Re: FWD:Re: refactoring
Joel de Guzman <[email protected]>
| Newsgroups | gmane.comp.documentation.synopsis |
|---|---|
| Message-ID | <[email protected]> |
Stefan Seefeld wrote: > David Abrahams wrote: > >> David Abrahams <[email protected]> writes: >> >> >>> Stefan Seefeld <[email protected]> writes: > > >>>> I don't think the backend really comes into play. >>> >>> >>> Yeah, I think Joel is using the term "backend" to mean "all the >>> semantic analysis," whereas most people mean "platform-specific code >>> generation and optimizations" when they say "backend." >> >> >> >> Err, sorry, I was wrong. Yes, architecture does come into play, or it >> can. For example, >> sizeof(int) == sizeof(long) >> >> is a compile-time constant that can affect which of two class template >> specializations is chosen, completely altering the types in play, and >> also such things as (member function) overload sets. > > > Fair enough. I just wanted to make sure we agree on what we are *not* > talking about, i.e. what most people refer to as a 'compiler backend', > i.e. code optimization, register allocation, etc. Agreed. I think I should set it straight: syntax analysis vs. semantic analysis. Pardon the confusion. Cheers, -- Joel de Guzman http://www.boost-consulting.com http://spirit.sf.net