Kannada letter design issue ಪು, ಪೂ, ಪೊ, ಪೋ, ಫು, ಫೂ, ಫೊ, ಫೋ

[MayuraVerma]

Defect Report

Kannada letter design issue with ಪು, ಪೂ, ಪೊ, ಪೋ and ಫು, ಫೂ, ಫೊ, ಫೋ
They should carry similar diacritics as ಕು, ಕೂ, ಕೊ, ಕೋ. The diacritics u+0cc1, u+0cc2, u+0cca, u+0ccb must sit to the right to the base letter ಪ and ಫ.

History:
This wrong of using diacritics mark for u+0cc1, u+0cc2, u+0cca, u+0ccb used with letter ವ was cultivated during the printing error. Since it was only letter to use the below form, the printing presses felt the below form u+0cc1, u+0cc2, u+0cca, u+0ccb "sort" was under-utilized. So, they used the below form u+0cc1, u+0cc2, u+0cca, u+0ccb "sort" with letters ಪ, ಫ, ಹ. Usage of below form diacritics is minimized with letter ಹ in recent past. But It is still continuing with ಪ and ಫ.

References:
Authentic Kannada texts uses side form of diacritics of u+0cc1, u+0cc2, u+0cca, u+0ccb with ಪ and ಫ, similar to that of ಕ.
Only ವ uses the below form to avoid confusion with ಮ.
Extracts from "Kannada dictionary" from "Kannada Sahitya Parishat". This is the most comprehensive dictionary in Kannada language.

Also, please refer to the Nudi fonts, these fonts are very good references. They are reviewed by Kannada scholars. Also, for offline Kannada documents, these Nudi fonts are the most used.
Comparision.

Nudi fonts direct link
http://www.kagapa.in/kannada/sites/default/files/downloads/Nudi-5.0/Fonts.zip
These fonts were original in ASCII format used for past 25 years. They have unicode fonts as well. NudiUni01k and NudiUni02k are good references.

Title

Kannada letter design issue ಪು, ಪೂ, ಪೊ, ಪೋ, ಫು, ಫೂ, ಫೊ, ಫೋ in both "Noto Sans Kannada" and "Noto Serif Kannada"

Font

'NotoSansKannada-Regular.ttf'.
'NotoSerifKannada-Regular.ttf'.

Where the font came from, and when

https://www.google.com/get/noto/

Font Version

1.03uh

OS name and version

Mac OS high Sierra

Application name and version

All

Issue

Design defect and suggestion

Character data

ಪು, ಪೂ, ಪೊ, ಪೋ and ಫು, ಫೂ, ಫೊ, ಫೋ

[MayuraVerma]

@JelleBosmaMT @kalapi Please review this.

[JelleBosmaMT]

The first thing that came to my mind: If I am going to remove the low u and uu variants for ಪ and ಫ then I will get bug reports that they are missing.
There seems to be a government standard in progress, setting out requirements for glyph support. For now I am thinking to wait and see if that has something to say about this issue: one way or the other.

[MayuraVerma]

Do you know which government? India or Karnataka?
Nudi fonts is what Karnataka government using for past 25 years.
http://www.karnataka.gov.in/Pages/downloads.aspx
Click on "ಡೌನ್ಡೋಡ್"->"ನುಡಿ"
All experts agree for this form.

[JelleBosmaMT]

On 24-Oct-17, 08:26, "ಮಯೂರ ವರ್ಮ" <notifications@github.com> wrote: Do you know which government? India or Karnataka? India

[MayuraVerma]

Thanks.
Do you know which department?

[MayuraVerma]

The reason I ask. If it is Indian govt. written the document. They write for all Indian languages. Typing they do not consult the language experts. If you know any information about who and which dept., we can let our language experts to contact them and provide necessary inputs.
If you have any information, please share.

[JelleBosmaMT]

I don’t really know much more than this : http://www.bis.gov.in. At the moment I am supposed to expand the existing designs and worry about the standard, once I receive more clarification from people who are looking into it. It would be easier if we can do with fewer special u and uu forms, most of all for ಫ, where the descender and the tip of the vowel get in each other’s way. However I don’t think printers in the old days would have decided to use these forms if they felt they had the choice not to. It was definitely more expensive to have these additional characters made. Or, if they were not using ligatures made as separate characters, having to typeset with separate vowel signs sorts that fit below the base consonant. That is more work than simply using the standard forms, that do not overlap and there is less risk of breaking the letter. There must be another story that explains. I wonder about the relation with Telugu that does the same. And there too, it is really needed to distinguish Vu from Ma but not so much for Pa and Pha.

[MayuraVerma]

Kannada and Telugu script were same 1100-1400CE until and was called HalagannaDa (Old Kannada or Kadama script). Telugu typeface has changed little from HalagannaDa. But modern Kannada typeface has evolved more from HalagannaDa, significantly.
But both script are same with different typeface.
https://en.wikipedia.org/wiki/Telugu-Kannada_alphabet/
If you are creative font designer, you can create both Kannada and Telugu typeface with one design and two hat forms and convince both the users!

I will do further more due-diligence on pu and phu design and provide more information as available.

Do you what is the reason for this glyph standard,
to help font designer?
Universal Shaping Engine?
does latin has this standard?

Thank you.

[MayuraVerma]

@sridatta1 Please review this issue.

[sridatta1]

I don't recommend the changes and recommend the existing glyphs.
<< This wrong of using diacritics mark for u+0cc1, u+0cc2, u+0cca, u+0ccb used with letter ವ was cultivated during the printing error.
These isn't any evidence for this claim.
These special forms are used since ಪು may be confusable with ಮ (the dot may be visible) and in Telugu dot doesn't exist though the headstroke is non-ligating.
If the changes are made then there would be bug reports that they are incorrect.
The current forms are ones which are commonly used and many Indic scripts have special forms for CV combinations (they are even greater in Tigalari, Tamil, Sharada, Traditional Malayalam) etc. This is not unnatural.

[Rameshkannada]

Sir please give me some extra kannada fonts

[nizarsq]

Look like there is no definite answer about the issue reported. Apparently the current design is the recommended. @marekjez86 should we close this bug and consider it WAI or get expert opinion?

[sridatta1]

After examining the sources based on Inscriptions, printed books etc.
In Kannada script, the forms with current design are most commonly used and I recommend them to be used as the default forms in the Kannada. Most ( or All other) Kannada Unicode fonts have the same design.
The lower/ special form of vowel sign in PA, PHA is not a recent introduction in Telugu-Kannada scripts due to printing error.
They are found in 16th century Telugu -Kannada script inscriptions as well e.g. here పు పూ

The CV combinations with regular right form of vowel signs are also found in Kannada inscriptions and some printed books etc, thus are not incorrect.
Thus my suggestion for the closure of issue is to add the requested forms with right side vowel signs with PA/PHA as stylistic sets and retain the current design as default forms.