Re: Language extensions

[Marek Janukowicz]

On Sunday 16 of January 2011, Keith Hopper wrote:
> On 16/01/11 12:17, Marek Janukowicz wrote:
> > Btw. there is another question - what is the proper way (if any) to go
> > with language extensions? One thing that I *really* miss since I
> > starting writing stuff bigger than a few classes is some kind of
> > namespaces. Because right now I'm ending up with class names like
> > MUSH_NET_HTTP_REQUEST_GET (MUSH being the general name of my library). I
> > could try taking a shot at implementing it on compiler level (doesn't
> > sound like something terribly complicated)
> 
>     ......eerrrrrr!! Whether modern linkers have improved the situation
> or not I am not sure, but in earlier days the problem was the name
> mangling involved to get identities for linkable items which were short
> enough - and still sufficiently recognizable - to work with the various
> linkers around (some of which just truncated too-long identifiers). This
> is not just a compiler issue - rather one of the compilation environment.

I get it, but whether I use long class name prefixes or namespaces the problem 
remains the same (IMHO). Anyway, thank you for bringing this up, I'm not 
acquainted with compilation environments too much, so I did not think about 
the issue in question at all.

-- 
Marek Janukowicz