Xref: utzoo comp.sources.wanted:8446 comp.text:4687 Newsgroups: comp.sources.wanted,comp.text Path: utzoo!henry From: henry@utzoo.uucp (Henry Spencer) Subject: Re: DITROFF question(s). A repost of my earlier mail. Message-ID: <1989Aug14.215310.2248@utzoo.uucp> Organization: U of Toronto Zoology References: <7040@cloud9.Stratus.COM> Date: Mon, 14 Aug 89 21:53:10 GMT In article <7040@cloud9.Stratus.COM> fmbutt@cloud9.Stratus.COM (Farooq Butt) writes: >it does not seem to have a "psc" (postscript) device driver system >( i.e. When I try to say: troff -Tpsc foo | psdit | lpr, troff >complains that I don't have "devpsc" in/usr/lib/font). > > 2. If ditroff is TRULY device independent why would I need a > a postscript device driver, as opposed to a generic DIT -> PS > converter such as psdit supplied by Sun with Transcript ? What it is complaining about is not the lack of a device driver, but the lack of a description file to tell it what fonts, sizes, characters, etc. are available for your device. There is a common misconception that ditroff output is device-independent; NOT SO. The output *format* is device-independent but the *content* is built with the capabilities and limitations of the particular device very much in mind. (For example, ditroff cannot possibly justify text without some idea of how wide each character is on the output device.) >I hope I have not asked many stupid questions but then again I am >a troff-illeterate TeX hacker...... In case you're wondering why TeX doesn't run into this problem... TeX deals with it by trying to pretend that all output devices are the same. This is close enough to being true that it works for a fairly wide range of the better devices, but it tends to fall down for poor ones. -- V7 /bin/mail source: 554 lines.| Henry Spencer at U of Toronto Zoology 1989 X.400 specs: 2200+ pages. | uunet!attcan!utzoo!henry henry@zoo.toronto.edu