silgraphite-fonts — Font-related discussion for Graphite

Re: [Silgraphite-fonts] Tamil font using Graphite

[Shriramana S. <sa...@gm...>]

Yes I am/do. However I will test more and get back later.

Sent from my Android phone
On Jul 13, 2012 2:44 AM, "Sharon Correll" <sha...@si...> wrote:

> Hi, are you still needing help on this?
>
> On 6/28/2012 2:20 AM, Shriramana Sharma wrote:
>
>> On Wed, Jun 27, 2012 at 11:31 AM, Shriramana Sharma <sa...@gm...>
>> wrote:
>>
>>> I'm already upgrading the font with some features. I'll send you an
>>> updated version presently
>>>
>> Hello people. So I've already added all features I can think of for
>> Tamil and replaced some not-really-necessary glyph substitutions with
>> diacritic positioning which looks OK (although not totally identical
>> with the original).
>>
>> However I'm having trouble with reordering numerical diacritics 1 2 3
>> 4 (as superscript or subscript) which are used in writing non-Tamil
>> Indic languages using the Tamil script. They need to be positioned
>> before vowel signs or parts thereof which stand to the right of the
>> consonant. However, this is not working correctly. Please can anyone
>> help me fix this? There is also some inexplicable inordinate spacing
>> between the consonant and the numerical diacritic in the case of super
>> 1 2 3. That needs to be fixed too.
>>
>> Please find the latest material at
>> https://sites.google.com/site/**jamadagni/files/temp/Krishna-**
>> Tamil-2.5.1shriramana2.1-**testing-20120628-1250.zip<https://sites.google.com/site/jamadagni/files/temp/Krishna-Tamil-2.5.1shriramana2.1-testing-20120628-1250.zip>
>>
>> @Martin/Sharon: I'll get back to you on the XeTeX matter later.
>>
>>
>
>

Re: [Silgraphite-fonts] Tamil font using Graphite

[Sharon C. <sha...@si...>]

Hi, are you still needing help on this?

On 6/28/2012 2:20 AM, Shriramana Sharma wrote:
> On Wed, Jun 27, 2012 at 11:31 AM, Shriramana Sharma <sa...@gm...> wrote:
>> I'm already upgrading the font with some features. I'll send you an
>> updated version presently
> Hello people. So I've already added all features I can think of for
> Tamil and replaced some not-really-necessary glyph substitutions with
> diacritic positioning which looks OK (although not totally identical
> with the original).
>
> However I'm having trouble with reordering numerical diacritics 1 2 3
> 4 (as superscript or subscript) which are used in writing non-Tamil
> Indic languages using the Tamil script. They need to be positioned
> before vowel signs or parts thereof which stand to the right of the
> consonant. However, this is not working correctly. Please can anyone
> help me fix this? There is also some inexplicable inordinate spacing
> between the consonant and the numerical diacritic in the case of super
> 1 2 3. That needs to be fixed too.
>
> Please find the latest material at
> https://sites.google.com/site/jamadagni/files/temp/Krishna-Tamil-2.5.1shriramana2.1-testing-20120628-1250.zip
>
> @Martin/Sharon: I'll get back to you on the XeTeX matter later.
>

Re: [Silgraphite-fonts] Forcing LibO to use Gr only

[Bob H. <Bob...@si...>]

On 2012-07-07 at 19:57 Shriramana Sharma wrote:

> What is LibO's default behaviour re Gr when a font has both OT and Gr 
> tables? 

I was suggesting a way you could find out :-)

> Surely the "string" function could be used here to make this shorter?

It'd be nice if that worked, but it doesn't, afaict.  The string() 
function is used to put UI strings in the font's name table, not to 
construct glyph sequences.

Bob

Re: [Silgraphite-fonts] Forcing LibO to use Gr only

[Shriramana S. <sa...@gm...>]

On Sat, Jul 7, 2012 at 11:48 PM, Bob Hallissy <Bob...@si...> wrote:
> To avoid having to remove the OT tables, I recently put this GDL code in
> a non-Roman font I'm working on:

Hello and thanks for your response. I suppose any self-devised easter
egg could be put in this way but the question was how to specifically
telling LibO to use Gr and Gr only.

What is LibO's default behaviour re Gr when a font has both OT and Gr tables?

>      // The following is a debugging tool so I can find out whether the
>      // app is rendering with Graphite or not.
>      // The string "ThisIsNotGraphite" will be rendered as "ThisIsGraphite":
>
>      codepoint("N") codepoint("o") codepoint("t") > _ _ _ /
>          codepoint("T") codepoint("h") codepoint("i") codepoint("s")
> codepoint("I") codepoint("s")
>          _ _ _
>          codepoint("G") codepoint("r") codepoint("a") codepoint("p")
> codepoint("h") codepoint ("i")  codepoint ("t") codepoint ("e") ;

Surely the "string" function could be used here to make this shorter?

string("Not") > _ _ _ / string ("ThisIs") _ _ _ string ( "Graphite" )?

-- 
Shriramana Sharma

Re: [Silgraphite-fonts] Forcing LibO to use Gr only

[Bob H. <Bob...@si...>]

On 2012-07-06 at 22:20 Shriramana Sharma wrote:
> I'm removing them
> before compiling with GDL so that I can be sure what smart rendering I
> see is a result of my GDL and not of the existing OT tables.

To avoid having to remove the OT tables, I recently put this GDL code in 
a non-Roman font I'm working on:

     // The following is a debugging tool so I can find out whether the
     // app is rendering with Graphite or not.
     // The string "ThisIsNotGraphite" will be rendered as "ThisIsGraphite":

     codepoint("N") codepoint("o") codepoint("t") > _ _ _ /
         codepoint("T") codepoint("h") codepoint("i") codepoint("s") 
codepoint("I") codepoint("s")
         _ _ _
         codepoint("G") codepoint("r") codepoint("a") codepoint("p") 
codepoint("h") codepoint ("i")  codepoint ("t") codepoint ("e") ;

This assumes that you have basic ASCII chars in your font.

Bob

[Silgraphite-fonts] Forcing LibO to use Gr only

[Shriramana S. <sa...@gm...>]

Hello list.

In debugging my GDLs, I'm mostly using LibO 3.5. Right now if a font
already has OT tables (JSTF, GDEF, GPOS and GSUB) I'm removing them
before compiling with GDL so that I can be sure what smart rendering I
see is a result of my GDL and not of the existing OT tables. Is there
a way to tell LibO to use Gr only for rendering so I can avoid
removing the OT tables?

-- 
Shriramana Sharma

Re: [Silgraphite-fonts] Krishna Tamil 2.0 based on Lohit Tamil 2.5.1

[Sharon C. <sha...@si...>]

On 6/30/2012 2:19 AM, Shriramana Sharma wrote:
> Dear friends,
>
> I'm happy to present a stable and feature-rich version of Krishna
> Tamil, a purely Graphite-based Tamil font based on Lohit Tamil 2.5.1.
>
> This font provides support for:
>
> 1) classical Tamil orthography
> 2) the two glyphic variants each of pulli (0BCD) and aydam (0B83)
> 2) old Tamil grammatical sequences
>
> It is available from
> https://sites.google.com/site/jamadagni/files/temp/krishna-tamil-2.5.1shriramana2.0-release.zip.

I've updated the link on the Graphite site to use this. If you change 
it again, let me know.

> A better URL would be desirable (note the "temp" in the URL above).
> Lemme see what can be done about that.

Re: [Silgraphite-fonts] Krishna Tamil 2.0 based on Lohit Tamil 2.5.1

[Sharon C. <sha...@si...>]

No. You can suppress it using the -w3521 option.

On 6/30/2012 5:05 AM, Shriramana Sharma wrote:
> BTW while compiling I get a message:
>
> krishna-tamil.gdl(182) : warning(3521): Vertical overlap between
> glyphs in items 1 and 2; attachment may be needed
>
> Should I be concerned about this?
>

Re: [Silgraphite-fonts] AttributeOverride default changed?

[Sharon C. <sha...@si...>]

On 6/29/2012 11:30 PM, Shriramana Sharma wrote:
> On Fri, Jun 29, 2012 at 7:34 PM, Sharon Correll <sha...@si...> wrote:
>> You are correct,  the default value is true. I've made a fix in the GDL
>> manual.
> LOL, the tutorial solutions on p 6 actually already mention this: "The
> following is true by default", but it was only the manual that read
> wrong it seems.

That's probably because I updated the tutorial recently and 
double-checked it. :-)

[Silgraphite-fonts] Fwd: GrCompiler 4.2

[Shriramana S. <sa...@gm...>]

In case there is someone here who's not on silgraphite-devel...

Shriramana.

---------- Forwarded message ----------
From: Shriramana Sharma <sa...@gm...>
Date: Sun, Jul 1, 2012 at 10:20 AM
Subject: GrCompiler 4.2
To: SILGraphite development <sil...@li...>


Looks like the GrCompiler 4.2 has not been announced yet -

http://scripts.sil.org/GraphiteCompilerDownload

and what's even better is that Debian packages are available too:

http://packages.debian.org/sid/grcompiler
http://packages.ubuntu.com/quantal/grcompiler

Right now I'm backporting it to my Kubuntu Precise system without any troubles.

:-) Nice work people!

--
Shriramana Sharma

Re: [Silgraphite-fonts] Krishna Tamil 2.0 based on Lohit Tamil 2.5.1

[Shriramana S. <sa...@gm...>]

BTW while compiling I get a message:

krishna-tamil.gdl(182) : warning(3521): Vertical overlap between
glyphs in items 1 and 2; attachment may be needed

Should I be concerned about this?

-- 
Shriramana Sharma

Re: [Silgraphite-fonts] Testing for items that *don't* belong to a class

[Martin H. <mar...@si...>]

Dear Shriramana,

> Hi -- I'm just wondering if you noticed this (because a couple of
> mails crossed in the cyberspace in either direction!)
> 
> Please reply when you can. Thanks.

Well I was leaving it as an exercise for the student ;) What do you *think* the answer would be?

> > Another thing: If I'm actually testing for the *preceding* character, would:
> >
> > gJ > gJ / clsVowels _ ^ ;
> > gJ > gI ;
> >
> > work?

You don't need the final ^, it's the default position for the cursor after a rule anyway.

Yours,
Martin

Re: [Silgraphite-fonts] Testing for items that *don't* belong to a class

[Shriramana S. <sa...@gm...>]

Hi -- I'm just wondering if you noticed this (because a couple of
mails crossed in the cyberspace in either direction!)

Please reply when you can. Thanks.

On Sat, Jun 30, 2012 at 10:22 AM, Shriramana Sharma <sa...@gm...> wrote:
> Another thing: If I'm actually testing for the *preceding* character, would:
>
> gJ > gJ / clsVowels _ ^ ;
> gJ > gI ;
>
> work?

-- 
Shriramana Sharma

[Silgraphite-fonts] Krishna Tamil 2.0 based on Lohit Tamil 2.5.1

[Shriramana S. <sa...@gm...>]

Dear friends,

I'm happy to present a stable and feature-rich version of Krishna
Tamil, a purely Graphite-based Tamil font based on Lohit Tamil 2.5.1.

This font provides support for:

1) classical Tamil orthography
2) the two glyphic variants each of pulli (0BCD) and aydam (0B83)
2) old Tamil grammatical sequences

It is available from
https://sites.google.com/site/jamadagni/files/temp/krishna-tamil-2.5.1shriramana2.0-release.zip.

As always, the fonts are under the OFL, and the GDL is under the
GPLv2+. (The test ODT/PDF is under the CC-BY-SA, FWIW.)

A better URL would be desirable (note the "temp" in the URL above).
Lemme see what can be done about that.

-- 
Shriramana Sharma

Re: [Silgraphite-fonts] Testing for items that *don't* belong to a class

[Martin H. <mar...@si...>]

On Sat, 30 Jun 2012 10:19:53 +0530
Shriramana Sharma <sa...@gm...> wrote:

> On Sat, Jun 30, 2012 at 10:17 AM, Martin Hosken <mar...@si...> wrote:
> > gJ > gJ / _ ^ clsVowels;
> > gJ > gI
> 
> Thanks. I suppose it would be better to use @ (or @1) instead of gJ in
> the first of these two?
> 
> [BTW does @ also automatically create associations like @1, @2 etc?)]

Yes. so:

gJ > @1 / _ ^ clsVowels;
gJ > gI

I think even grcompiler can't fail to associate gI with gJ in that case ;)

Yours,
Martin

Re: [Silgraphite-fonts] Testing for items that *don't* belong to a class

[Shriramana S. <sa...@gm...>]

On Sat, Jun 30, 2012 at 10:17 AM, Martin Hosken <mar...@si...> wrote:
> gJ > gJ / _ ^ clsVowels;
> gJ > gI

Another thing: If I'm actually testing for the *preceding* character, would:

gJ > gJ / clsVowels _ ^ ;
gJ > gI ;

work?

-- 
Shriramana Sharma

Re: [Silgraphite-fonts] Testing for items that *don't* belong to a class

[Shriramana S. <sa...@gm...>]

On Sat, Jun 30, 2012 at 10:17 AM, Martin Hosken <mar...@si...> wrote:
> gJ > gJ / _ ^ clsVowels;
> gJ > gI

Thanks. I suppose it would be better to use @ (or @1) instead of gJ in
the first of these two?

[BTW does @ also automatically create associations like @1, @2 etc?)]

-- 
Shriramana Sharma

Re: [Silgraphite-fonts] Testing for items that *don't* belong to a class

[Martin H. <mar...@si...>]

On Sat, 30 Jun 2012 09:57:32 +0530
Shriramana Sharma <sa...@gm...> wrote:

> It is possible to fire a rule by testing for a nearby item belonging
> to a particular class, such as in:
> 
> gI > gJ / _ clsVowels ;
> 
> If I want to fire a rule when a nearby item *does not* belong to a
> particular class, what is the way to go?
> 
> I suppose I could define a boolean attribute for all members of the
> class, and then test it:
> 
> clsVowels { isVowel = true } ;
> gJ > gI / _ any { ! isVowel } ;

That's one way, although a pretty expensive one since it involves conditions and setting user attributes. You could build your own negative class, (a class of everything no in clsVowels), but that is often too hard to do. An alternative trick I have used is to create a guard rule:

gJ > gJ / _ ^ clsVowels;
gJ > gI

The first rule is longer than the second so will take precedence over it, so if you have a gJ clsVowels, the first rule will fire, advancing the 'cursor' after the gJ so the second rule won't fire. Processing will continue with the glyph after the gJ input in either case.

HTH,
Yours,
Martin

Re: [Silgraphite-fonts] AttributeOverride default changed?

[Shriramana S. <sa...@gm...>]

On Fri, Jun 29, 2012 at 7:34 PM, Sharon Correll <sha...@si...> wrote:
> You are correct,  the default value is true. I've made a fix in the GDL
> manual.

LOL, the tutorial solutions on p 6 actually already mention this: "The
following is true by default", but it was only the manual that read
wrong it seems.

-- 
Shriramana Sharma

[Silgraphite-fonts] Testing for items that *don't* belong to a class

[Shriramana S. <sa...@gm...>]

It is possible to fire a rule by testing for a nearby item belonging
to a particular class, such as in:

gI > gJ / _ clsVowels ;

If I want to fire a rule when a nearby item *does not* belong to a
particular class, what is the way to go?

I suppose I could define a boolean attribute for all members of the
class, and then test it:

clsVowels { isVowel = true } ;
gJ > gI / _ any { ! isVowel } ;

Is this right, wrong, or right but there is a better way?

-- 
Shriramana Sharma

Re: [Silgraphite-fonts] AttributeOverride default changed?

[Sharon C. <sha...@si...>]

You are correct,  the default value is true. I've made a fix in the 
GDL manual.

But it never hurts, in fact it may be wise, to specify 
AttributeOverride explicitly if you are really counting on it to 
behave a certain way.

Yes, I'm planning to look at the problem you were having, but it may 
take a few days.

On 6/29/2012 12:33 AM, Shriramana Sharma wrote:
> Hello. I wonder if the default for AttributeOverride = false has been
> changed to true since the writing of the GDL manual? (P 18)
>
> """The AttributeOverride directive is
> used to determine whether the first or last value is used. If this
> directive is false (default), then
> overriding doesn't happen so the first value will be used. If it is
> true, then the last value is used."""
>
> Basically I have two rules:
>
> 	clsBasesForAboveMarkCentered { aboveTopCenter = point ( advancewidth
> / 2, 625m ) } ;
> 	clsBasesForAboveMarkVisualCentered { aboveTopCenter = point ( 820m, 625m ) } ;
>
> where the latter class is a subset of the former. First I had the
> visual centering line above the generic centering line, but the
> rendering was not as expected. When I changed it to be as above, the
> rendering is correct. I have not manually set AttributeOverride in my
> GDL nor have I #include-d a file that does so. So I'm assuming the
> default has changed.
>
> Frankly I feel the current behaviour is right, AttributeOverride=true
> should be the default, as one feels natural hearing/stating a general
> rule first and a special rule i.e. exception later. I would just like
> to have confirmation so that my program works correctly hereafter.
>
> BTW when you people get some time, I'd also appreciate any guidance as
> to fixing the reordering problem I reported.
>
> Thanks!
>

[Silgraphite-fonts] AttributeOverride default changed?

[Shriramana S. <sa...@gm...>]

Hello. I wonder if the default for AttributeOverride = false has been
changed to true since the writing of the GDL manual? (P 18)

"""The AttributeOverride directive is
used to determine whether the first or last value is used. If this
directive is false (default), then
overriding doesn't happen so the first value will be used. If it is
true, then the last value is used."""

Basically I have two rules:

	clsBasesForAboveMarkCentered { aboveTopCenter = point ( advancewidth
/ 2, 625m ) } ;
	clsBasesForAboveMarkVisualCentered { aboveTopCenter = point ( 820m, 625m ) } ;

where the latter class is a subset of the former. First I had the
visual centering line above the generic centering line, but the
rendering was not as expected. When I changed it to be as above, the
rendering is correct. I have not manually set AttributeOverride in my
GDL nor have I #include-d a file that does so. So I'm assuming the
default has changed.

Frankly I feel the current behaviour is right, AttributeOverride=true
should be the default, as one feels natural hearing/stating a general
rule first and a special rule i.e. exception later. I would just like
to have confirmation so that my program works correctly hereafter.

BTW when you people get some time, I'd also appreciate any guidance as
to fixing the reordering problem I reported.

Thanks!

-- 
Shriramana Sharma

Re: [Silgraphite-fonts] Tamil font using Graphite

[Shriramana S. <sa...@gm...>]

On Wed, Jun 27, 2012 at 11:31 AM, Shriramana Sharma <sa...@gm...> wrote:
> I'm already upgrading the font with some features. I'll send you an
> updated version presently

Hello people. So I've already added all features I can think of for
Tamil and replaced some not-really-necessary glyph substitutions with
diacritic positioning which looks OK (although not totally identical
with the original).

However I'm having trouble with reordering numerical diacritics 1 2 3
4 (as superscript or subscript) which are used in writing non-Tamil
Indic languages using the Tamil script. They need to be positioned
before vowel signs or parts thereof which stand to the right of the
consonant. However, this is not working correctly. Please can anyone
help me fix this? There is also some inexplicable inordinate spacing
between the consonant and the numerical diacritic in the case of super
1 2 3. That needs to be fixed too.

Please find the latest material at
https://sites.google.com/site/jamadagni/files/temp/Krishna-Tamil-2.5.1shriramana2.1-testing-20120628-1250.zip

@Martin/Sharon: I'll get back to you on the XeTeX matter later.

-- 
Shriramana Sharma

Re: [Silgraphite-fonts] Tamil font using Graphite

[Shriramana S. <sa...@gm...>]

On Wed, Jun 27, 2012 at 1:28 AM, Sharon Correll <sha...@si...> wrote:
> Wow, that was fast! (I think my attempt at Tamil took a lot longer. :-) )

Well I have an advantage since I'm a native user. I've also been
contemplating the evolution and orthographic characteristics of South
Indian scripts for quite some time building up to some Unicode
proposals.

> And some pretty concise GDL code!!

Thanks!

> I'd like to put a link to this on the Graphite Fonts page--would you
> consider it to be a "production quality" font, or just at the "experimental"
> (ie, beta) level?

I'm already upgrading the font with some features. I'll send you an
updated version presently with glyph repertoire on par with Lohit
Tamil. (The initial testing font I sent does not have the generic
punctuation etc.)

> I notice your comments about dotted circles. An idea might be to make dotted
> circles a user-selectable feature. In theory, languages beside Tamil might
> be written with Tamil script and want different letter combinations and
> sequences than Tamil itself. Just a thought.

I know -- and I am quite supportive of this. I should modify the
comments to indicate my actual intent, that a good font should not
produce identical visual output for different input codepoints which
would result in confusion, bad text entry/handling, and possibly
security problems too.

I don't throw in a dotted circle for reordrant glyphs which can't be
reordered then sequences then such confusable sequences would occur.
Observe:

Minority orthographies might require vowel signs to be applied to
independent vowels (and not only consonants). However, I do not think
vowel signs would be applied to numbers or punctuation marks! [That's
too far fetched IMO, what do you think?] So a reordrant vowel sign
would not reorder around such a character. In the case of such a
"NonReordChar" that does not permit reordering against a "ReordChar"
that would permit reordering, an input encoded sequence:

NonReordChar + ReordrantVS + ReordChar

and another input encoded sequence:

NonReordChar + ReordChar + ReordrantVS

would be confusable if NO dotted circle is inserted, because the
latter sequence would be reordered to the former. (You might have to
think of that carefully to get the picture.)

So I figure that in such cases, and in cases where Unicode explicitly
tells us (see the various Indic chapters) to use औ and not अ + ौ etc,
dotted circles *should* be displayed.

Dotted circles should only NOT be displayed out of a pedantic
assumption that sequences like VOWEL_LETTER + VOWEL_SIGN are "wrong",
and I am not doing any such assumption.

Please consider the above carefully. I think you will agree with me.

> Thanks for working on this!

A pleasure to work with Graphite and you people! :-)

-- 
Shriramana Sharma

Re: [Silgraphite-fonts] Tamil font using Graphite

[Sharon C. <sha...@si...>]

Wow, that was fast! (I think my attempt at Tamil took a lot longer. 
:-) ) And some pretty concise GDL code!!

I'd like to put a link to this on the Graphite Fonts page--would you 
consider it to be a "production quality" font, or just at the 
"experimental" (ie, beta) level?

I notice your comments about dotted circles. An idea might be to make 
dotted circles a user-selectable feature. In theory, languages beside 
Tamil might be written with Tamil script and want different letter 
combinations and sequences than Tamil itself. Just a thought.

Thanks for working on this!


On 6/25/2012 3:13 PM, Shriramana Sharma wrote:
> To Sharon and everyone else doing great work on Graphite and related,
>
> I present Krishna Tamil, with my compliments. :-)
>
> https://sites.google.com/site/jamadagni/files/temp/Krishna%20Tamil%202.5.1.tar.gz
>
> Licence: GPLv2+ for the GDL, OFL for the font (of course).
>
> All glyphs are from Lohit Tamil (https://fedorahosted.org/lohit/)
> under the OFL. (Some were designed/touched up by me as my contribution
> to that project.)
>
> Remarks on the name:
>
> Lohit means "red" in Sanskrit, and the Lohit fonts were named so since
> "Red" Hat had them made. But graphite (note absence of capitalization)
> is black, and Krishna means black in Sanskrit [and it's also the name
> of my three-year old daughter's favourite Hindu god :-)] so Krishna
> Tamil it is. Of course, the OFL also requires derivative fonts to have
> the name changed, so...
>
> BTW Pravin Satpute of Red Hat who maintains the Lohit fonts is a very
> nice person and would probably not have problems with including Gr
> tables in his official distributed fonts, but I did want to make the
> point that Graphite could independently render Indic well [and also be
> very clearly expressive about it in the GDL unlike the abstruse OT
> tables which I quite dislike...] so I think it would be appropriate to
> have it as a separate family (of which Tamil is hopefully only the
> first member).
>
> Other remarks:
>
> Well Sharon, you got me going on this, and once I closed my ConTeXT
> reading (for now) I started thinking hey how much time could it need
> to write GDL for Tamil, and so here you are. :-) IIRC the GDL itself
> didn't take more than half an hour (or even less probably, I didn't
> time myself). I had a problem with the base font, since just stripping
> the GPOS, GSUB, GDEF tables from Lohit Tamil wasn't enough and I got
> some illegal glyph substitutions which looked like some stray OT
> remnants overriding my Gr programming. So I had to do a fresh empty
> font, set the metrics, copy paste the glyphs from Lohit Tamil. I then
> just adjusted my GDL for the new gids, and compiled.
>
> Note that as I just wanted to get it done, I didn't bother with
> including the glyphs for characters not specific to Tamil (such as
> international punctuation, digits etc), so if we were to actually
> release this font it would be better to include them to be one-to-one
> par on character repertoire with Lohit Tamil.
>
> BTW as an exception to the above I should note that no proper Tamil
> font should include 0B82 TAMIL SIGN ANUSVARA, since the encoding
> itself was a mistake, which I asked the Unicode people to deprecate
> (https://sites.google.com/site/jamadagni/files/utcsubmissions/12018-tamil-anusvara-depr.pdf).
> However it seems they will only annotate it saying "don't use it for
> the virama". [sigh]
>
> Anyhow, I need to go to bed now, (it's 0135 in India) and I'll read
> all your nice reviews in the morning! [hint hint ;-)]
>
> --
> Shriramana Sharma
>
> ------------------------------------------------------------------------------
> Live Security Virtual Conference
> Exclusive live event will cover all the ways today's security and
> threat landscape has changed and how IT managers can respond. Discussions
> will include endpoint security, mobile security and the latest in malware
> threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
> _______________________________________________
> Silgraphite-fonts mailing list
> Sil...@li...
> https://lists.sourceforge.net/lists/listinfo/silgraphite-fonts