At 2023-05-19T21:30:31-0500, Dave Kemper wrote: > Regarding the conventional space width, this patch added this > parenthetical: > > > current font's space width (usually one-third em for Western scripts) > > But this claim is not reflected in the data for the fonts shipped with > groff for the two most common typesetting devices, ps and pdf. > > $ fgrep unitwidth font/devp*/DESC.in > font/devpdf/DESC.in:unitwidth 1000 > font/devps/DESC.in:unitwidth 1000 > > So for these devices, 1/4 em would be 250 units, and 1/3 would be 333. > The actual space widths for the fonts (pdf inherits the metrics from > ps, I believe) are:
Yes, because the base fonts for PS and PDF are supposed to be metrically compatible. And in the early days, the font files themselves were often identical, as I understand it. > Disregarding the four monospace fonts, which require their own spacing > rules, 29 of those fonts are closer to 1/4 em, and 5 of them to 1/3 em > (though few are exactly either 1/4 or 1/3). > > This aligns with what our friend Heraclitean River > (http://web.archive.org/web/20171217060354/http://www.heracliteanriver.com/?p=324) > observed in his research: "Margins became smaller [in the first half > of the 20th century], and standard interword spaces often went from > about 1/3 em to 1/4 em." That's a resource that just keeps paying dividends. > Incidentally, where the same parenthetical phrase is inserted into > groff_diff(7), it introduces mismatched parentheses (not obvious as > they go three levels deep). Thanks, Dave. I've fixed these problems in my working copy and they'll be in my next push to master--a matter of hours, perhaps. Regards, Branden
signature.asc
Description: PGP signature
