ftp.cs.arizona.edu/ftp.cs.arizona.edu.tar/ftp.cs.arizona.edu/icon/newsgrp/group93a.txt/000052_icon-group-sender _Fri Jan 29 19:09:07 1993.msg

Internet Message Format | 1993-04-21 | 1KB

Received: by cheltenham.cs.arizona.edu; Sat, 30 Jan 1993 05:51:05 MST
Date: Fri, 29 Jan 1993 19:09:07 MST
From: "Clint Jeffery" <cjeffery>
Message-Id: <[e-mail]>
To: [e-mail]
Cc: [e-mail]
In-Reply-To: [e-mail]'s message of Fri, 29 Jan 1993 11:06:24 +0300 (EET) <[e-mail]>
Subject: Removing various features
Status: R
Errors-To: [e-mail]

    > Arne Larsson writes:
    ...By the way, what is the overhead of having a few unused
    builtins in a programming language... [I'll say, Icon, for example]

For users, the overhead is not bad, except on systems with tiny memories,
such as MS-DOS.

For language implementors, some little-used built-ins are no trouble
and others impose increased difficulty of implementation and increased
cost of changing the implementation.  For example, when we developed
the Icon compiler, we had to make Large changes to all of these
built-ins.  For a math function like sin() it was no sweat.  But for
entab()/detab() it was a headache. The "mathematical minimalist" notion
is not so bad whenever code plays tricks with the garbage collector.
I hope entab() and detab() will get used more, considering the effort
that has gone into them and all the free advertising they have been
getting lately... :-)