lack of conjuncts, diacritic stacks and general typography for Javanese

[GoogleCodeExporter]

What steps will reproduce the problem?
1. While all of letters have conjuct forms, they cannot be used with any of the 
combining diacritic mark such as U+A9B8 (suku), U+A9BD (keret), or U+A9BD 
(pengkal). Instead of joining with the conjuct, the diacritic joins the main 
letter and obstruct the conjuct 
2. U+A9B3 (cecak telu), U+A9B6 (wulu), U+A9B7 (wulu melik), U+A9BD (pepet), 
U+A980 (panyangga), U+A981 (cecak), and U+A982 (layar) should form a specific 
stacking pattern when used together. However, noto sans lack any of these 
stack. Diacritic are smashed together in a single space instead.
3. The diacritics are misplaced in many cases, the position slide into the 
following letter instead of the intended letter
4. There’s a space between the letters and U+A9C0 (pangkon)
5. The shape of most characters are incomprehensible, as initial n-stroke is 
reduced into a single line
6. U+A99D (dda) and U+A9A3 (da mahaprana) should be differentiated in shape

What is the expected output? What do you see instead?
(please see attached image)

What version of the product are you using? On what operating system?
android kit-kat, windows 7, graphite and non-graphite applications

Please provide any additional information below.
These feedback are from a Javanese-script user community based in Indonesia. 
Some member of the community expressed how noto is easier to use in android 
compared to other javanese fonts. However, because of the various problems 
stated above, it is far from ideal and are not recommended to be used. For 
reference of proper Javanese shape and typography (based on 19th century 
typesetting) please refer to the "Tuladha Jejeg font" by R.S. Wihananto at 
https://sites.google.com/site/jawaunicode/main-page.

Thank you,
Bayu

Original issue reported on code.google.com by jimeildo...@gmail.com on 11 Jan 2015 at 7:25

Attachments:

• 10933218_393746370789811_1407346826_n.jpg

[GoogleCodeExporter]

It seems some of the issues reported are defects or mismatches in the rendering 
software. The font uses OpenType tables to inform rendering. OpenType is 
different from Graphite, so using Graphite software does not help. Also, 
support for Javanese in OpenType is quite recent; old systems such as Windows 7 
don't support it. A good application to use for testing is the current version 
of Firefox – bugs you see there are very likely bugs in the font.

Detailed comments:

1. The font's GPOS table adjusts the positions of U+A9B8 (suku) and U+A9BF 
(cakra), but not of other characters that can collide with pasangan. The 
adjustment of suku and cakra can be seen in Firefox.

2. The font's GSUB table forms a nice ligature of U+A9BD (pepet) and U+A981 
(cecak). All other combinations are, as you say, smashed together. The ligature 
of pepet and cecak can be seen in Firefox.

3. The font's GPOS table has rules to properly position individual marks above 
or below the base consonant. The positioning can be seen in Firefox. In 
rendering software that doesn't observe the GPOS rules, some above-base marks 
may float to the right because of bug 162.

4. The font's GPOS table does nothing to position pangkon, so it floats to the 
right because of bug 162.

5. I don't know what motivated the Monotype design team to simplify the glyphs 
this much, but I'm curious: does the lack of the initial upstroke really make 
the character incomprehensible, or is it just an unusual design? Or does for 
Javanese unusual imply incomprehensible because people don't see much text in 
Javanese script in the first place?

6. That's bug 161.

Original comment by googled...@lindenbergsoftware.com on 18 Jan 2015 at 7:37

[GoogleCodeExporter]

Original comment by stua...@google.com on 28 Feb 2015 at 12:31

[GoogleCodeExporter]

Original comment by roozbeh@google.com on 2 Apr 2015 at 4:02