I couldn't help myself. This stuff is 100% not Unicode. One of the ways you can test this yourself - open up the Glyphs menu, and choose your favorite Tibetan font. Mouse over one of the glyphs, and wait for the mouseover popup. What does it say? If it is actually Unicode, it won't say "NULL" or "Private Use Area" - it'll say the name of the character.

These are actually worse than "NULL" or "Private Use Area" or "Unassigned" - when I mouse over anything in a font that starts with "Monlam" I see things like "LAO TONE MARK NGO" or "THAI CHARACTER LO LING." It's a hack job - and oh boy, I have way too much experience with hack jobs like this font. They're one of the best ways to hamstring a complex-script publishing project. They've written over the perfectly valid Unicode code points for other scripts with their own Tibetan glyphs. I mean, there's a right way to put glyphs into a Unicode font if you don't want to put 'em at their appropriate codepoints for the language, and these folks did not do that.

So, the World-Ready Composer and Contextual Alternates won't help you at all, if you're using these fonts. I might be able to help you figure out a solution - but it could take many, many hours to figure out how things were encoded, if there's a conversion utility available, if there's a trustworthy conversion utility available... or, lastly, if you can keep on using your old pre-Unicode hack-fonts with some deft config work in ID or elsewhere. My initial guess on that last one is "no" at least without duct-taping those "legacy" fonts with even more hack-work.

I am using InDesign CS5 on Mac OS X 10.6.4. It appears that CS5 has a severe regression in the handling of Tibetan versus CS4. It simply doesn't work with any of the usual fonts. Here are screenshots showing CS4 and CS5 entered using the Wylie keyboard supplied with Mac OS X:

as can be seen from the above Chris Fynn's font and Tony Duff's both work correctly in CS4 whereas the Mac OS X supplied font does not. In CS5:

here we see that Jomolhari sort of works except for the odd glyphs appearing to the left of the margin and the weird block at the end that seems to be an unselectable entity. Tibetan Machine Uni fails on virtually all stacks but not as badly as Kailasa - which fails also of course on CS4 as seen above.

Too bad that we have regression but at least CS4 is still working with CS5 present on my system.

Thanks,

Chris

I'm suddenly too busy (as of twenty minutes ago) to spend any significant time on this today, but I will be checking up on this later this evening. I will certainly be testing those fonts you posted. But here are the basics:

1) World-ready composer has to be on for any genuine Tibetan Unicode font to work

2) The world-ready composer is hidden, because it's obviously not ready for prime-time. Tibetan is in the list of potentially functional complex-scripts, but Adobe hasn't tested it. That means that we (complex-script typesetters) don't get any guarantee that WRC will function without problems.

3) Complex-script fonts for languages not spoken by millions of wealthy people tend to lag behind in font development, adherence to Unicode spec, et cetera. So, there are a wide variety of fonts out there, built with widely varying skills (drawing letterforms, using OpenType tables, respecting encoding specifications, et cetera). Seeing Tibetan mapped on to Thai and Lao codepoints looks very, very bad to me.

4) 2 & 3 taken together mean: be very careful! Your font may not be built to spec. It may be built with certain assumptions (i.e. "will be rendered by Uniscribe") that may not pertain in your case.

5) The font-rendering parts of InDesign (which I knew as "CoolType" back when it first emerged, although I think that may be an internal Adobe name for their type-rendering engine) operate completely independently of the host platform. So, there should be no difference between font rendering on Windows and OS X.

6) Why would CS5 render things differently that CS4? This is aimed at xristyNepal: I think that all of the fonts you listed in your post but one are, to one extent or another, workarounds. So, my warning in no.4 applies very strongly here. The one exception is the Mac font supplied with OS X: I'd bet that it's an AAT font, not an OpenType font, and therefore wouldn't render correctly in InDesign at all, ever. I don't know, and I'll look into it, but that is my first guess.

I didn't get a chance to do any testing last night; maybe tonight.

I don't know what spec you're talking about. Is that a spec for how to design fonts for the Adobe rendering engine used in ID?

No, I mean adherence to the Unicode spec for character encoding (don't encode Tibetan as Lao, use the PUA for glyphs with no codepoint defined in Unicode) and OpenType for building fonts (use OpenType methods to compose characters, lots of other details I don't know).

Jomolhari ID was specifically developed with the purpose of being rendered by ID - I guess that's a workaround but since ID has its own rendering engine, then designing for it wouldn't seem to constitute a workaround.

I wonder if it was designed to a) render in ID, or designed to b) render in ID with the World-Ready Composer turned on. If it's a), I'd wager that it would need to be updated from version to version as InDesign matures, and that there are probably workarounds (unlike the dirty hack of mapping Tibetan to codepoints for other scripts) built into the font. If it's b), then it should adhere to the OT spec. I won't know until I get a chance to look at my Tibetan sample text in all of these fonts. The page claims that it was made for "InDesign CS3 and above" but I don't know if that covers CS5, as workarounds for one version of InDesign may not work well in another version. I did pop it open and it looked well-made (lots of ccmp and rlig in there) but to be honest, I'm a little out of my depth in Tibetan - I can eyeball Khmer or Lao, for example, and know if they're composed correctly, but I don't do Tibetan often enough to spot all of the potential drops/misalignments/instances of poor composition/etc. I will do some tests and post some screenshots for you and get your feedback.

As a daily user of Tibetan in ID, it seems clear that there was a regression from CS4 to CS5 that remains in the current version 7.0.2.

 

The IndicPlus plugin does solve the problem in CS5 but it was not needed in CS4 for typing Tibetan but was essential in CS3.

Please satisfy my curiousity - are you using the World-Ready Composer?

In CS5, Jomolhari-ID looks like it might work with the WRC, and it certainly fails without. I sent you a PM - let us know how well it works.      

Chris:

I am a dummy computer user. Can you explain what is the verdict on using monlam font on INDESIGN CS5?

Thanks  you very much Peter. That was extremely helpful. Did save tons of hours.

Thanks to domischi!

In the Chinese version of InDesign CS5

This script should be change to <<paragraphStyle.composer = "Adobe 全球適用段落視覺調整";>> or it will not work.

我寫了這個步驟說明在InDesign中文版中如何編排藏文

Just to clear things up....

The Jomolhari fonts support ordinary Unicode for Tibetan script using only characters in the 0Fxx block; as well as  "precomposed Tibetan"  conforming to the Chinese national standard. This maps glyphs for precomposed Tibetan stacks to characters in the private use area (PUA) of the Unicode / iso10646 standards.[Precomposed Tibetan Set-A  U+F300-U+F8FF; Precomposed Tibetan Set-B  U+F0000-U+F16FF]

The ordinary Jomolhari font uses some of standard OpenType features for Tibetan (ccmp, blws, abvs, calt and kern) under the Tibetan "tibt" script tag  and "dflt" language tag to compose glyphs. Where a precomposed glyph exists in the repetoire mapped to PUA characters, of course the lookups point to those glyphs (but not to the PUA characters). For ordinary Unicode it irrelevant whether or not those glyphs have PUA mappings as well.

InDesign CS3 and later versions without WorldReady composer did not support Tibetan rendering. However I found that by putting lookups under the DFLT script tag (instead of the tibt script tag), and using only OpenType features used in Latin fonts, then the font would work in InDesign. However these lookups do not work in Microsoft applications etc. as only some Adobe apps support lookups under the "DFLT" script tag.

So the Jomolhari ID font  was made to work in InDesign CS3 it uses OpenType features under the Default "DFLT" script tag and  "dflt" language tag tag to compose glyphs. These lookups are under features  (ccmp, rlig, calt  and kern) also used in Latin fonts. This means the font should work in InDesign without WRC.

The reason for creating two versions of the font at the time was that, Microsoft applications, using Uniscribe, rendered Tibetan only using features and lookups under the "tibt" script tag and Adobe applications using Cooltype would render Tibetan only using lookups under the DFLT script tag. So Microsoft's Uniscribe layout engine ignored lookups inder the DFLT script tag but applied those under the tibt script tag - while Adobe's Cooltype engin in InDesign did the opposite.

At first Jomolhari and Jomolhari ID each had only one set of lookups. So you could enter your text in MS Word using Jomolhari then bring the text into InDesign and everything worked fine.

Later I put both sets of lookups into Jomolhari ID (version 3d and above) - so  a single font worked in both Windows and in Adobe applications. DDC Uchen, another Tibetan script font I made  later, works this way too.  There is no need for WRC with these fonts. If you do use WRC, there may be some potential for conflict between the two sets of lookups because both are being applied. They should work together fine - but that has not been tested.  I made the Jomolhari font on my own time and at my own expense and can't afford to update to the latest versions of CS to do the testing. However both these fonts have been successfully used in Bhutan for over three years by all the  Dzongkha newspapers and publishers - as well as by publishers of classical Tibetan religious texts - and they have not reported any problems to me. If you do have problems with particular combinations, let me know and I will try to fix them, when I have time.

I suspect any "regression" you see between CS4 and CS5 may be because CS5 is no longer applying the features and lookups under the default "DFLT" script tag but, using WRC, is trying to apply the features and lookups under theTibetan "tibt" script tag but some of these features may not yet be fully supported by Adobe's CoolType rendering engine. 

Concerning the support for "precomposed Tibetan" in these fonts. First, it should in no way interfere with them being used as regular Unicode fonts. The support is there to enable the fonts to be used in some applications used in China which support this encoding, but have no support for complex script rendering. It also enables the use of these fonts in applications like the Corel Draw suite which "support Unicode" - but have no support for complex script rendering of Tibetan text. This also applies to some versions of PhotoShop. Of course if you have need to use this encoding you need a way of converting from standard Unicode to pre-composed Tibetan. Andrew West's BabelPad, a free Unicode text editor, enables this.

Using the BhoView application, the DDC Uchen and Jomolhari fonts can also be used on Android, which currently has no support for Tibetan rendering. This application converts between standard Unicode and the precomposed Tibetan encoding on-the-fly using Java to display Tibetan text correctly.  There are also some Android phones, available in China using the DDC Uchen font which have the whole interface and applications translated into Tibetan and use a similar method of converting between standard Unicode and pre-composed Tibetan.

- Chris Fynn

Hello,

Put the attached scripts in the install like:

Then in the Window menu select Utilities, Scripts:

Then you will see the Scripts window and can select for example the para composer - which is what I usually use - with a Tibetan paragraph selected:

Regards,

Chris