Path: utzoo!attcan!utgpu!jarvis.csri.toronto.edu!rutgers!tut.cis.ohio-state.edu!rpi!pawl5.pawl.rpi.edu!kibo From: kibo@pawl.rpi.edu (James 'Kibo' Parry) Newsgroups: comp.fonts Subject: Re: Re^2: How flexible are the Macintosh outline fonts? Message-ID: <5981@rpi.edu> Date: 14 Jul 89 18:34:47 GMT References: <1989Jul12.111024.10324@sics.se> <812@acorn.co.uk> <1989Jul14.104844.17925@sics.se> Sender: usenet@rpi.edu Organization: The Site Of The Kibology Annex for Scotia/Schenectady NY Lines: 28 In article <1989Jul14.104844.17925@sics.se> dan@sics.se (Dan Sahlin) writes: > "The excellent kerning function in Calamus is also using the >vectorised fonts and using them automatically computes the most >appropriate distance between the characters". I believe the mechanism Calamus uses is that in the font file, for each character, there are 16 numbers -- 8 describing the widths of the character at eight different "latitudes" on the left, and 8 for the right. (I think eight is the number.) This way it can fit letter together by saying to itself, "This letter is wide at the top and narrow at the bottom, but this one's perfectly straight... hmm, how close can I make them..." Of course, not being privy to the secrets of the designers of Calamus, etc., I may be off the mark somewhat. I've deduced the above by dumping some Calamus demo fonts (like CGLOGO.CFN, the CompuGraphic insignia) in hex and spending a few hours examining them; I'm trying to write a program to interchange font formats. Whatever mechanism Calamus uses, ultimately, the characters wind up very nicely spaced... the output is beautiful. (I like how it uses the outline fotns on both the screen and printer; there are no seperate bitmapped screen fonts. WYSIWYG down to the pixel level...) james "kibo" parry, 138 birch lane, scotia, ny 12302 usa / All colors are kibo%pawl.rpi.edu@itsgw.rpi.edu _________________________/ arbitrary. kibo@mts.rpi.edu / Kibology / Anything I say is the opinion userfe0n@rpitsmts.bitnet / is better! / of myself, and not of Xibo.