Follow-up Comment #6, bug #67544 (group groff): At 2026-01-27T18:33:14-0500, Deri James wrote: [snip] > I think this means that the meta-data in both fonts (at least the > glyph widths (the text example did not exercise the italic > correction)) is sufficiently close to not be visibly detectable.
Thanks for checking this out!
> Of course the actual glyph shapes are identical because the glyphs for
> both TR and U-TR come from the same URW font. So, I did a third run,
> again no kerning, but ne-TR.pdf produced via grops->ghostscript. This
> time there is a visible difference - but not to the glyph shapes, just
> a very slight vertical shift. This is likely to be a rounding issue,
> grops uses 2 decimals gropdf 3 decimals rounded, don't know what
> ghostscript uses. The fact it repeats every 4 lines points towards
> rounded calculation landing on pixel boundaries when the glyphs are
> filled.
I'm intrigued. Do you think it's worth trying to increase the precision
of grops's calculations? Or is this limitation on significant figures
imposed on us by the PostScript file format?
_______________________________________________________
Reply to this item at:
<https://savannah.gnu.org/bugs/?67544>
_______________________________________________
Message sent via Savannah
https://savannah.gnu.org/
signature.asc
Description: PGP signature
