ooc-dev team mailing list archive

[<ndd@xxxxxxxxxx>]

On Wed, 01 Sep 2010 17:53:52 +0200, <ndd@xxxxxxxxxx> wrote:
> On Wed, 01 Sep 2010 12:01:40 +0200, Vincent Richomme
> <forumer@xxxxxxxxxxxxxxx> wrote:
>> On Thu, 12 Aug 2010 08:35:22 -0500, Amos Wenger <ndd@xxxxxxxxxx> wrote:
>>> Yes changing designated initializers to regular struct initializers isn't
>>> impossible at all. It just frightens me a little, because if you forget a
>>> member, everything goes awry, whereas designated initializers don't have
>>> this problem.
>>>
>>> I'll try to modify rock to use regular initializations for class
>>> structures and let you know.
>>>
>>> ndd
>>>
>>
>> Just a reminder because I am still curious to see if it would be
>> possible to compile with MSVC.
>>
> 
> Yes, as a matter of fact I did attempt to get that working while I was
> in a "Hey let's check if the build still works on Win32" spree, but
> it's in fact a significant change of state of mind when writing class
> structures
> 
> What currently happens now is that rock cycles through the class and
> write
> every function  that is implemented and/or overriden (that also works
> for abstract classes)
> 
> But with regular initializations (ie. C89 ones) and abstract classes we
> need to set some fields
> to null, and in the right order, etc. So long story short,
> ClassDeclWriter needs a serious
> cleanup if we're gonna make it work with C89 initializers. It's not
> impossible, but
> it takes time. While I'm working on other issues right now (see on
> GitHub's rock issues the 'nddrylliog'
> tag to see what I'm working on) I definitely haven't forgotten about
> this. (I even went
> so far as to actually install VS2010, hehe)
> 
> ndd
> P.S: sorry for the double-post.. new mail client fail.
> 
>> Thanks
>>
>>
>> _______________________________________________
>> Mailing list: https://launchpad.net/~ooc-dev
>> Post to     : ooc-dev@xxxxxxxxxxxxxxxxxxx
>> Unsubscribe : https://launchpad.net/~ooc-dev
>> More help   : https://help.launchpad.net/ListHelp