Mark ==== wow. What an overreaction from Mark Pilgrim[1]. While sound & fury may grab eyeballs, he is ignoring solutions already in front of his face[2]. The

Mark ==== wow. What an overreaction from Mark Pilgrim[1]. While sound & fury may grab eyeballs, he is ignoring solutions already in front of his face[2]. The Solution ============== I worked with fonts and licensing all the time in a previous job, generating automated artwork. The big thing missing in Open Source [his 'usable'] fonts is kerning and hinting. There is nothing stopping someone setting up a Kern-or-Not? website where web people vote on which kerning values look best, and incorporating that data set into an OSS font foundry[3]. I already did some of this myself to kern HTML using spans with negative margins. To get excellent results you only need a few key characters [in English]. I used Web fonts, but made up the kerning offsets by trial and error. There is nothing stopping other people doing the same [see 2]. 'T', 'W' are the big offenders. You can get excellent results concentrating on these and all the vowels. T e , T o , T y , T i , T e, W * etc. This is not exhaustive. ------------------------- [1] I came across his emails all the time when I was working on the OSS Greasemonkey in the early days of its development. He threatened the lead dev with 'forking the project and working on it until the day I die' if the project did not take a direction he wanted. I believe it was another of one of the fequent LGPL/GPL arguments that whirlwind in many OSS projects. The mailing list greasemonkey-[user|dev] doesn't seem to include such old emails [c 2005], so no link, sorry. [2] Mark seems to prefer that other people fix the problems under his guidance. [3] May I suggest Public Domain for the dataset to avoid horrendous license choice debates?