Out of interest, why did you pick the hhea values to always be primary? Two main issues that I see: 1) it ignores a foundry’s decision to set the useTypoMetrics flag (or not). There could be a reason why winAscent/Descent is better than usTypo for a given font.
-
-
Introducing that flag, like introducing any other flag late in the game, was an incredibly silly idea. (Entertaining type designers with how to define data that application developers would ignore anyway – including those working for the company that had introduced it.)
3 replies 0 retweets 0 likes -
While I think MS Office should follow the OT standard, I understand too well why it doesn’t right now (years of cruft and large corporate policy). But the standards are public. So new applications should be expected to align. Or do you just want to give them a pass?
1 reply 0 retweets 0 likes -
That’s a pretty peculiar way to look at it.
2 replies 0 retweets 0 likes -
Replying to @KLTF @aaronbell and
IIRC: Apps correctly chose OS/2 Typo values. Except for MS Office apps. MS didn’t like to fix that, there, instead introduced a funny flag – leaving it to type designers to fix MS Office apps’ bug. Or not. MS Office apps, turned out, didn’t bother to respect that flag for years.
2 replies 0 retweets 1 like -
Replying to @KLTF @aaronbell and
Use of Win metrics for linespacing wasn't limited to Office apps: it was near universal on Windows. And so many fonts had subsequently been made with that assumption, that a lot of Typo metrics were unreliable (early 2000s, metrics dara reliability has improved since).
2 replies 0 retweets 1 like -
Replying to @TiroTypeworks @KLTF and
Simply changing from using Win metrics to using Typo metrics would have reflowed or broken a lot of existing documents using fonts that had inconsistent or incorrect Typo metrics. So MS wanted a solution that could be used in new fonts, but not affect layot for existing ones.
1 reply 0 retweets 1 like -
Replying to @TiroTypeworks @KLTF and
The useTypoMetrics flag is supposed to indicate that the Typo metrics data are reliable, and hence should be used.
1 reply 0 retweets 1 like -
Replying to @TiroTypeworks @KLTF and
Thats a great explanation. Do you know if they are, in practice? (If there is one thing I learned from this side of the pond is that font metadata cannot often be trusted verbatim en masse without an extra layer of historical context and interpretation.)
2 replies 0 retweets 4 likes -
My own approach has been mostly to set the useTypoMetrics flag only when the Win metrics need to differ from the Typo metrics, and always when I've carefully considered the latter and really do want them to be used. So I can say, yeah, the flag is reliable for my fonts.
2 replies 0 retweets 1 like
Thanks! Might do some analysis on a large corpus of fonts.
-
-
Will be interested to see stats for a) Typo/Win equivalent, bit not set b) Typo/Win equivalent, bit set c) Typo/Win not-equivalent, bit not set d) Typo/Win not-equivalent, bit set Might also be interesting to cross reference to date of font creation.
1 reply 0 retweets 0 likes -
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.