On Sat, Sep 3, 2011 at 5:37 PM, Matthew Flatt <[email protected]> wrote: > At Sat, 3 Sep 2011 17:25:58 -0400, Sam Tobin-Hochstadt wrote: >> On Thu, Sep 1, 2011 at 5:20 PM, <[email protected]> wrote: >> > >> > src/racket/src/place.c >> > ~~~~~~~~~~~~~~~~~~~~~~ >> > + REGISTER_SO(quote_symbol); >> > + quote_symbol = scheme_intern_symbol("quote"); >> >> This is the third definition of 'quote_symbol' in the source. Are >> repeated definitions a problem? > > It doesn't seem like a problem to me. What did you have in mind?
I was just generically worried about duplicated code. Looking more deeply, it seems the only potential problem is one more (very small) GC root, and root traversal surely isn't the current GC bottleneck. >> Can these be shared (other than via >> the intern table)? > > A `scheme_quote_symbol' could be defined somewhere and exported to the > other places. The tradeoff is writing `(define QUOTE 'quote)' multiple > times versus making a global change. Make sense. -- sam th [email protected] _________________________________________________ For list-related administrative tasks: http://lists.racket-lang.org/listinfo/dev
