Path: csiph.com!eternal-september.org!feeder.eternal-september.org!reader01.eternal-september.org!.POSTED!not-for-mail From: Ben Bacarisse Newsgroups: comp.programming Subject: Re: Is Pascal utterly useless? Date: Sun, 08 Sep 2019 15:12:59 +0100 Organization: A noiseless patient Spider Lines: 21 Message-ID: <87ef0qx4es.fsf@bsb.me.uk> References: <87pnkg149f.fsf@bsb.me.uk> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: 8bit Injection-Info: reader02.eternal-september.org; posting-host="ababc0e3e4750e98f5e1869c5aa0a110"; logging-data="12823"; mail-complaints-to="abuse@eternal-september.org"; posting-account="U2FsdGVkX1/6u/qqnEbt6p77MxSbYahSGrUSBa5e9W0=" Cancel-Lock: sha1:KNV9YrH6HPZPx9DjejGRGrgVA2E= sha1:uxfmqQMUB96cbEfuyc/eJo9GhfE= X-BSB-Auth: 1.3a8530159e2aa4fa22a0.20190908151259BST.87ef0qx4es.fsf@bsb.me.uk Xref: csiph.com comp.programming:10184 Peter Percival writes: > Ben Bacarisse wrote: >> There's an excellent book (old now) that compares languages by >> examining the extent to which they could implement their own IO >> facilities. It's a very revealing point of view. Sadly I can't >> remember anything else about it. > > If you should come to remember its title & author let me know, I'd > like to read it. Sorry, can't find it. I suspect I may have conflated two things, a paper or technical report by Ron Morrison and a book by D M Harland "Polymorphic programming Languages – Design And Implementation". That book (if I recall correctly) advocates for designs where nothing is privileged, but the "how well can a language implement its own IO" is something else. -- Ben.