RE: Two Final Threads - Diphthongs / Final glyph checks

Hi Jirimutu,

OK, I think I follow you now.
Your algorithm looks complete.
I know when I was working on the U+1836, it was very difficult, however it did seem to work.
Give me two weeks while in China to look it over and I will get back to you.
So, just to recap, you are suggesting that we do not need the Medial+FVS2 over-ride glyph whereas Badral is suggesting that we do need the Medial+FVS2 over-ride glyph.
I will compare your implementation with the Baiti implementation and see what I come up with.
As of now, just looking at the algorithm, I cannot say much, but in testing on the actual font, will be able to see more clearly.

Greg

>>>>>>
Sent: Saturday, November 21, 2015 12:07 PM
Subject: RE: Two Final Threads - Diphthongs / Final glyph checks

Hi Greg

>But could I ask Jirimutu – is this what you are proposing that ligature sequence <U+1836><U+1822> [cid:image003.png@01D12470.DE1266A0]  be encoded as one glyph?
>This would be typed as code-point U+1836 plus one code-point U+180C?
No. It is not my proposal meaning. The medial form [cid:image003.png@01D12470.DE1266A0]   should always be  <U+1836><U+1822> exactly.  No extra FVS after<U+1836> in anytime and any case.
What I am insisting here is that we cannot  sacrifice this regular medial form become any kind of irregular encoding.
(Like Badral’s propose in recent mail to use FVS2 or FVS3 to over-ride, actually it is not over-riding, it is replacing beams and pillars with inferior ones)

Let me clearly summarize about the <U+1836_YA> medial form encoding proposal rule entirely again.


1.     We accept the current NP Variant Form encoding for <U+1836_YA> exactly. It is the [https://r12a.github.io/scripts/mongolian/v/1836i.png]  is the first medial form o and [https://r12a.github.io/scripts/mongolian/v/1836m.png]  is the second medial form.

For the people who want to use <U+1836_YA> and <U+1822_I> to encode [cid:image009.png@01D12470.DE1266A0] .

2.     No extra contextual rule for [cid:image009.png@01D12470.DE1266A0]  if anyone want to use <U+1836_YA> and <U+1822_I> to encode it, the code should be explicitly as <U+1836_YA><FVS1><U+1822_I>.

(Note: we do not agree to use contextual rule here like if <U+1836_YA> before <U+1822_I> become [cid:image009.png@01D12470.DE1266A0] ,

because that will sacrifice The medial form [cid:image003.png@01D12470.DE1266A0]   should always be  <U+1836><U+1822> requirements.)

3.     But the [cid:image009.png@01D12470.DE1266A0]  followed after <U+1822_I>, <U+1825_OE>, <U+1826> in first syllable, it will become [https://r12a.github.io/scripts/mongolian/v/1836m.png] .

We need to explicitly define its encoding for the people who want to use <U+1836_YA> for this long tooth.

I have provided examples in previous mail like ᠦᠢᠯᠡᠰ · ᠰᠦᠢᠳᠦᠯ · ᠰᠢᠢᠳᠪᠦᠷᠢ etc.

For example, ᠦᠢᠯᠡᠰ should be encoded as <U+1826_UE><U+1836_YA><FVS1><U+182F_L><U+1821_E><U+182F>.

The first long tooth is the part of <U+1826_UE> and The second long tooth is <U+1836_YA><FVS1>. We cannot neglect this FVS1 too.

For the people who want to use <U+1822_I> to encode [cid:image009.png@01D12470.DE1266A0] .


4.     The [cid:image009.png@01D12470.DE1266A0]  is encoded as <U+1822_I> and it will follow the <U+1822_I> character rule in here.

(I have skip the details here. Please refer previous Greg’s over-riding rule definition document for <U+1822_I>)


5.     But in this case the [cid:image009.png@01D12470.DE1266A0]  followed after <U+1822_I>, <U+1825_OE>, <U+1826> in first syllable, it will become [https://r12a.github.io/scripts/mongolian/v/1836m.png]  too.

We need to explicitly define its encoding for the people who want to use <U+1822_I> for this long tooth.

I have provided examples in previous mail like ᠦᠢᠯᠡᠰ · ᠰᠦᠢᠳᠦᠯ · ᠰᠢᠢᠳᠪᠦᠷᠢ etc.

For example, ᠦᠢᠯᠡᠰ should be encoded as <U+1826_UE><U+182F_L><U+1821_E><U+182F>.

The first long tooth is the part of <U+1826_UE> and The second long tooth is <U+1822_I>. We do not need any more FVS here.

All of Mongolian font should implement all of the encoding rule listed above.
No exception here is for we can correctly display each other’s context in any fonts.

The same thing will be listed for <U+1838_W>. Let me skip it here for simplify my mail.

Jirimutu