-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 22-May-07, at 2:03 PM, Bradley Lucier wrote:
Marc:
One finds
[brad:~/programs/gambc-4.0b22/gsc] lucier% grep make-table *.scm _host.scm:(define (make-table . args) _t-c-1.scm: (set! targ-glo-vars (make-table 'test: eq?)) _t-c-1.scm: (set! targ-sym-objs (make-table 'test: eq?)) _t-c-1.scm: (set! targ-key-objs (make-table 'test: eq?)) _t-c-1.scm: (set! targ-num-objs (make-table 'test: eqv?)) _t-c-1.scm: (set! targ-prc-objs-seen (make-table 'test: eq?)) _utils.scm: (vector (make-table 'test: test)
and extended-bindings is not declared, so when a user redefines make-table it screws up the compiler.
I see this is hardly unique, so I'll ask you: is this kind of stuff a bug?
Brad
Yes, I think this is a bug that needs to be addressed. Marc -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.3 (Darwin) iD8DBQFGU6R7//V9Zc2T/v4RAgpsAKC8NuYXs8R5qx7vWLCEeTL/nMw9sgCeKw2t 6XZ+FmY71K6P54cWZkHjQvw= =M270 -----END PGP SIGNATURE-----