Re: Measuring "ideographic character face" and commonest characters

On 09/24/2016 10:20 PM, Martin J. Dürst wrote:
> Hello Fantasai,
>
> What I have done is download ISO-10646 (publicly available) and looked at the columns/sources. Essentially, you want a
> character that shows up in all columns, with the most basic source (G0, HB1, T1, J0, K0, V1).
> (G=China, H=Hong Kong, T=Taiwan, J=Japan, K=Korea, V=Viet Nam)
>
> On 2016/09/25 00:20, fantasai wrote:
>> The CSS Inline spec has an appendix for synthesizing missing baseline
>> information
>> from glyph outlines of specific characters. From a comparison of several
>> fonts I
>> have on my system, it seems that the top and bottom bounds are best
>> found using
>> 丅 and 丄,
>
> Please be careful. 丄 doesn't appear in K or V sources, and 丅 also not in H sources.

Ah, right. I meant 上 and 下!

>> 回因
>
> These seem okay.
>
>> 困
>
> Seems okay.

> These seem okay:
>> 囚回因固圃圈國圍園圓圖團
>
> I'd probably go with some of the later ones rather than 回 or 因, because as a general tendency, the more content, the higher
> the chance that they are a bit wider. But please check for yourself, too.

My concern as we go higher in complexity & rarity is that some fonts
might leave them out. 因 is reasonably common. 回 is also common but
I'm not sure that it's thought of as part of the same series, whereas
I'm pretty sure 因 and 困 are linked up to the rest and would therefore
share their proportions.

(That said, my education in writing was quite delinquent. "Here are
some random characters which are a collection of strokes with no
particular relation to each other, go memorize them for next week.")

~fantasai

Received on Tuesday, 27 September 2016 02:32:28 UTC