Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!shadooby!samsung!brutus.cs.uiuc.edu!apple!hercules!gilham
From: gilham@csl.sri.com (Fred Gilham)
Newsgroups: comp.sys.amiga.tech
Subject: Re: ^G in WorkBench 1.4
Message-ID: <GILHAM.89Nov10165501@cassius.csl.sri.com>
Date: 11 Nov 89 00:55:01 GMT
References: <3742@altos86.Altos.COM> <1159@maestro.htsa.aha.nl> <3260@jhunix.HCF.JHU.EDU>,
	<3770@nigel.udel.EDU> <0ZKoagy00UkaFnd_R7@andrew.cmu.edu>
Sender: usenet@csl.sri.com
Organization: Computer Science Lab, SRI International, Menlo Park, CA.
Lines: 22
In-reply-to: bader+@andrew.cmu.edu's message of 10 Nov 89 22:20:28 GMT

>> No, no, no!  Keep S: for scripts.  Put it in DEVS:BeepSound.  -- Darren

>Why is a sound-sample any more appropiate in "devs:" (devices) than in
>"s:"???????
>-Miles

Since the subject came up, let me say that I wish no one would
hard-code their programs to use any system-defined directories to find
auxilliary files.  When someone does this, I have to put the
auxilliary file there whether I like it or not.  So, for example,
bison skeletons go in s:.  Well, I want to put s: on my rad: disk, but
I don't want to have the bison skeleton taking up space all the time.
Why not have bison look at (say) SKEL:, then I could, if I wanted, say

assign SKEL: s:

or whatever, thus allowing ME to put the file where I want it.  This
seems to make more sense.  I really appeal to program writers, don't
use system directories for auxilliary files unless for things where
they have to go there.

-Fred Gilham gilham@csl.sri.com