Path: utzoo!mnetor!uunet!lll-winken!lll-lcc!ames!xanth!kent
From: kent@xanth.cs.odu.edu (Kent Paul Dolan)
Newsgroups: comp.sys.amiga.tech
Subject: Re: IFF ILBM File to Gadgets etc...
Message-ID: <4953@xanth.cs.odu.edu>
Date: 20 Apr 88 09:35:24 GMT
References: <880418104522.0eg@Ins.MRC.AdhocNet.CA>
Reply-To: kent@xanth.UUCP (Kent Paul Dolan)
Organization: Old Dominion University, Norfolk Va.
Lines: 29

In article <880418104522.0eg@Ins.MRC.AdhocNet.CA> svermeulen%Ins.MRC.AdhocNet.CA@UNCAEDU.BITNET writes:
[...]
>This is very true, also there are no limits as to the UPPER LIMIT for
>picture sizes (appart from about 65K pixels wide and high :-).  A
>convenient size for laser printing is 2400x3204...

>               Stephen Vermeulen
>               Author: Express Paint


How I wish that smiley face made me smile.  There have been display
devices for at least ten years with more resolution than that!  For
one example, the Scitex laser onto film textile design and
cartographic workstation that long ago used a plot resolution along
the 70" long drum circumference of at least 1000 points per inch.
Another example of a reasonable looking design limit being overtaken
by technology, in this case, before the fact.

This is one of many reasons I lean toward 256 bit longs in C.  ;-)
(Perhaps that smiley face will outlive me, but I wouldn't bet my life
on it.)

The moral lesson is: when designing a system limit, by choice pick a
link list or variable length integer as appropriate; if you think you
_must_ make a hard limit, try to wiggle off the hard limit hook as
hard as you can; if unsuccessful, try the _square_ of the largest
example you've ever seen, and hope you aren't overtaken by progress
before you sell your first unit.

Kent, the man from xanth.