search

gongshoudao / noto

Automatically exported from code.google.com/p/noto
0 stars 0 forks source link

Shape of GURMUKHI LETTER LLA (Gurmukhi) need to correct. #113

Open GoogleCodeExporter opened 9 years ago

GoogleCodeExporter commented 9 years ago 
What steps will reproduce the problem?
1. http://utrrs-testing.rhcloud.com/language/pa/codepoint
2.  OR Check http://www.unicode.org/charts/PDF/U0A00.pdf Unicode Chart for 
Gurmukhi
3. ceck Image of Unicode Character A033 (ਲ਼)
4. Check Position of  GURMUKHI SIGN NUKTA (਼) with 0A32 (ਲ) GURMUKHI LETTER 
LA

What is the expected output? What do you see instead?
Current:
GURMUKHI SIGN NUKTA is on Left Side

Expected:
Should be on Right Side as per Unicode Image.

What version of the product are you using? On what operating system?
commit : e517fb85f7ee6beaa7204063cbe30709c4649617

Image Attached for Reference (compared with Unicode Chart (Gurmukhi))

Original issue reported on code.google.com by amanpree...@gmail.com on 4 Aug 2014 at 12:19

Attachments:

GoogleCodeExporter commented 9 years ago

Original comment by roozbeh@google.com on 4 Aug 2014 at 3:50

GoogleCodeExporter commented 9 years ago

Original comment by roozbeh@google.com on 16 Jan 2015 at 6:05

GoogleCodeExporter commented 9 years ago 
Yes, the shape of LLA (either as U+0A33 or as <0A32, 0A3C>) in Noto Gurmukhi 
fonts is different from the Unicode charts and the Windows fonts. The dot is on 
the lower left side in Noto, while it's on the lower right side in the Unicode 
charts and the Windows fonts.

It seems that the location of the dot on the bottom right in Unicode was due to 
a request from the Government of India in 2003. Here is their document: 
http://www.unicode.org/L2/L2003/03101-indic-dir/Gurmukhi.pdf (see top of page 
3).
Jelle, is the positioning in Noto intentional, or a bug?

Original comment by roozbeh@google.com on 2 Apr 2015 at 6:14

GoogleCodeExporter commented 9 years ago 
The positioning is intentional, in the sense it follows the general principle 
for nukta positioning: if possible use a place where the nukta does not get in 
the way of other low marks.

My Panjabi book happens to be from 2003, and while it mentions the other 
"paireen bindee" characters, not this one. 

Searching through the list with Panjabi words I do not see U+0A33, but various 
words with <0A32, 0A3C>, which is occasionally followed by virama, uu vowel or 
a subscript va. So that is why the nukta is on the left.

Original comment by jelle.bo...@monotypeimaging.com on 2 Apr 2015 at 6:50

GoogleCodeExporter commented 9 years ago 
Jelle, the letter LLA is part of Unicode's normalization composition 
exclusions, so that's why you've been seeing it as <0A32, 0A3C>. But for all 
practical purposes, the two sequences <0A33> and <0A32, 0A3C> should act 
exactly the same way, since they're canonically equivalent.

I just asked our reviewers for some feedback on the position of the dot.

In the meanwhile, I'm attaching the rendering of <lla>, <lla, virama>, <lla, 
uu>, and <lla, virama, va> in Noto, Nirmala, and Raavi, for reference. Raavi is 
peculiar in the sense that its dot moves to the left when there's a virama 
after it, but not in other cases.

Original comment by roozbeh@google.com on 6 Apr 2015 at 1:01

Attachments:

GoogleCodeExporter commented 9 years ago 
In the case of Raavi: In Windows 7 the dot is still to the left. In Windows 8.1 
it has moved to the right, but the ligature with virama was overlooked. Which 
explains.

It is actually used with the ha-subscript rather than va. 

It seems very odd that one of the consonants has the dot to the right and the 
others to the left. It makes no visual sense. To the left, the nukta belongs to 
the consonant and is always clear. Nirmala shows clearly why the right is less 
than ideal: the dot belongs to the subscript rather than consonant.  Noto is 
hardly the only font which has it to the left.

Original comment by jelle.bo...@monotypeimaging.com on 6 Apr 2015 at 1:28

GoogleCodeExporter commented 9 years ago 
... and sometimes it is centred!

Original comment by jelle.bo...@monotypeimaging.com on 6 Apr 2015 at 1:41

GoogleCodeExporter commented 9 years ago 
Here's feedback we received from our reviewer, who is also a font designer:

"We had a similar bug logged against the initial builds of two other fonts, in 
which we had located the nukta on the lower left of the La. Like you, we 
researched the issue when the bug was logged, and came across the Gov't of 
India document you cited. So we changed the position of the nukta do to the 
lower right.

I don't have print examples, because this letter is very rare. It isn't part of 
the standard Punjabi alphabet, and I believe is only used to transcribe foreign 
loan words. It doesn't occur in either of our Punjabi grammars, or in the 
volume of poetry in which one of the fonts was first used. Given this, we 
decided the best course of action was to follow the Gov't of India request."

Amanpreet, do you have access to more resources discussing or showing the 
position of the dot on lla?

Original comment by roozbeh@google.com on 6 Apr 2015 at 7:00