Re: Issues with DA,NA,GA default medial variants

Hi Greg,

On 20.11.2015 08:21, Greg Eck wrote:
>
> Hi Badral,
>
> I think we have already said that one of our most clear guiding 
> principles is that we do not disturb the existent encoding.
>
Did you read the point 2?
>
> The current Medial defaults for U+1828, U+182D, U+1833 have been set 
> back as far as the year 2000.
>
Do you know, that Mongolian script was the only existing Unicode font 
until 2004? We have never changed our encoding for above characters. 
Current problem established due to ignorance of us and TR 170.
Please see page 13-14 of TR 170 more carefully. We have to respect 
standards!
>
> So, there should be no discussion on changes of this proportion.
>
?

Badral
>
> Greg
>
> *Sent:*Thursday, November 19, 2015 10:31 PM
> *Subject:* Re: Issues with DA,NA,GA default medial variants
>
> We have decided, that we don't change our implementation (BS column) 
> of medial default form of DA, NA, GA due to following reasons.
> 1. Logic and Ambiguation
> I already mentioned the representation logic in my previous emails. 
> Now plus:
> NA is distinguished only by dot from A.
> GA is distinguished only by dot from QA.
> DA is distinguished only by horizontal tie from TA.
>
> 2. Established almost no destabilization
> As Jirimutu described there exist almost no destabilization because no 
> FVSs stored in middle of a word. 
> https://lists.w3.org/Archives/Public/public-i18n-mongolian/2015OctDec/0123.html
> (cutting:
>
> Actually we are not typing and storing FVS1 in the complete Mongolian Word actually,
> It is automatically been selecting correct display form in the word.
> Only when we show the separated Medial Form individually, the FVS1 is necessary.
> For this reason, maybe you are misunderstanding the encodings.
>
> )
> I didn't understand what is the separated Medial Form. If it's not 
> ISOLATED one, then we should use ZWJ or?
> In our stem database, there exist no stem in middle form of DA, NA, GA 
> with FVSx. The most words in Greg's example are initial DA which is 
> also stored in our DB with FVS1 except suffixes. (Dun and Dugnelt 
> stored in our DB also with FVS because it's an initial second form of DA.)
> I think, the misunderstanding was established by magic of OT grammar. 
> The actual problem could be just changing OT grammar rules for font 
> developers. If someone has difficulties, we are ready to help.
> Can you accept it Jirimutu?
>
> Greg: Please accept and update your NP column as we requested unless 
> you have contra arguments.
>
> cheers,
> Badral
> >>>>>
>


-- 
Badral Sanlig, Software architect
www.bolorsoft.com | www.badral.net
Bolorsoft LLC, Selbe Khotkhon 40/4 D2, District 11, Ulaanbaatar