Path: utzoo!dciem!trigraph!john
From: john@trigraph.UUCP (John Chew)
Newsgroups: comp.lang.postscript
Subject: Re: A little PostScript hacking for NeXT screens
Message-ID: <458@trigraph.UUCP>
Date: 21 May 89 19:36:53 GMT
References: <3338@tank.uchicago.edu> <5031@pt.cs.cmu.edu>
Reply-To: poslfit@gpu.utcs.utoronto.ca (John J. Chew)
Organization: Trigraph Inc., Toronto, Canada
Lines: 35

In article <5031@pt.cs.cmu.edu> moore@PULSAR.FAC.CS.CMU.EDU (Dale Moore) writes:
...
>>%%BoundingBox: 0.000 0.000 512.000 480.000
...
>According to the document structuring conventions,
>a bounding box is an orderd set of four "integers".
>
>Now, the above numbers certainly look like floating point
>numbers.  And we all know that floating point numbers aren't
>integers.
...
>But I've a couple of utilities that look at such things as 
>"%%BoundingBox"  and I wonder if I should reject these things for
>not being formatted properly.  If I should accept these
>floating point numbers, should I round 'em off or truncate 'em?
>See, the spec doesn't really say.

We got hit by this problem a little while back when *Adobe*
Illustrator '88 (v1.8.3) created a file with a BoundingBox
comment that had real numbers such as "54.125" (I don't know
if it's significant, but all the fractions were multiples of
an eighth).  I would suggest that any properly written utility
be able to accept not-quite-PS-Adobe-conformant comments, possibly
issuing a warning indicating what action it is taking.

Such a utility should neither round nor truncate, but should
rather round the lower left co-ordinates down and the upper right
co-ordinates up, so as to guarantee a valid bounding box for the
document.

John
-- 
john j. chew, iii   		  phone: +1 416 425 3818     AppleLink: CDA0329
trigraph, inc., toronto, canada   {uunet!utai!utcsri,utgpu,utzoo}!trigraph!john
dept. of math., u. of toronto     poslfit@{utorgpu.bitnet,gpu.utcs.utoronto.ca}