Noto Javanese using unusual glyphs

[tjahaja]

Noto Javanese is using glyphs that are uncommon for Javanese people. It turns the letters hardly legible for most people. Some even say it is Balinese, instead of Javanese.

This is a comparison of three Javanese fonts. Tuladha Jejeg is the most used font in web (used by Microsoft), while Aturra is created recently with simpler glyphs. The style used in Tuladha Jejeg & Aturra is found in books, thus easy to recognize and considered standard.

Not all Javanese letters in Noto are problematic. These are what I found to be problematic. Most of them are only missing the initial stroke (marked in black background), while there are some that are switched (marked in red background). I put Tuladha Jejeg (in yellow background) and Aturra (in blue background) as comparisons.

[biyanto]

Thank you, Cahyo for commenting Javanese script. The initial stroke is really important, without it, it just like another version of Balinese. Matur nuwun

[NorbertLindenberg]

Please label as Script-Javanese.

[adtbayuperdana]

Regarding the switched conjuncts of ṣa and sa, it is indeed "switched" if the script is used to write sanskrit or old Javanese. Contemporary rules (as this excerpt from a guide line published by the provincial government of Yogyakarta, east, and central Java in 2002) and indeed many Javanese manuscripts made in the colonial era has already switched the conjunct of these two aksara (as shown in an excerpt from the Raffles paper below, written in 1816).

excerpt from Pedoman Penulisan Aksara Jawa (2002) (circled) pasangan sa amongst the Raffles Papers, 1816. "True" pasangan of sa is never observed in handwritten Javanese manuscript.

"True" conjunct of sa are never observed in everyday writing of modern and contemporary Javanese. It is used in specialized writing, commonly scholarly work of old Javanese or Sanskrit literature. I also observed that this is quite popular among purists or enthusiast. While I believe that neither is wrong or right, it should be noted that the "switched" pasangan is the norm and "true" pasangan is not common. I think current pasangan of sa is more reflective to common Javanese.

[dougfelt]

@waksmonskiMT Sue, can you have someone involved with the font design respond to this thread?

[jungshik]

@kmansourMT : Do you have any input on this issue?

[kmansourMT]

I am traveling now but upon my return, I will review the details.

Kamal

From: jungshik notifications@github.com<mailto:notifications@github.com> Reply-To: googlei18n/noto-fonts reply@reply.github.com<mailto:reply@reply.github.com> Date: Monday, 24 October 2016 at 22:44 To: googlei18n/noto-fonts noto-fonts@noreply.github.com<mailto:noto-fonts@noreply.github.com> Cc: Kamal Mansour kamal.mansour@monotype.com<mailto:kamal.mansour@monotype.com>, Mention mention@noreply.github.com<mailto:mention@noreply.github.com> Subject: Re: [googlei18n/noto-fonts] Noto Javanese using unusual glyphs (#765)

@kmansourMThttps://github.com/kmansourMT : Do you have any input on this issue?

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHubhttps://github.com/googlei18n/noto-fonts/issues/765#issuecomment-255844551, or mute the threadhttps://github.com/notifications/unsubscribe-auth/AMJehXKa0v6loQk0p4AMVNHyTBrQxduFks5q3QqEgaJpZM4KTMhW.

[marekjez86]

Thank you @cahyoramadhani for a very thorough comparative investigation of various implementations of Javanese scripts. Could you provide me with some sample Javanese text (e.g., in an uploaded file that I could use as a test case)?

Also if I read correctly @mangajapa , the only issue is missing initial stroke in 7 glyphs characters (see " I think current pasangan of sa is more reflective to common Javanese.")

[adtbayuperdana]

@marekjez86 I also agree with cahyo that 34 glyph characters are missing initial stroke (all of those marked black in his table), of which the most illegible are GA MURDA (A993), YA (A9AA), and LA (A9AD) with their respective pasangans as well as PA (A9A5).

In printed Javanese, initial stroke loops are always maintained and never reduced. Some of them may be thin, but the general loop is always visible. These are some samples from 19-20th century publications:

Very thin-line like loops similar to noto can be found in handwritten manuscripts. But such loops are only present in select glyph characters (most notably KA) and are not uniformly applied. Here is a great sample from the Raffles Paper, written in Semarang, 1816:

Numbered 1 are KA and NA with a thin loop resembling the Noto stroke, but this is more a kin to a quirk of fast handwriting. As you can see in number 2, the KA has a more visible loop. In other glyph characters (numbered 3) the initial loop stroke are visible. Thin loops are more pronounced in letters such as KA, NA, and YA; in the extremes, they almost look like a single line even though they are not.

If we put this in a table, it should be clear that chopping the initial loop stroke on all glyph characters would made the font confusing, as the case in Noto:

[tjahaja]

Thank you @mangajapa for the 19th century samples!

@marekjez86 I attach the more recent examples of Javanese usage in textbooks. As @mangajapa has told, Noto likely makes use of handwriting style of Javanese script where loops tend to be written narrowly that they are visible like a single line instead. In prints, the loops are always as clear as shown below. I also put the handwriting style as it is still found in a Javanese magazine, Jaya Baya.

Javanese textbook for 5th graders in Yogyakarta (1994). Javanese textbook for 7th graders in Yogyakarta (2008). Jaya Baya magazine, 4th week of October 2016 edition. This one is showing the handwriting style. I mark some letters resembling YA (A9AA) in red and blue to show that the letter is written in two ways: one with clear visible loop (in red) and another with hardly visible narrow loop (in blue). Panjebar Semangat magazine, 22 October 2016 edition Djaka Lodang magazine, 22 October 2016 edition

I also encourage you to have a look at these proposals on Javanese scripts from 1992 to 2014.

[dougfelt]

@kmansourMT @waksmonskiMT did you investigate this?

[kmansourMT]

We’ve been aware for some time that the style chosen for Noto Javanese has not been well received by many users, primarily because it omits a typical lead stroke. The style had been accepted by the Google team at that time (2012/2013). Redoing Javanese would be a contractual decision at this point.

Kamal

From: dougfelt notifications@github.com<mailto:notifications@github.com> Reply-To: googlei18n/noto-fonts reply@reply.github.com<mailto:reply@reply.github.com> Date: Friday, 24 February 2017 at 09:48 To: googlei18n/noto-fonts noto-fonts@noreply.github.com<mailto:noto-fonts@noreply.github.com> Cc: Kamal Mansour kamal.mansour@monotype.com<mailto:kamal.mansour@monotype.com>, Mention mention@noreply.github.com<mailto:mention@noreply.github.com> Subject: Re: [googlei18n/noto-fonts] Noto Javanese using unusual glyphs (#765)

@kmansourMThttps://github.com/kmansourMT @waksmonskiMThttps://github.com/waksmonskiMT did you investigate this?

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHubhttps://github.com/googlei18n/noto-fonts/issues/765#issuecomment-282356607, or mute the threadhttps://github.com/notifications/unsubscribe-auth/AMJeheh9Hn8Gu3m9nayEOmgcHRk9PUg9ks5rfxfwgaJpZM4KTMhW.

[marekjez86]

@waksmonskiMT @kmansourMT : what would it take to make it compatible with UNICODE and what others say here? I don't think straight conversion is enough.

[kmansourMT]

There is no problem with Unicode compatibility here, but of design.

[marekjez86]

I stand corrected.

On Thu, Jun 22, 2017 at 12:21 PM, kmansourMT notifications@github.com wrote:

There is no problem with Unicode compatibility here, but of design.

From: Marek Jeziorek <notifications@github.com<mailto: notifications@github.com>> Reply-To: googlei18n/noto-fonts <reply@reply.github.com<mailto: reply@reply.github.com>> Date: Thursday, 22 June 2017 at 11:56 To: googlei18n/noto-fonts <noto-fonts@noreply.github.com<mailto: noto-fonts@noreply.github.com>> Cc: Kamal Mansour <kamal.mansour@monotype.com<mailto: kamal.mansour@monotype.com>>, Mention <mention@noreply.github.com<mailto: mention@noreply.github.com>> Subject: Re: [googlei18n/noto-fonts] Noto Javanese using unusual glyphs (#765)

@waksmonskiMThttps://github.com/waksmonskimt@kmansourMT<h ttps://github.com/kmansourmt> : what would it take to make it compatible with UNICODE and what others say here? I don't think straight conversion is enough.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHubhttps://github.com/ googlei18n/noto-fonts/issues/765#issuecomment-310470946, or mute the threadhttps://github.com/notifications/unsubscribe-auth/AMJehSvKjvZAsxr- 3sPT7UHrc97bzdcKks5sGrjpgaJpZM4KTMhW.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/googlei18n/noto-fonts/issues/765#issuecomment-310477097, or mute the thread https://github.com/notifications/unsubscribe-auth/AMtJf636a9eESzHKO0R2Csod8GR9Zz60ks5sGr64gaJpZM4KTMhW .

--

Marek Z Jeziorek [ 老马 ] | marekj@google.com | 312 725-6958

[adtbayuperdana]

What is the status on the Javanese glyph redesign? Has the missing initial lead stroke been fix? I've noticed that Noto sans Javanese seems to have became the default font displayed in android devices, while the issue of unusual glyph design has not been addressed properly. @dougfelt @kmansourMT @marekjez86

[Apsodia]

@kmansourMT please do what you can to make noto more legible as it is very important in preserving the usage of javanese script, especially as android phones will ship with them. It is fine if it's only a design decision but the illegibility of la ꦭ and pa ꦥ is a real dealbreaker that hampers its widespread usage...

[kmansourMT]

Thank you. I’ve noted your comments and have understood the implications for legibility.

From: Apsodia notifications@github.com Reply-To: googlei18n/noto-fonts reply@reply.github.com Date: Sunday, 25 March 2018 at 01:02 To: googlei18n/noto-fonts noto-fonts@noreply.github.com Cc: "Mansour, Kamal" Kamal.Mansour@monotype.com, Mention mention@noreply.github.com Subject: Re: [googlei18n/noto-fonts] Noto Javanese using unusual glyphs (#765)

@kmansourhttps://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_kmansour&d=DwMFaQ&c=euGZstcaTDllvimEN8b7jXrwqOf-v5A_CdpgnVfiiMM&r=AQ2sZoFCIfx9-sfXNsrCuXEZByIrA4R9G6bMa8TgemY&m=NtM0RqOQ7sI8s0gA851II8-2Hkp08b8CXCV57_FYVOM&s=qQCVFIKzEUZsXvIH-5oiveJ0lo-Z1yAbqERQheXUHWA&e= please do what you can to make noto more legible as it is very important in preserving the usage of javanese script, especially as android phones will ship with them. It is fine if it's only a design decision but the illegibility of la ꦭ and pa ꦥ is a real dealbreaker that hampers its widespread usage...

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHubhttps://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_googlei18n_noto-2Dfonts_issues_765-23issuecomment-2D375952672&d=DwMFaQ&c=euGZstcaTDllvimEN8b7jXrwqOf-v5A_CdpgnVfiiMM&r=AQ2sZoFCIfx9-sfXNsrCuXEZByIrA4R9G6bMa8TgemY&m=NtM0RqOQ7sI8s0gA851II8-2Hkp08b8CXCV57_FYVOM&s=OWnkqM-cXpQGNKifQzwHuuO6tTrbFmylveYT-gK2r5I&e=, or mute the threadhttps://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_notifications_unsubscribe-2Dauth_AMJeha7B8hyh9m4SpUTjb6N-2Dxg2QxGjEks5th078gaJpZM4KTMhW&d=DwMFaQ&c=euGZstcaTDllvimEN8b7jXrwqOf-v5A_CdpgnVfiiMM&r=AQ2sZoFCIfx9-sfXNsrCuXEZByIrA4R9G6bMa8TgemY&m=NtM0RqOQ7sI8s0gA851II8-2Hkp08b8CXCV57_FYVOM&s=fppiSc8Px4UBK4pUfqDDEDtwsqSEuMaJjeFM4bg594M&e=.

Kamal Mansour Linguistic Typographer Monotype 127 Second Street Unit 2 Los Altos, California 94022 Phone +1 (650) 305 2876 Our Brand Family ​ Monotype.com Olapic.com

[davelab6]

@marekjez86 I'm curious if you have any news on this? I ran into @mangajapa in a chat room and it seems that since noto is in Android it can "shift the culture" but if the Noto design here is "cut off" this is not a good thing

[marekjez86]

@cahyoramadhani @adtbayuperdana : this is unrelated to the problem listed here. However, I'd like to hear your opinion/feedback on two proposals for the replacement of the current Javanese which is included in the PDF below (the original is grey, the suggested change is in black). Please look at the pages # 2 and # 3 in the PDF and tell use what version do you prefer and why?

Proposal for NotoSansJavanese.pdf

[marekjez86]

@cahyoramadhani @adtbayuperdana : BTW, we are leaning toward page # 3 (low contrast design option) because it is matching better Noto Sans and Roboto font design. If you are opposed to it, what would be the major critique or argument against page # 3 (low contrast design option) becoming the new design?

[adtbayuperdana]

@marekjez86 I prefer the one on page # 3 as well, for the same reason that you have described

[tjahaja]

@marekjez86 Hi, the new design is more legible to me than the revised design, so I will pick that one.

[bre73]

Just a thought on the Noto Sans Javanese, maybe with a little more stroke contrast the "revision of the original" could potentially become Noto Serif Javanese, helping to expand the serifed options.

[marekjez86]

please check fonts in https://github.com/googlei18n/noto-fonts/tree/master/phaseIII_only/unhinted/otf/NotoSansJavanese or https://github.com/googlei18n/noto-fonts/tree/master/phaseIII_only/unhinted/ttf/NotoSansJavanese

[ipuzz]

Thanks dev,

꧋ꦩꦠꦸꦫ꧀ꦱꦼꦩ꧀ꦧꦃꦤꦸꦮꦸꦤ꧀꧉

so far so goooddd..... 😘👍👍