[cfe-dev] Should we build semantically invalid nodes?

[Ted Kremenek]

On Oct 26, 2008, at 12:16 PM, steve naroff wrote:

>> I'm still struggling to figure out what problem you're trying to
>> solve.
>>
>
> I think any breakthrough in modularizing Sema (or not) will come from
> developing new clients.

I agree with Steve and Chris here.  While there is value in looking  
forward, the demands of real clients will clarify our (future) needs  
from Sema, further modularity in Clang, new Action modules, and so  
forth.  At this point much of this discussion, although extremely  
interesting, seems dominated by discussion about hypothetical issues.