Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!usc!cs.utexas.edu!uunet!munnari.oz.au!ariel!ucsvc.ucs.unimelb.edu.au!wehi.dn.mu.oz!baxter_a
From: baxter_a@wehi.dn.mu.oz
Newsgroups: comp.sys.amiga.programmer
Subject: Re: Line-thickness drawing attribute
Message-ID: <1991Apr7.174600.24545@wehi.dn.mu.oz>
Date: 7 Apr 91 17:46:00 GMT
References: <1991Mar27.155056.27758@decuac.dec.com> <24200@wehi.dn.mu.oz> <965@boing.UUCP> <20310@cbmvax.commodore.com>
Organization: Walter & Eliza Hall Institute
Lines: 36

In article <20310@cbmvax.commodore.com>, spence@cbmvax.commodore.com (Spencer Shanson) writes:
> In article <965@boing.UUCP> dale@boing.UUCP (Dale Luck) writes:
>>In article <24200@wehi.dn.mu.oz> BAXTER_A@wehi.dn.mu.oz writes:
>>>Ah! Drawing - routines - that - need - to - be - available - and - aren't:
>>>
>>>Line thickness is needed.
>>>
>>>Lines that bend at nodes restart the line pattern instead of continuing it.
>>>Thus you can't draw a patterned curve using small segements.
>>
>>What? Graphics.library does not restart the pattern  at line joins.
>>Make sure you are not using a move.
>>Move resets the pattern count, Draw does not, string your lines
>>together with Draws.
>>
>>It is a new bug if it happens any other way.
> 
> Dale - the Draw() code was *really* confused about the rp->linpatcnt value
> after drawing through cliprects. This is now fixed.


Thanks. But now I've spent 1/2 hour stuffing around with it, I thought I'd
post what I found, anyway...

My line pattern was half pixels on, then the other half off.
My line was horizontal.
The line was made of segments (variable number) using Draw(), no Move()'s.
One segment, no problems.
Several segments, looks okay, but the more segments, the worse it looks,
line becomes moth eaten.
Segment ending at every pixel, no pattern at all.

Looks a bit like Draw looses track in the pattern and drops back one
pixel at each node.

Regards Alan