That is indeed the situation with the current technology, which is kind of my point. It would seem more intuitive to base the logic of a designspace file more on the design than the technical limitation. Otherwise, where should one define the default design representation?
-
-
Replying to @NickSherman @kioskfonts and
In an 8 master, 3 axes system, if a non-aligned off-axis location is assigned as the new default, the generator than has to create a lot more masters to “keep the tent up” (
@djrrb’s metaphor). 1/2pic.twitter.com/USY1c9t80a
2 replies 0 retweets 14 likes -
Replying to @letterror @NickSherman and
Great sketches! Very helpful. But ... I'm still a bit confused why it wouldn't work to just add the single, calculated master. If the 8 corners can "keep the tent up" already, why would an intepolated source disrupt that? I may just need to test it to see why
1 reply 0 retweets 0 likes -
Replying to @ArrowType @NickSherman and
If you want that interpolated source to be the default, with current technology and math, that master needs to be on the default of all axes. Hence lots of masters to reconstruct those. But maybe I'm not understanding your question..
2 replies 0 retweets 1 like -
Replying to @letterror @NickSherman and
I think you understood my question, and I appreciate the added explaination. It is still hard for me to get *why* the default must be on all the axes in the first place. But, I probably just need to look at how the gvar table relates to the glyf table.
1 reply 0 retweets 0 likes -
Replying to @ArrowType @letterror and
Attempt: the default master is the only “font” that is stored in the VF as actual, old-school outlines. Everything else is stored in the VF as deltas from the points of those default outlines.
1 reply 0 retweets 4 likes -
Replying to @kioskfonts @letterror and
Thanks – I do understand that it is a current limitation, but I'm just not sure why/whether it is a hard mathematical limitation. If a designspace works for a given area, why would pre-calculating one location in that area require pre-calculating any others?pic.twitter.com/O8bsa49nmy
1 reply 0 retweets 1 like -
Replying to @ArrowType @kioskfonts and
My thinking: In a future OpenType spec, maybe there could be a "default" location in the gvar table, and other deltas would be relative to that. However, the glyf table outlines would be pre-calculated to load first / load in legacy software.
1 reply 0 retweets 0 likes -
Replying to @ArrowType @kioskfonts and
So, a better illustration of that would be this:pic.twitter.com/69LQW5IkFS
1 reply 0 retweets 3 likes -
Replying to @ArrowType @kioskfonts and
Definitely believe v1.8 has issues here
1 reply 0 retweets 1 like
We can (re-)define the basis functions so they're zero at the default setting. Then you can change the default freely. This is one part of a proposal I have to improve interpolation, which I will publish shortly.
-
-
Replying to @raphlinus @davelab6 and
And by shortly I mean now :)https://github.com/OpenType/opentype-variations/issues/7 …
1 reply 2 retweets 10 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.
