Re: Changing Kannada font used in website.

[विश्वासो वासुकिजः (Vishvas Vasuki)]

नमस्ते -

Since I generally read online vedic texts in devanAgarI, I don't have a good suggestion. But perhaps someone in the +sanskrit-programmers mailing list does...

On Sat, Sep 15, 2018 at 9:33 PM, Shree Devi Kumar <shree...@gmail.com> wrote:

Namaste Vishvas Ji,

Can you recommend a good kannada font for displaying the transliterated sanskrit text.  I was earlier using Goda but that seems to have some rendering problem.   Noto Serif Kannada does not support the Vedic accents.

I have changed font for vedic texts to Goda - color Teal so that it is easy to verify while testing. There are still some broken circle signs coming.

Please compare

https://sanskritdocuments.org/doc_veda/s-sukta-accent.html?lang=kn

https://sanskritdocuments.org/doc_veda/s-sukta.html?lang=kn

and

https://sanskritdocuments.org/doc_devii/aShTalakShmIstutiH.html?lang=kn

Thanks.

---------- Forwarded message ----------
From: Sanskrit Team member <sans...@cheerful.com>
Date: Thu, Sep 13, 2018 at 9:41 PM
Subject: Fwd: Changing Kannada font used in website.
To: shree...@gmail.com

--
Sent from my Android phone with mail.com Mail. Please excuse my brevity.

On 09/13/2018, 7:30 AM rama prakasha <ramapr...@gmail.com> wrote:

Respected Sir,

As you can see in the screenshot attached on the mail the rendering of kannada font is not good. Luckily Google has now released new fonts in Kannada called Noto Serif Kannada and Noto Sans Kannada fonts. 

https://www.google.com/get/noto/#sans-knda

https://www.google.com/get/noto/#serif-knda

It would be very benefiting for Kannada readers, I hope webmasters will look into this.

Jai Shri Raama,

Raama Prakasha

--

____________________________________________________________
भजन - कीर्तन - आरती @ http://bhajans.ramparivar.com

--

--
Vishvas /विश्वासः

[Shreevatsa R]

+Sridatta A who might know. If I understand correctly, the question is whether there exist Kannada fonts with correct/good rendering of Vedic accents.

--
You received this message because you are subscribed to the Google Groups "sanskrit-programmers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sanskrit-program...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Shree Devi Kumar]

Thanks, that's correct.

Alternatively, if there is  no Kannada font with correct/good rendering of Vedic accents, then is there is way to setup fallback font in css to support this. I tried the following but the results didn't seem right.

*[lang="kn"] {

font-family: 'NotoKannada', 'Goda', serif !important;

font-size: 1.2em;

line-height:2.0;

}

To unsubscribe from this group and stop receiving emails from it, send an email to sanskrit-programmers+unsub...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

[Sridatta A]

As per Jelle Bosma of Monotype, the designer of Noto Sans Kannada it should work in the latest Phase III fonts. See here https://github.com/googlei18n/noto-fonts/issues/1193#issuecomment-391728115 I haven't checked it yet. 

You can get the fonts here https://github.com/googlei18n/noto-fonts/tree/master/phaseIII_only . You may test in hb-view, if the positioning etc. isn't correct you can file issue in the GitHub repository.

However the Vedic signs doesn't render properly in Noto Serif Kannada.

Sridatta

To unsubscribe from this group and stop receiving emails from it, send an email to sanskrit-program...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

[Shree Devi Kumar]

Thank you so much for your prompt response.

Yes,  Noto Sans Kannada (Phase III) indeed seems to work (on Chrome, Winows 10) for Vedic Marks.

For now, I have changed css to use the Serif version for regular documents and Sans version for the ones marked as Vedic.

Appreciate the help.

PS. If any positioning related problems are pointed out by Kannada users, I will file an issue in the Github repo.

Regards,

To unsubscribe from this group and stop receiving emails from it, send an email to sanskrit-programmers+unsub...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

--

____________________________________________________________
भजन - कीर्तन - आरती @ http://bhajans.ramparivar.com

[Bhasha IME]

What's the exact problem with Goda you have encountered. Send me a sample text. I may be able to fix it. ( I am the creator of Goda)

regards

Venkatesh

--
You received this message because you are subscribed to the Google Groups "sanskrit-programmers" group.

To unsubscribe from this group and stop receiving emails from it, send an email to sanskrit-program...@googlegroups.com.

[विश्वासो वासुकिजः (Vishvas Vasuki)]

On Sun, Sep 16, 2018 at 2:13 AM, Bhasha IME <bhas...@gmail.com> wrote:

What's the exact problem with Goda you have encountered. Send me a sample text. I may be able to fix it. ( I am the creator of Goda)

IIRC:  anusvaara following a svara sign is rendered badly. +shreedevi might have handy examples.

Btw, would you publish the source for this font on github with a suitable open source license? I am curious to take a look.

 

regards

Venkatesh

To unsubscribe from this group and stop receiving emails from it, send an email to sanskrit-programmers+unsubscrib...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "sanskrit-programmers" group.

To unsubscribe from this group and stop receiving emails from it, send an email to sanskrit-programmers+unsubscrib...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

[Shree Devi Kumar]

On 09/14/2018, 2:48 AM rama prakasha wrote:

The font outline is getting very thin making the text unreadable. I have marked those portions. I tried to replicate it in libre office but unable to do so. This is happening in browsers though

From: rama prakasha 
Date: Fri 14 Sep, 2018, 4:58 PM
Subject: Re: Changing Kannada font used in website.
To: Sanskrit Team member <sans...@cheerful.com>

I created a sample and hosted here: https://rpkaranth.bitbucket.io/Goda_test/

The CSS contains links to external font resources, so make take some time.

The problem is found both on Google Chrome and Firefox on Ubuntu OS. And even on Android 8.1 Chrome and Firefox browser.

regards

Venkatesh

To unsubscribe from this group and stop receiving emails from it, send an email to sanskrit-program...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "sanskrit-programmers" group.

To unsubscribe from this group and stop receiving emails from it, send an email to sanskrit-program...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

[Bhasha IME]

anusvara following svara is non-unicode. It should be anusvara+svara.

The dotted circles are not created (in fact, non-controllable) by font. They happen from the uniscribe engine (or Universal shaping engine in newer ver of Windows), (harbuzz in Libreoffice) without referring the font itself. You need to correct the text for proper ordering of anusvara+svara

You can already open the ttf file in fontforge to look. It's as good as the source. Will 'git' when I find time.

regards

Venkatesh

To unsubscribe from this group and stop receiving emails from it, send an email to sanskrit-program...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "sanskrit-programmers" group.

To unsubscribe from this group and stop receiving emails from it, send an email to sanskrit-program...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

--

--
Vishvas /विश्वासः

--
You received this message because you are subscribed to the Google Groups "sanskrit-programmers" group.

To unsubscribe from this group and stop receiving emails from it, send an email to sanskrit-program...@googlegroups.com.

[विश्वासो वासुकिजः (Vishvas Vasuki)]

On Sun, Sep 16, 2018 at 8:44 AM, Bhasha IME <bhas...@gmail.com> wrote:

anusvara following svara is non-unicode. It should be anusvara+svara.

So there is a unicode standard confirming that "should be anusvara+svara" expectation? Any citation (just want to save it for future reference)?

Nonetheless, even if it were the case, just from the point of view of shAstra and even typography (where the svara sign is visually placed), it is more natural to type the svara after the maatraa and before the anusvAra - since the svara sign usually clarifies the tone of the vowel rather than the anusvaara (which often is infact a placeholder in kannada texts for a vargapanchama like nakaara). This being the case, it might have been good for a font +‌ rendering engines to support the "anusvara following svara" sequence.

 

The dotted circles are not created (in fact, non-controllable) by font. They happen from the uniscribe engine (or Universal shaping engine in newer ver of Windows), (harbuzz in Libreoffice) without referring the font itself. You need to correct the text for proper ordering of anusvara+svara

Ah alas - so there is nothing to be done about it. +shridevi-ji https://sanskritdocuments.org/doc_veda/s-sukta-accent.html?lang=kn seems to retain a couple of such "anusvara following svara" cases. Generally seems to happen where anusvaara is used in kannaDa as a placeholder in kannada texts for a vargapanchama like nakaara

 

You can already open the ttf file in fontforge to look. It's as good as the source. Will 'git' when I find time.

But there must have been some scripts or framework to generate these ttf files? (I don't know - just guessing.) I am curious what steps you went through while developing this font.

 

regards

Venkatesh

To unsubscribe from this group and stop receiving emails from it, send an email to sanskrit-programmers+unsub...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "sanskrit-programmers" group.

To unsubscribe from this group and stop receiving emails from it, send an email to sanskrit-programmers+unsub...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

--

--
Vishvas /विश्वासः

--
You received this message because you are subscribed to the Google Groups "sanskrit-programmers" group.

To unsubscribe from this group and stop receiving emails from it, send an email to sanskrit-programmers+unsub...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "sanskrit-programmers" group.

To unsubscribe from this group and stop receiving emails from it, send an email to sanskrit-programmers+unsub...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

[Shreevatsa R]

On Sun, Sep 16, 2018 at 5:04 PM विश्वासो वासुकिजः (Vishvas Vasuki) <vishvas...@gmail.com> wrote:

On Sun, Sep 16, 2018 at 8:44 AM, Bhasha IME <bhas...@gmail.com> wrote:

anusvara following svara is non-unicode. It should be anusvara+svara.

So there is a unicode standard confirming that "should be anusvara+svara" expectation? Any citation (just want to save it for future reference)?

The Unicode standard is found at http://www.unicode.org/versions/latest/
Many major Indic scripts (including Devanagari and Kannada) covered in Chapter 12, South and Central Asia-I (https://www.unicode.org/versions/Unicode11.0.0/ch12.pdf)

In the current version (Unicode 11), see p. 456 R10:

Also p. 467:

These are both in the "Devanagari" section, but for other scripts the standard generally refers to the Devanagari section and only notes the differences.

 

Nonetheless, even if it were the case, just from the point of view of shAstra and even typography (where the svara sign is visually placed), it is more natural to type the svara after the maatraa and before the anusvAra - since the svara sign usually clarifies the tone of the vowel rather than the anusvaara (which often is infact a placeholder in kannada texts for a vargapanchama like nakaara).

In general Unicode does not follow visual order: e.g. कि is encoded as U+0915 DEVANAGARI LETTER KA followed by U093F DEVANAGARI VOWEL SIGN I unlike earlier systems where the "i" vowel sign would be encoded first (as it occurs first visually). 

The situation is arguably unsatisfactory in the case of Kannada where the anusvāra doesn't mean much... there's some formal process for proposing changes to Unicode after verifying that it makes sense (Sridatta can tell you the details; he's an expert and has done this many times AFAIK), but at the moment fonts or layout/rendering engines are not expected to support that. 

[विश्वासो वासुकिजः (Vishvas Vasuki)]

On Sun, Sep 16, 2018 at 7:19 PM, Shreevatsa R <shree...@gmail.com> wrote:

On Sun, Sep 16, 2018 at 5:04 PM विश्वासो वासुकिजः (Vishvas Vasuki) <vishvas...@gmail.com> wrote:

On Sun, Sep 16, 2018 at 8:44 AM, Bhasha IME <bhas...@gmail.com> wrote:

anusvara following svara is non-unicode. It should be anusvara+svara.

So there is a unicode standard confirming that "should be anusvara+svara" expectation? Any citation (just want to save it for future reference)?

The Unicode standard is found at http://www.unicode.org/versions/latest/
Many major Indic scripts (including Devanagari and Kannada) covered in Chapter 12, South and Central Asia-I (https://www.unicode.org/versions/Unicode11.0.0/ch12.pdf)

In the current version (Unicode 11), see p. 456 R10:

Also p. 467:

But the above prescriptions doesn't apply to the current case, as far as I see - atleast not unless one somehow (strangely) extrapolates the chandra-below rule to anusvAra. Since the chandra-below rule, as well as the chandrabindu and the AA verb-sign directly qualify the vowel presented before them, it is fine for any arbitrary order to be imposed/ prescribed for them. Not so with the anusvaara - given that it is a separate sound that follows the vowel being modified by the svara sign. (And the nukTa which qualifies the vyanjana sign appearing before the virAma/ vowel signs sort of makes sense as well - since viraama "modifies" - ie denotes absence of- the succeeding vowel.)

Thanks for looking, though! Please let me know if you find something directly prescriptive about the svara-anusvAra order.

+ shriramaNa in case he might favor us with his expert comments. (context - https://groups.google.com/d/msg/sanskrit-programmers/sp-FwGFr2PQ/ktmMjtwnCQAJ )

 

These are both in the "Devanagari" section, but for other scripts the standard generally refers to the Devanagari section and only notes the differences.

 

Nonetheless, even if it were the case, just from the point of view of shAstra and even typography (where the svara sign is visually placed), it is more natural to type the svara after the maatraa and before the anusvAra - since the svara sign usually clarifies the tone of the vowel rather than the anusvaara (which often is infact a placeholder in kannada texts for a vargapanchama like nakaara).

In general Unicode does not follow visual order: e.g. कि is encoded as U+0915 DEVANAGARI LETTER KA followed by U093F DEVANAGARI VOWEL SIGN I unlike earlier systems where the "i" vowel sign would be encoded first (as it occurs first visually). 

The situation is arguably unsatisfactory in the case of Kannada where the anusvāra doesn't mean much... there's some formal process for proposing changes to Unicode after verifying that it makes sense (Sridatta can tell you the details; he's an expert and has done this many times AFAIK), but at the moment fonts or layout/rendering engines are not expected to support that. 

@sridatta - what do you think? can we make unicode prescribe the svara-before anusvaara order? and then make the rendering engines properly deal with such an order?

 

--
You received this message because you are subscribed to the Google Groups "sanskrit-programmers" group.

To unsubscribe from this group and stop receiving emails from it, send an email to sanskrit-programmers+unsubscrib...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

[Shreevatsa R]

On Sun, Sep 16, 2018 at 9:37 PM विश्वासो वासुकिजः (Vishvas Vasuki) <vishvas...@gmail.com> wrote:

On Sun, Sep 16, 2018 at 7:19 PM, Shreevatsa R <shree...@gmail.com> wrote:

On Sun, Sep 16, 2018 at 5:04 PM विश्वासो वासुकिजः (Vishvas Vasuki) <vishvas...@gmail.com> wrote:

On Sun, Sep 16, 2018 at 8:44 AM, Bhasha IME <bhas...@gmail.com> wrote:

anusvara following svara is non-unicode. It should be anusvara+svara.

So there is a unicode standard confirming that "should be anusvara+svara" expectation? Any citation (just want to save it for future reference)?

The Unicode standard is found at http://www.unicode.org/versions/latest/
Many major Indic scripts (including Devanagari and Kannada) covered in Chapter 12, South and Central Asia-I (https://www.unicode.org/versions/Unicode11.0.0/ch12.pdf)

In the current version (Unicode 11), see p. 456 R10:

[snip]

But the above prescriptions doesn't apply to the current case, as far as I see

Sorry not sure why I included the second screenshot. The relevant thing is in the first one:

R10: "Other modifying marks, in particular bindus and svaras, apply to the orthographic syllable as a whole and should follow [...] all other characters that constitute the syllable. The bindus should follow any vowel signs, and the svaras should come last."

-- this includes the anusvara, which is considered a "bindu" (see "Combining marks" on p. 462).

(BTW, quite separately, try this: copy the text "ಮಾಂಗಲ್ಯ" into a text editor, put your cursor just before it, and hit the right-arrow key, it will go after the anusvara.)

[Bhasha IME]

Since when did Unicode standard/Implementation follow shastra?

Does shastra have आ and ा (maatra)

Does shastra allow क + ि) => कि

and so on.

Unicode code standard for Indian script is unscientific. It was proposed with yesteryear's typewriter layout in mind. I don't think anyone can convince MS or even people behind harfbuzz to change their implementations.

I created Graphite (open ended technology) fonts as an alternate for Opentype. This would allow, for eg. combining Kannada text with sama-vedic svaras; alas, the harfbuz creator decided to introduce the very same restriction that plague OT into Libreoffice, sapping out the very essence of Graphite as being open ended. I was rebuked for pointing this out.They tell you what you can do and what you cannot.

Have filed many bug-issues with Libreoffice over years w.r.t  Indic-scripts - not one corrected so far.

Have learnt to get on with what's is possible.

regards

Venkatesh

To unsubscribe from this group and stop receiving emails from it, send an email to sanskrit-program...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "sanskrit-programmers" group.

To unsubscribe from this group and stop receiving emails from it, send an email to sanskrit-program...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

--

--
Vishvas /विश्वासः

--
You received this message because you are subscribed to the Google Groups "sanskrit-programmers" group.

To unsubscribe from this group and stop receiving emails from it, send an email to sanskrit-program...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "sanskrit-programmers" group.

To unsubscribe from this group and stop receiving emails from it, send an email to sanskrit-program...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

--

--
Vishvas /विश्वासः

--
You received this message because you are subscribed to the Google Groups "sanskrit-programmers" group.

To unsubscribe from this group and stop receiving emails from it, send an email to sanskrit-program...@googlegroups.com.

[विश्वासो वासुकिजः (Vishvas Vasuki)]

Great - just what I sought - thanks!

 

--
You received this message because you are subscribed to the Google Groups "sanskrit-programmers" group.

To unsubscribe from this group and stop receiving emails from it, send an email to sanskrit-programmers+unsub...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

[विश्वासो वासुकिजः (Vishvas Vasuki)]

On Mon, Sep 17, 2018 at 2:38 AM, Bhasha IME <bhas...@gmail.com> wrote:

I created Graphite (open ended technology) fonts as an alternate for Opentype. This would allow, for eg. combining Kannada text with sama-vedic svaras; alas, the harfbuz creator decided to introduce the very same restriction that plague OT into Libreoffice, sapping out the very essence of Graphite as being open ended.

Ah! I see - https://scripts.sil.org/cms/scripts/page.php?site_id=projects&item_id=graphite_aboutOT

Here is my current overview of how fonts are rendered (which I am recording in https://sanskrit-coders.github.io/site/pages/content/fonts.html ) - how do Opentype and Graphite fit in?

-------------

Several things are required for digitally encoded script to properly appear on the screen.

• First, the representation should be something one’s font files and rendering engines understand. Currently, Unicode is the well-entrenched standard.
• Then, the font should have information about how to render particular code points.
• Then rendering engines should be able to put the glyphs read from the font dictionary together properly.
  • Major complex-text rendering software:
    • uniscribe engine or Universal shaping engine in newer version of Windows
    • harfbuzz in Libreoffice, Chrome, Firefox etc..
    • CoreText (on macOS)
  • General rendering software: Cairo or Xft.

-------------

 

I was rebuked for pointing this out.They tell you what you can do and what you cannot.

Do you have a link to that conversation online?

 

regards

Venkatesh

To unsubscribe from this group and stop receiving emails from it, send an email to sanskrit-programmers+unsubscrib...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "sanskrit-programmers" group.

To unsubscribe from this group and stop receiving emails from it, send an email to sanskrit-programmers+unsubscrib...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

--

--
Vishvas /विश्वासः

--
You received this message because you are subscribed to the Google Groups "sanskrit-programmers" group.

To unsubscribe from this group and stop receiving emails from it, send an email to sanskrit-programmers+unsubscrib...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "sanskrit-programmers" group.

To unsubscribe from this group and stop receiving emails from it, send an email to sanskrit-programmers+unsubscrib...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

--

--
Vishvas /विश्वासः

--
You received this message because you are subscribed to the Google Groups "sanskrit-programmers" group.

To unsubscribe from this group and stop receiving emails from it, send an email to sanskrit-programmers+unsubscrib...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "sanskrit-programmers" group.

To unsubscribe from this group and stop receiving emails from it, send an email to sanskrit-programmers+unsubscrib...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

[Madhva Raj]

NAMASTHEY,

                            I think this might be right  thread to clear my doubts.

In my website i want Sanskrit slokas (Site on Sandhyaavandanam) . Following are my doubts

1.Do i need to use UTF fonts. What are they UTF. How can i download them.

2. Once a website opens in Chrome Google Asks for translate then sanskrit slokas might not be translated in Right manner. How to avoid Translation.

3. If i want Transliteration (in order to help user to write slokas in his own handwriting or printout form) how to achieve this.

4. What is the website or software to Type sanskrit letters (which are UTF ready).

5. In future i want to apply for Google Ads . What are their Rules regarding language or font.

6. I am  a .net Developer . I want to Develop Android App. How can i achieve it . Without Knowing Android Development or using .net

Please suggest me a way to proceed.(In future also Regarding Sandhyaavandnam Information) 

Thanks and Regards

Naga Mahima

To unsubscribe from this group and stop receiving emails from it, send an email to sanskrit-programmers+unsub...@googlegroups.com.

[विश्वासो वासुकिजः (Vishvas Vasuki)]

Your non-font questions merit separate threads - so won't take them up here.
 

By UTF font, I presume you mean "Unicode font" or a font which can present UTF-8 codes as characters. If so, just google "devanagari unicode font" and you will get plenty of choices.

"What is the website or software to Type sanskrit letters (which are UTF ready)." - > Nearly every modern software lets you type unicode devanagari - such as shrI venkatesh's bhasha IME  (google it).

[Shreevatsa R]

On Mon, Sep 17, 2018 at 8:25 AM विश्वासो वासुकिजः (Vishvas Vasuki) <vishvas...@gmail.com> wrote:

Here is my current overview of how fonts are rendered (which I am recording in https://sanskrit-coders.github.io/site/pages/content/fonts.html ) - how do Opentype and Graphite fit in?

-------------

Several things are required for digitally encoded script to properly appear on the screen.

• First, the representation should be something one’s font files and rendering engines understand. Currently, Unicode is the well-entrenched standard.
• Then, the font should have information about how to render particular code points.
• Then rendering engines should be able to put the glyphs read from the font dictionary together properly.

(As no one else has answered yet...)

OpenType and Graphite are font formats. The text rendering engine picks up data from the font (stored in that format), and uses that data (along with its "understanding" of the text) to produce the final (not-quite) "image".

In more detail:

At minimum, a font is a collection of shapes (what each letter should look like -- these days this is typically drawing instructions for the outline of the shape, also possibly programs ("hinting") for specific resolutions).

But that's not all, a good font should also contain kerning: e.g. when you have "V" followed by "A", the letters should be brought closer together than when you have "V" followed by "T". This is subjective and up to the font designer.

Moreover: a font can also contain substitution rules (like GSUB in OpenType): replace "f" followed by "i" with a particular shape, replace "ह" followed by vowel-sign-vocalic-R with another shape.

A font can also contain more and more instructions, like for positioning, etc.: it is "data" but it is also "code".

Some font formats allow more kinds of instructions / data than others.

Both Unicode and the font [format] want to specify how to position characters, in what context something should be replaced with something else etc., so there is some tension. The rendering engine might implement one set of instructions better than the other.

[विश्वासो वासुकिजः (Vishvas Vasuki)]

Thanks so much! I've updated my understanding and notes.

Now what about TTF? I gather that it used to be an "earlier generation font format". Is this correct? How come popular sanskrit fonts come in *.ttf files?

--
You received this message because you are subscribed to the Google Groups "sanskrit-programmers" group.

To unsubscribe from this group and stop receiving emails from it, send an email to sanskrit-programmers+unsub...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

[विश्वासो वासुकिजः (Vishvas Vasuki)]

Incidentally, a couple of rendering demos inspired by shrI ajit:

• svaras-siddhaanta.html
• svaras-noto.html

Specially note "visarga after svara" and svaras on gm. The use of "private-use" code points in siddhaanta seems a bit caution-inspiring.

To unsubscribe from this group and stop receiving emails from it, send an email to sanskrit-programmers+unsubscrib...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

--

--
Vishvas /विश्वासः

[Shreevatsa R]

OpenType is an extension of TrueType, which is still around. And OpenType fonts can (and often do) have extension .ttf, which can add to the confusion.

To unsubscribe from this group and stop receiving emails from it, send an email to sanskrit-program...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

--

--
Vishvas /विश्वासः

--
You received this message because you are subscribed to the Google Groups "sanskrit-programmers" group.

To unsubscribe from this group and stop receiving emails from it, send an email to sanskrit-program...@googlegroups.com.

[vatsa vatsi]

You can use google input tools.

[विश्वासो वासुकिजः (Vishvas Vasuki)]

graphite format fonts can be in ttf files too?

[Shreevatsa R]

Yes. In fact if I understand correctly, all graphite-format fonts will usually be .ttf files, as the Graphite format (like OpenType) extends Truetype and uses some custom tables for the "program".

This is my understanding of Graphite and Opentype (partly repeating what I said earlier):

- For simple scripts like English (Latin script), the "program" is simple and the "data" is mostly the shape of each letter.

- If you want to get fancier, you can have "programs" (custom instructions) for things like "when 'i' follows 'f', replace with the 'fi' shape". But given that these instructions are so simple, they too can be encoded as "data" in a table, and the rendering engine be expected to know how to use this data.

- With more complex scripts, the instructions get more and more complex. E.g. consider something in Devanagari like (making this up) अर्द्धि -- here, given the sequence  ‎[A] [RA] [VIRAMA] [DA] [VIRAMA] [DHA] [VOWEL SIGN I], it is needed to do various things, like attach the "dha" below the "da", make the vowel sign (which appears last in the unicode order and first in the visual order) have a top curve long enough to reach the top part of the "da", and finally put the repha after the whole thing.

Now, for scripts that require complex instructions like this, the Unicode+OpenType approach (slowly becoming predominant) is for such instructions to be standardized (rules like "whenever [RA] [VIRAMA] occurs at the beginning of a consonant sequence, it must be converted to a repha placed at the end...") in common places, then implemented in the various rendering engines which are supposed to follow these standards; then the fonts only need to describe the "data" (perhaps the shapes of "da", "dha", the repha, etc), and the rendering engine will know how to assemble them together. 
For SIL (makers of Graphite), given their missionary goals, they often want to work with tribal groups and the like ("unreached peoples"), and don't want to wait for obscure scripts to be slowly (and often imperfectly) standardized, then (at least initially) poorly implemented, etc. The Graphite approach is to put such "programs" in the font itself.

From a long-term perspective of wanting more fonts in the world, the Unicode+OpenType approach makes sense, as there is less "data" requires from the font designer -- we can't expect *all* designers of (say) Kannada fonts to each of them properly implement all the rules of Kannada consonant re-positioning in every font they make, so it makes sense to require the bare minimum from the font designer (and put the burden of implementing the rules in the rendering engine, which only have to do it once for that script, for all fonts).

From a *good* font designer's perspective, though, Graphite (if better supported) would make more sense, as with arbitrary programs they can completely control the appearance, and don't have to care which rendering engine will be used. (In practice the rendering engine may poorly implement Graphite, as mentioned in this thread earlier by Bhasha IME.)

[Sridatta A]

One of issues discussed in this thread on Syllable + Vedic svara + Anusvara/Visarga is recently discussed in this thread on Unicode mailing list.
https://www.unicode.org/mail-arch/unicode-ml/y2019-m12/0046.html

[Shreevatsa R]

Thanks for the reference -- a clear post and glad to see some discussion happening. Let's see what comes out of it!

On Tue, 31 Dec 2019 at 20:26, Sridatta A <sridatta....@gmail.com> wrote:

One of issues discussed in this thread on Syllable + Vedic svara + Anusvara/Visarga is recently discussed in this thread on Unicode mailing list.
https://www.unicode.org/mail-arch/unicode-ml/y2019-m12/0046.html

--
You received this message because you are subscribed to the Google Groups "sanskrit-programmers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sanskrit-program...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/sanskrit-programmers/b8652a9e-fdc6-4cb3-9c68-c2b1262482fc%40googlegroups.com.

[Vishvas Vasuki]

Relevant to this thread - https://groups.google.com/g/sanskrit-programmers/c/cWShtglz_1o

[Daya Aithal]

I have been using Noto Sans Kannada for Rigveda with Vedic Accents. Check this

https://sanskritdocuments.org/kannada/

https://sanskritdocuments.org/doc_veda/RVMKF2-1.html

Could you elaborate the issue you are facing in using the font?

Thanks,

Dayananda Aithal