On Nov 16, 2015 5:37 PM, "Lisi Reisz" <lisi.re...@gmail.com> wrote: > > On Monday 16 November 2015 19:33:51 David Wright wrote: > > On Mon 16 Nov 2015 at 06:54:40 (+0100), Martin Str|mberg wrote: > > > In article <qvjtp-2c...@gated-at.bofh.it> David Wright > <deb...@lionunicorn.co.uk> wrote: > > > > As for script-file extensions in DOS, there was really only .BAT > > > > wasn't there?, so the idea of distinguishing .bash, .csh, .py, .pl, > > > > .sh, .zsh etc as being inherited from DOS is difficult for me to > > > > understand. > > > > > > Perhaps it's because (MS)DOS begat WINDOWS that only knew how to run > > > something based on the extension? > > > > > > And that is why we shudder on the sight of a (unnecessary?) extension? > > > > I wouldn't know. My experience of windows is far less than DOS, and my > > use of DOS was pretty much restricted to an AUTOEXEC.BAT that started > > an emulation system which was my area of expertise. Interesting choice > > of language, though; shudder. > > > > I'm the person questioning the relevance of DOS to putting ".sh" at > > the end of an on-PATH executable script's filename, when DOS was > > brought up in https://lists.debian.org/debian-user/2015/11/msg00453.html > > > > I take it there's a whole generation of folk who gained their > > experience of filename endings (a less loaded word than extension) > > through DOS/windows, perhaps entirely so. With it, they picked up a > > load of negative associations, causing shuddering here and unhappiness > > in another part of this thread. > > > > I'm sorry for you. I didn't touch DOS until 1992-06-01 (to be precise) > > about twentyone years into my computing career. To say I *used* it > > would be an overstatement: I ran one package on it. > > > > Putting meaningful endings onto filenames (excepting, I hasten to > > add lest people jump down my throat, executable scripts) had been a > > way of life for years. Their necessity was variable from system to > > system; sometimes they were just a convention. Look at man gcc. > > It has meaningful endings. They've been there since at least > > 15 March 1972 when, allegedly, the number of Unix installations had > > grown to 10. (At that time, gcc was obviously called cc; Stallman > > hadn't yet graduated.) > > > > As for unix scripts, well, yes, there's no *need* for any endings, > > but that doesn't preclude their use. If that makes you unhappy or > > into a shudderer, please get over it. > > I take it those who are so against file endings are equally upset by > sources.list and menu.lst? > > Though it is very annoying when they are *needed*. Xsane usually puts them > in, and I used not to bother to check. A few months ago I sent my lawyer a > scan of a document he needed. An hour or two later, back came an > email: "I'm so sorry, we have no software that can open that file. The IT > department has been trying for an hour". Puzzled, because I thought I had > sent a .pdf, and had checked that it opened fine in Evince, I looked at the > file - groaned - and renamed scan-foo to scan-foo.pdf. When resent it opened > fine. >
Magic shouldn't be used when a file type is known and should be communicated (via its extension). If you create a pdf, it is bad to not have the pdf extension - you've lost data. Also, I'm fine with installed executables not including extensions. However, in a repo with different types of executable code, I want to know at a glance what something is (there's also that vim filetype uses extension and I want stuff to work).