Path: csiph.com!feeder.erje.net!2.eu.feeder.erje.net!fu-berlin.de!uni-berlin.de!individual.net!not-for-mail From: Bob Eager Newsgroups: alt.folklore.computers Subject: Re: PL/I advertising Date: 4 Jun 2016 07:12:41 GMT Lines: 53 Message-ID: References: <873a27e2-6baf-4772-9dc8-45a8f1bc466e@googlegroups.com> <5750739e.11401875@nntp2.rawbw.com> <873a27e2-6baf-4772-9dc8-45a8f1bc466e@googlegroups.com> <5750739e.11401875@nntp2.rawbw.com> <873a27e2-6baf-4772-9dc8-45a8f1bc466e@googlegroups.com> <5750739e.11401875@nntp2.rawbw.com> <2ab91e3f-6f75-45fb-aca3-4c91d7f8f883@googlegroups.com> <20160603171616.b525b4786f06a44e92e3efbb@eircom.net> <20160604061638.e40b31c3d7118fe7ad434b27@eircom.net> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit X-Trace: individual.net 5sMGa4NUS+N8wgpMqvoylQI0JLVAGV0lWfGlSihxzBvgBzNByB Cancel-Lock: sha1:tcUKA+D9P8UEwl7XLPC/NdubFlg= User-Agent: Pan/0.139 (Sexual Chocolate; GIT bf56508 git://git.gnome.org/pan2) Xref: csiph.com alt.folklore.computers:164460 On Sat, 04 Jun 2016 06:16:38 +0100, Ahem A Rivet's Shot wrote: > On 3 Jun 2016 17:38:26 GMT Bob Eager wrote: > >> On Fri, 03 Jun 2016 17:16:16 +0100, Ahem A Rivet's Shot wrote: >> >> > On Fri, 3 Jun 2016 07:37:25 -0700 (PDT) >> > Quadibloc wrote: >> > >> >> On Friday, June 3, 2016 at 7:17:58 AM UTC-6, jmfbahciv wrote: >> >> >> >> > Note that all of you keep assuming that a compiler exists before >> >> > the never-produced soft/hardware of a system is created. That's >> >> > not how it worked in the aulden days. >> >> >> >> It's certainly true that something has to begin the process of >> >> bootstrapping... and that something will have to be in machine >> >> language. >> >> >> >> However, it would merely need to be as little as a loader program, >> >> as one could always use a compiler on another system to write a >> >> cross-compiler. People did that sometimes even in the olden days. >> > >> > Yep, another approach that was used (originated I think) with >> > BCPL was to ship the compiler as source and compiled for a very >> > simple virtual machine (although it was called an intermediate code >> > not a virtual machine in those days) designed to be easy to write an >> > interpreter or translator for. Compiling the compiler made a great >> > test of your work because it should produce output identical to that >> > in the kit. >> > >> > You could then go on to write a better (optimising perhaps) code >> > generator in BCPL having bootstrapped yourself to a (fairly) HLL >> > easily. >> >> I did my M.Sc. dissertation on that. >> >> With an interpreter written in Standard FORTRAN. > > I did it for fun with a translator written in 6809 assembler. Oh, I've done that too. But I was aiming for portability, of course. The interpretive code was closely packed into fixed format 'records' that would work on any machine that had a reasonable FORTRAN implementation. -- Using UNIX since v6 (1975)... Use the BIG mirror service in the UK: http://www.mirrorservice.org