From ... Path: archiver1.google.com!news1.google.com!newsfeed.stanford.edu!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!uio.no!nntp.uio.no!ifi.uio.no!not-for-mail From: Erik Naggum Newsgroups: comp.lang.lisp Subject: Re: Common Lisp wish list item Date: 20 Aug 2002 05:55:10 +0000 Organization: Naggum Software, Oslo, Norway Lines: 27 Message-ID: <3238811710520717@naggum.no> References: <3238779012022126@naggum.no> <87u1lqjzb2.fsf@cinifa.internal> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Trace: maud.ifi.uio.no 1029822910 7084 129.240.64.16 (20 Aug 2002 05:55:10 GMT) X-Complaints-To: abuse@ifi.uio.no NNTP-Posting-Date: 20 Aug 2002 05:55:10 GMT Mail-Copies-To: never User-Agent: Gnus/5.09 (Gnus v5.9.0) Emacs/21.2 Xref: archiver1.google.com comp.lang.lisp:38367 * Bulent Murtezaoglu | Would you not consider (*-byte 32) declarations along with block compilation | as a possible solution? For the particular case where I work specifically with modulo-32-bit integers, that would suffice, but where I need real integers, I do not want C semantics, but I would really appreciate being able to use the full machine width without going to bignums with the last 2 bits. | Am I missing the point? What I really want is better integration with the hardware and other tools on the same hardware. On the other hand, if we had boxed integers in the normal case, I could imagine using interned integers and single-word boxes that could be upgraded very efficiently to a bignum when whole machine words were used in the computations. The assembler programmer in me still protests against the shifting and reduced value space. So I also wonder what people have done, not whether they are satisfied with the status quo, which I guess most people are, at least as far as doing anything about it. -- Erik Naggum, Oslo, Norway Act from reason, and failure makes you rethink and study harder. Act from faith, and failure makes you blame someone and push harder.