Mon 25 Sep 2017 04:16:38 PM UTC, comment #10:
Hello Steve,
actually I am working on Sanskrit, not Hindi. I was not aware that by default the font gets loaded with the Hindi option, so I had not loaded the font with
UprightFeatures={Language=Sanskrit},
BoldFeatures={Language=Sanskrit},
but after doing so, I did get these ligatures, and will now use it some time in a real file and then let you know if I find anything missing.
BTW., for when can a new release be expected?
|
Mon 25 Sep 2017 01:27:36 PM UTC, comment #9:
Hi Peter,
I just noticed that I didn't directly answer your question about "italics".
Of course, the term "italics" is meaningful only in Latin-based scripts, and by convention, in Greek and Cyrillic. In any other script, there is a notion of "oblique".
I haven't implemented oblique for Devanagari. If there's a strong need for it, we could open a bug report on it, which I would postpone for a future release. My opinion is that a bold face is of more importance, for a couple of reasons.
(The only Indic face with an oblique in GNU FreeFont is Bengali. I had no part in implementing that, but have regretted as long as I've worked with it.)
|
Mon 25 Sep 2017 01:17:36 PM UTC, comment #8:
Hi Peter,
dma (द्म) and dbhya (द्भ्य) are supported by GNU FreeFont in Sanskrit text. They are not formed by default for Hindi. Check out the velthuis-conj test file. I think this is correct behavior.
The conjunct of dbha (द्भ) with ya (द्भ्य) is likewise supported, there is a conflict between the velthuis font manual and the TDIL "Devanagari Script Behaviour for Hindi" v. 1.4.10 on this point: velthuis shows no ligature, while the TDIL shows a ligature.
It would help if you can show that such syllables are conventionally written as ligatures in modern Hindi. We need to see images of printed text -- pre-computer-era, preferably.
Thanks!
|
Mon 25 Sep 2017 12:09:50 PM UTC, comment #7:
dma (द्म) , dbha (द्भ), dbhya (द्भ्य), which are rather common, do not work. There are probably some more.
|
Sun 26 Jun 2016 02:56:21 PM UTC, comment #6:
FreeSerif supports Devanagari in normal and bold, as does FreeSans. FreeMono only has normal face Devanagari.
Please amuse yourself with the HTML and XeTeX files I uploaded.
The FreeSerif Devanagari is based on the Velthuis font. But even that doesn't contain all conjuncts.
It's pretty easy to edit the TeX file to look at the other two faces.
Let me know if I can say this bug is fixed.
(file #37581, file #37582, file #37583, file #37584)
|
Sun 26 Jun 2016 12:59:48 PM UTC, comment #5:
It looks good, so far I haven't seen a conjunct consonant that didn't render properly. After some fiddling the bold font also works, but I didn't get italics to work (with xelatex), I assume they're not supported?
|
Sun 26 Jun 2016 08:20:33 AM UTC, comment #4:
Hi Peter,
I'll send you a snapshot of a recent build.
Please let me know what you see.
The last released version of FreeFont is the one dated 20120503. Anything else is not an authorized release. Unfortunately, due to the long wait in making a release, many people have taken it on themselves to distribute bootleg versions.
There are a few known bugs remaining, and a large amount of testing to be done.
I worked hard with several people on Sanskrit especially. Note that by default, the font features support Hindi -- Sanskrit has to be enabled specially. See the usage documentation in the package.
Yes, the 2012 release was meant to run under the old Linux font rendering engine Pango and on Windows XP, and there it had been heavily tested. In later Windows versions and with Harfbuzz, the rendering especially of Indic scripts has changed so much, that very little that worked before still works. This has been one of the biggest efforts in FreeFont, which supports several Indic languages.
|
Sat 25 Jun 2016 01:57:29 AM UTC, comment #3:
thanks for your response. Indeed I recently learned (http://tug.org/pipermail/xetex/2016-June/026671.html) that the released version of freefont does not work with HarfBuzz, which might be the cause for the bad rendering of the devanagari conjuncts, but that the development version doesn't have that problem and has also much improved support for Sanskrit. Yes, I am very happy to test it. In the attachment you can see an example tex file to be compiled with xelatex and its output using the released version of freefont.
(file #37567, file #37568)
|
Sat 25 Jun 2016 12:04:18 AM UTC, comment #2:
Hello Peter,
Could you please provide us with a unicode-encoded document with text that uses these letters, as well as graphical representation of the text, with the glyphs correctly rendered?
Much work has been done with Devanagari since the 2012 release, specifically regarding Sanskrit support. Let us know if you would like to look at development versions.
|
Sun 28 Feb 2016 03:05:17 PM UTC, comment #1:
I had changed my original Heading from "Sanskrit" to "Devanagari", that's how the "as well" in my first sentence is to be understood: They're primarily relevant to Sanskrit, but to other languages as well.
|
Sun 28 Feb 2016 02:17:11 PM UTC, original submission:
Some of them might be wished for in Hindi and other modern South Asian Languages as well. The first one coming to mind are combinations where "r" follows another consonant, such as in "pra" or "tra", "प्र" "त्र", where the "r" should be represented by a diagonal stroke in the lower part of the letter.
|