Re: Myanmar questions

From: Antoine Leca (
Date: Mon Feb 19 2001 - 06:30:11 EST

Hi William,

Many thanks for your valuable help.

William W.L.K wrote:
>> 1) for the "au" dependent vowel, I believe (extrapolating from the one for "o")
>> the correct encoding is U+1031 U+102C U+1039. However the use of the virama
>> inside of a "matra" part looks surprising to me (and it creates problem to my
>> renderer).
> *************williaMM
> what looks surprising to you?

The fact that the virama have to be part of a non-halanta syllable.

> what problem does that cause?

This is different from what happens with the other scripts.

> what renderer? please give me the Internet Link so that I can download and
> test it thru.

You can't: I am designing an Indic script engine to be put in front of
FreeType 1.x OpenType (and if things work well, it will be used with FreeType 2).
But there is nothing to release publicly at the moment.

>> 2) There appears to exist a special vowel usually named "ui", which looks like
>> as a combination of i (above) and u (below). How is it supposed to be encoded
>> in Unicode? u before i (as pronounced)? i before u (as usual with Unicode,
>> above before below)?
> *************williaMM
> a special vowel? "ui"?

Sorry, I do not know enough of Myanmar grammar to explain properly.

> talking about the separate one,
> i before u (above before below) is only accepted in our approved
> handwriting literature.
> For e.g. U+1005 U+102D U+102F = the word "wet"

OK, you got it I believe. From what I read, this means that "wet" canNOT be
written as U+1005 U+102F U+102D. Do I get it right?

>> 3) The vowel bearer (1021) is reported to be the one to use at intial when
>> there is no consonant, along with the appropriate vowel sign. However,
>> Unicode also
>> encode the individual glyphs for the independant vowel which does not look like
>> the bearer+the vowel sign. I.e. there does not exist Long A (a space is
>> available at 1022), so I understand I have to encode it as U+1021 U+102C.
>> However, for
>> short i, I can use either U+1021 U+102D, or U+1023. What is the preference?
> *************williaMM
> The preference is that they have only got the same pronuciation.
> You use either U+1021 U+102D, or U+1023 based on the words you are writing.

You mean, there are words which are written with U+1021 U+102D, and the initial
"i" sound canNOT be written with U+1023?

> For a better visualization, U+1021 U+102D plus U+102F gives you the word
> ("Old") whilst U+1023 plus U+102F gives you a totally "meaningless" word
> that we don't have it in our dictionary.

I believe this is different. Here, I see only the independent "ui" vowel (and
I know it is written with the bearer+both signs for "i" and "u"). I believe the
problem for me is that as far as I understand that in fact the "u" sign here
really indicates a tone, not a different sound. Correct?

> My emphasis is that even when two or more consonet-vowel or vowel-vowel
> signs combinations produce the same sound (like above),
> they can't replace each other....this is our language barrier. You just
> have to learn the writing techinques..that would only help.

That is correct. That is exactly what I am searching after...

>> 5) I can't figure how looks like "kywe". Is it base_ka + wa_below +
>> ya_to_the_right (but then what is the difference with "*kwye"?), or is it
>> base_ka +
>> ya_to_the_right + a_special_wa_deep_below, the latter being under the
>> "arch" of the ya?
> Ehhhh!!! I'm not sure if I understand your questions "kywe" and "*kwye".....

Sorry, I should have made it explicit. the asterisk is the standard mark
to design a reconstructed or invented form that gets no evidence.

Also, I should add that I missed the the leading "e" vowel in my designs.

> but, understanding from your drawing, "*kwye" is never acceptable in our
> writing....
> I enclose the best you should expect out of your rendering.......

OK, for the benefit of the people on the Unicode list (where the attached
file has been dropped), here is the sketch:

      ___ ____ ____ |
     / \ / \ / \ |
    / / \ / \ |
   | | | | |
   | _____ | | | |
    \/ / \ / / |
__ \___/ ___/ | __=baseline
                            / \ |
                           | | |
                            \__/ \__|

Yes, that is ugly drawing. The point, as I understand, is that the "w"
does not transform it self into a triangle, but keeps its normal rounded
from although it goes below.
Another special case for my renderer, in fact.

Again, many thanks for your answers, they are very precious to me.
As soon as I got viewable results, I certainly will give the informations
to test them. Keep connected...


This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:21:19 EDT