Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!decvax!ucbvax!ucbcad!nike!lll-crg!seismo!gatech!cuae2!ihnp4!houxm!hoqam!hoqax!twb
From: twb@hoqax.UUCP (BEATTIE)
Newsgroups: net.graphics
Subject: GKS string and char precision
Message-ID: <277@hoqax.UUCP>
Date: Tue, 14-Oct-86 15:24:29 EDT
Article-I.D.: hoqax.277
Posted: Tue Oct 14 15:24:29 1986
Date-Received: Thu, 16-Oct-86 07:34:17 EDT
Distribution: na
Organization: AT&T Bell Laboratories, Holmdel
Lines: 35

The implementation of GKS I am using behaves as follows:
	1. If I request a HARDWARE font before any other fonts - I get
		the correct font.
	2. If I request a SOFTWARE font anytime - I get the correct
		font.
----->	3. If I request a HARDWARE font, after a SOFTWARE font, and
		request other than STRING precision - I get a SOFTWARE
		font. I'd rather have the HARDWARE font and char
		precision.

The reasoning behind #3 seems to be that since a hardware font can't be
clipped to stroke or char. precision and a software font is available --
the software font is used.

I don't understand why a hardware font can't be clipped to char
precision - although it may mean the device driver must plot one char at
a time in this case (too bad :-).

This is a hassle for me because I need to check which font has been
requested (by a user) before I can specify the precision.

It would be much preferable to have GKS determine if the requested font
could be clipped as requested and if not then use a more restrictive clip.
(i.e. demote stroke to char and char to string as needed)

Comments? (Did I have to ask? :-)
---
Yinkel - n. A person who combs his hair over his bald spot, hoping
no one will notice.

Tom.
T.W. Beattie
...!{ihnp4 | houxm | whuxl | ulysses}!hoqax!twb
...!{decvax | ucbvax}!ihnp4!hoqax!twb
These opinions are mine, all mine, and nobody else's