[Eng] [DejaVu-fonts] New glyph for U+014A

[Michael Everson]

On 9 Nov 2013, at 22:39, Gilbert Adjoyi <gilbert_adjoyi at att.net> wrote:

> Please permit me to comment on the issue under discussion.
> Contrary to Michael's observation that disunifying  would be unnecessary,

My contention is that it is necessary to disunify Eng from African Eng (or Eng from Sami Eng, whichever way it goes).

> I would say that for Ewe, one of the Ghanaian (West African) Languages that need the n-form capital Eng, the dissociation would ensure that the capital Eng never picks the N-form irrespective of the font in use.

Right. You don’t want the N-form capital Eng ever. And the Sami don’t want the n-form capital Eng ever. 

> Ewe does not need the N-form Capital Eng but other languages need it. So both are necessary but we need Ewe not to pick up the N-form Capital Eng at anytime as that would distort the word for which it is used.

By “distort” I take it that Gilbert means that using the N-form capital Eng is a *misspelling*. That is a definite sign of a character split. 

On 9 Nov 2013, at 21:54, Denis Jacquerye <moyogo at gmail.com> wrote:

>> It is still a character-based distinction that would leave the Sami letter alone.
> 
> What makes you think n with long leg is not a wrong glyph for n-form eng, considering the first can have a bottom serif while the second usually has a left hook?

I don’t. I provoke discussion. 

> Furthermore, when capital N with long leg was encoded, it wasn't clear whether it was going to be N-form or n-form. 

What?

> What if tomorrow some users decide they prefer the N-form?

> It seems you didn't read what I wrote.
> The eng/n with long leg disunification made things harder for Lakota,
> not better in the long term.

The n-with-long-leg was encoded for support of older Lakota documents. We have many, many, many characters encoded for support of older orthographies. I use many of these characters every day. 

> We are deeply grateful for you[r] work, that doesn't mean disunification
> never creates problems.

Eng remains a problem. 

>>> FileMaker, or the application that displays your file system, needs to
>>> support OpenType properly in general, not just for this case.
>> 
>> You expect database software to treat individual strings of plain text differently from field to field? You’re dreaming.
> 
> Use XML or databases that allow metadata.

Spelling should be correct in plain text. 

>> This is a very short-sighted view. The UCS will be with us for centuries. Why not FIX this problem?
> 
> Yes, let’s fix it. There are other tools than disunification.

They won’t work in plain text. They already don’t work in plain text. 

> In the long run better localization and OpenType support is a necessity. Why not fix this problem?

It’s NOT necessary for every other Latin letter in every other orthography out there. 

> UCS is about character identity, not about preferred glyph which might change depending on the context, culture, period or area.

The lines have already been drawn. 

Michael Everson * http://www.evertype.com/