From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from mimir.eigenstate.org ([206.124.132.107]) by ewsd; Fri Nov 6 11:07:56 -0500 2020 Received: from abbatoir.fios-router.home (pool-74-101-2-6.nycmny.fios.verizon.net [74.101.2.6]) by mimir.eigenstate.org (OpenSMTPD) with ESMTPSA id aeb29d85 (TLSv1.2:ECDHE-RSA-AES256-SHA:256:NO) for <9front@9front.org>; Fri, 6 Nov 2020 08:07:47 -0800 (PST) Message-ID: To: 9front@9front.org Subject: Re: [9front] truetypefs Date: Fri, 06 Nov 2020 08:07:45 -0800 From: ori@eigenstate.org In-Reply-To: <012440E4E4FD8FD9B1E1955DA8640092@sdf.org> MIME-Version: 1.0 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: 8bit List-ID: <9front.9front.org> List-Help: X-Glyph: ➈ X-Bullshit: proxy rich-client cloud > Hi all, > > truetypefs used to crash for me; it used to happen when trying to display > some non-ASCII glyph, e.g., an accented vowel: à. So, I remember looking into this a while back, and I asked aiju about it at the time -- here's the response: aiju | Ori_B: there is a dumb corner case i didn't expect any font to hit aiju | but some fonts hit it anyway Ori_B | oh? aiju | Ori_B: where ALL the points on a curve are off-curve aiju | this might be the same bug Ori_B | oh... hm. aiju | it's probably not too hard to figure out but i have all this shit paged out right now Ori_B | I can verify that at least. aiju | i think you need to pick two points and calculate the average and use that as a starting point Ori_B | that sounds reasonable. aiju | i literally just didn't implement it because i figured no font would do this shit Ori_B | yeah, it's some unicode junk that I appear to have accidentally gotten into the file, so I didn't hit it on a common glyph. Ori_B | should we also return a 'bogus glyph' instead of killing font rendering? Ori_B | '?', pjw face, something?