6 ms·
Performance and Evaluation of Lisp Systems (1985) [pdf]
- norswap 10y agoThis is by Richard Gabriel, who coined "worse is better" (https://www.dreamsongs.com/RiseOfWorseIsBetter.html https://www.dreamsongs.com/RiseOfWorseIsBetter.html). He's also the author of Patterns of Software (https://news.ycombinator.com/item?id=13799628 https://news.ycombinator.com/item?id=13799628). Incidentally I just finished reading it and found it quite fascinating.
- tmccrmck 10y agoCould you write about what you found fascinating? I read through the first 10 pages and it was mostly about how system architecture applies to Lisps. As someone who has written the back-end of a compiler, I was surprised to learn that roughly 25% of execution time is building stack frames for a caller and a callee.
- norswap 10y agoPoor writing, I meant that Patterns of Software was fascinating. I'd be interested to give a glance to the above though, precisely for insights like that which you highlighted. I'm not surprised at all. Java has the same kind of situation going on, and its performance depends on the ability of the JIT compiler to kick in and perform inlining. When this is not possible (e.g. calls sites where you call a virtual method with many implementations), performance degrades a lot. This is especially bad with interfaces, which have intrinsically higher costs. As for Lisp, it can be even worse if you take things like multimethods into account.
- samth 10y agoThe benchmarks gathered for this book continue to be incredibly significant (though not really for good reasons) 30+ years later. Quite remarkable.
- daliwali 10y agoI think one of the reasons why C became popular in industry and Lisp didn't, is because Lisp is harder to translate into efficient machine code. C was closer to how the hardware works. In the earlier days, specialized hardware was built to run Lisp code: Lisp Machines. When computers were sufficiently fast enough, the specialized hardware was no longer necessary. The software industry could have been vastly different, if Lisp took off instead of C, which I think had more to do with business, culture, & politics than the technology itself.
- pjmlp 10y agoNot really, Lisp Machines were expensive, UNIX was free with a plain symbolic license during the AT&T was prevented from selling it. If UNIX had a price tag similar to all other OSes, the revolution would never have happened.
- baldfat 10y agoI don't think people under 45 remember what the software was like before. Everything including AT&T Unix had a high price point and many "versions" of Unix that caused incompatibilities. LISP systems were extremely expensive and rare. The number is around 7,000 units sold up till 1988. In 1987 Unix became a "Thing" and in 1989 as a teenager I built a fiber optic network with a friend of mine and used System V running Word Perfect and Lotus 1-2-3. It ran into the 21st Century :) > "In a move intended to unify the market in 1987, AT&T announced a pact with Sun Microsystems, the leading proponent of the Berkeley derived strain of UNIX. However, the rest of the industry viewed the development with considerable concern. Believing that their own markets were under threat they clubbed together to develop their own "new" open systems operating system. Their new organization was called the Open Software Foundation (OSF). In response to this, the AT&T/Sun faction formed UNIX International. The ensuing "UNIX wars" divided the system vendors between these two camps clustered around the two dominant UNIX system technologies: AT&T's System V and the OSF system called OSF/1. In the meantime, X/Open Company held the center ground. It continued the process of standardizing the APIs necessary for an open operating system specification." - http://www.unix.org/what_is_unix/history_timeline.html http://www.unix.org/what_is_unix/history_timeline.html