raise this for the 1.0 version
>>> > here.
>>> >
>>> > There's a PEP for the 2to3 changes which I don't have at hand right
>>> > now.
>>> 
>>> I've been thinking of renaming the 'PyICU' module name to 'icu' just
>>> like I did with PyLucene when switching from gcj to jcc, taking
>>> advantage of the backwards compatibility break this introduced.
>>> 
>>> If backwards compat is not an issue - or if this is the least bad time
>>> to do this - before 1.0 - I'd actually like to do so.
>>> 
>>> Andi..
>> 
>> What about:
>> 
>> PyICU.py:
>> import warnings
>> warnings.warn("Module 'PyICU' is deprecated, use 'import icu' instead'",
>>    category=DeprecationWarning, stacklevel=2)
>> from icu import *
>> 
>> Of course followed by updating imports in _PyICU.cpp.
>
> This won't cover the necessary egg name change, right ?
> I think I'm just going to swallow the backwards compatibility break and do 
> the rename...

But isn't that going to cause confusion on PyPI ? A new project would 
appear, called 'icu', instead of 'PyICU'. Or is there a way to name the egg 
with 'icu', yet continue to name the project 'PyICU' on cheeseshop ?

Andi..