Tuesday, February 27, 2007

Typing Romanian in Word

Recently in the Apple forums a user complained that he couldn't do Romanian in MS Word 2004 using the standard fonts Arial and Times New Roman. At first I was puzzled, since MS specifically provides special enlarged versions of these fonts with Word that contain Arabic, Hebrew, Greek, Russian, and Vietnamese in addition to more common Latin languages. But sure enough, when I looked more closely, I found that Version 3.05 of Arial and Times New Roman are missing the letters s-comma and t-comma required by Romanian.

It turns out that updated versions of these fonts with the Romanian characters added are available to WinXP users via download from MS since Dec. 2006, and that the new versions are also automatically included in WinVista. Unless you can get copies of these, you'll have to use other fonts for Romanian in OS X -- Courier, Geneva, Helvetica, Hoefler, Lucida Grande, Monaco, Palatino, and Times should all work, along with various downloadable fonts that cover the whole Latin range, like Doulos, Gentium, Charis, and Everson Mono.

See this article for another issue relating to Romanian in OS X.

Monday, February 26, 2007

More Ways to Type Spanish

If you want to type Spanish on your Mac and are used to Windows, you may run into problems figuring out which keys to use for accented characters and the inverted punctuation marks. That's because, out of the 6 commonly-used Spanish keyboard layouts provided on the two platforms, only two (Mac Spanish ISO and Windows Spanish) are the same. How this situation came about, I have no idea. You can see graphics comparing the six here.

My understanding is that many Spanish users prefer the Windows US International or Latin American layouts. For the US International layout, a version can be downloaded from this page, and for Latin American (winlatinamerican.keylayout) you can go to my iDisk.

Of course you can also type Spanish on the normal US layout: The deadkey for the acute accent is Option + e and for the tilde it is Option + n, with ¿ at Option + Shift + / and ¡ at Option + 1.

Sunday, February 25, 2007

Fixing the Brazilian Keyboard Layout

If you try to type Brazilian Portuguese, you might be puzzled by the OS X keyboard layout that comes with the Brazilian flag attached to it. It's neither Portuguese nor Brazilian, but simply the normal US layout, where you type accented characters using Option deadkeys.

As I understand it, Brazilians are used to typing with a layout called ABNT or with the layout called US International PC. You can get an ABNT layout (winbrazabnt.keylayout) here.

Saturday, February 24, 2007

Missing Vietnamese Characters?

If you are having trouble browsing Unicode Vietnamese web pages, like the BBC Vietnamese site, seeing boxes or question marks instead of characters with two diacritics, it probably means you need an extra font. The problem arises because certain Vietnamese sites stipulate the use of the Arial font in their html code, but the Arial (2.60) that comes with OS X 10.4 does not contain the precomposed Vietnamese characters in the Latin Extended Additional Unicode block. Most people do not notice a problem, because at some point they installed a trial or other version of MS Office2004, which puts a more complete version of Arial (3.05) in your Home/Library/Fonts folder. If you *do* have missing characters, try to find a copy of this font for your machine.

Wednesday, February 21, 2007

Typing Yoruba

Yoruba is one of the major languages of Nigeria, and many speakers have also emigrated to the UK, Brazil, and the US. It is written with the normal English alphabet plus a few extra "dotted" letters: Ẹ/ẹ, Ọ/ọ, Ṣ/ṣ. Also tone marks are required, grave ` for low and acute ´ for high. Lucida Grande is probably the best font to use.

The dotted letters can be made with the US Extended keyboard layout, using Option + x, followed by the base letter. Tones can then be added using Option + Shift + ` and Option + Shift + e. You can also download a custom Yoruba keylayout from my iDisk. This puts the dotted letters on the 1, 2, and 3 keys and all the vowels are deadkeys. You type the vowel and then choose a) 4 for grave accent/low tone, b) 5 for acute accent/high tone, and c) any other key for no accent/mid tone. The Naira symbol is on the = key. This layout also has the special characters needed for Hausa and Igbo (Ɓ ɓ Ɗ ɗ Ƙ ƙ Ƴ ƴ Ṅ ṅ Ị ị Ụ ụ).

Sunday, February 18, 2007

Typing Hausa

Hausa is an official language in northern Nigeria and widely used in other Muslim areas of West Africa as well. The standard Latin orthography uses the normal English alphabet, plus the hooked letters Ɓ ɓ Ɗ ɗ Ƙ ƙ Ƴ ƴ. Of the fonts included with OS X, only Lucida Grande has these, but others can be downloaded. In addition, some publications have used a dot underneath instead of the hooked letters, and some dictionaries and teaching aids may also mark vowel length using a macron ¯ and vowel tones using grave ` (low), acute ´ (high), and circumflex ˆ (falling).

All of these can be made using the US Extended keyboard layout, a chart for which is here. The hooked letters are made using Option + Shift + . (period), followed by the base letter. Another solution is to use the custom Hausa keyboard layout from my iDisk. The hooked letters are located on the Option version of the base letter.

Hausa has also been written in Arabic script, and examples of that can be found on Nigerian banknotes.

Saturday, February 17, 2007

Blog Widget Available

For anyone who wishes to have access this this blog via a Dashboard Widget, you can download one here.

Friday, February 16, 2007

Translating To and From Foreign Languages

Mac OS has long provided a language translation facility, first via Sherlock and, starting with OS X 10.4, the Translation Widget in Dashboard. In both cases there is no translation application on your machine: Sherlock and the Widget are simply interfaces to an online translation service, in this case Systran, which can also be accessed via any web browser.

Anyone who uses this to translate foreign text into his native language will quickly see the limitations. Relying on it to translate your native language into one you don't yourself know, in the expectation of making yourself correctly understood, is not realistic. The reasons for that lie in the inherent shortcomings of Machine Translation. Only humans can really do it right.

For alternatives to what comes with OS X, there is a good list here. Some worth checking out are LiveDictionary, WordLookup, TranslateIt, Google's Language Tools, and Langenburg.com, which is a gateway to multiple services.

For translation software that runs on your Mac rather than via the Internet, there is NeuroTran and InteractiveTran. For your iPod, you can try iLingo.

Monday, February 12, 2007

Future Additions to Unicode?

There are various languages, scripts, and characters which have not yet made it into the Unicode standard. If you are interested in where something stands in that regard, the best resource is the Proposed New Scripts page at Unicode.org. There you will currently find about 4 dozen in 4 categories -- Exploratory Stage, Committee Review, ISO Balloting, and Pre-Publication. Another useful page is the Roadmaps of proposed allocations of Unicode codepoints. If you want to see what characters might be added to existing scripts, the right place to look is the Character Pipeline.

Saturday, February 3, 2007

Support for Dravidian Languages

The Dravidian languages, which form a group apparently unrelated to any other, are spoken in Southern India and in Sri Lanka. OS X 10.4 comes with support for Tamil, and you can find info on how to use it here. For typing in the other main family members -- Kannada, Malayalam, and Telegu -- you need to download and install fonts and keyboards, using either the commercial kits available from XenoTypeTech or the free stuff listed below:

For Kannada, you can get the Kedage font and a keyboard layout from Nick Shanks' site.

For Malayalam, you can get the Rachana font and a keyboard from the MacMalayalam Site.

For Telugu, use the Pothana font from Nick Shanks and the keyboard layout from my iDisk.

Web sites in these languages may still be using custom encodings instead of Unicode. In that case you will need to download whatever font they require (e.g. manorama for manoramaonline.com) and you may need to try different browsers as well.

Thursday, February 1, 2007

Windows Vista Internationalization

With the official release of Windows Vista, the MS pages giving info on international support have been updated here.

To see keyboard layouts, I think you need to use either the old IE for Mac or Opera, as neither Safari nor Firefox seem to display them.