- From: Erik van der Poel <erik@netscape.com>
- Date: Wed, 11 Dec 1996 09:08:03 -0800
- To: "Martin J. Duerst" <mduerst@ifi.unizh.ch>
- CC: Larry Masinter <masinter@parc.xerox.com>, avine@dakota-76.Eng.Sun.COM, rosenne@NetVision.net.il, www-international@w3.org
> In contrast to the Latin/Cyrillic issue > for Serbian/Kroatian, where automated conversion on the client side > is the solution, conversion to Latin letters (called Romaji in > the case of Japanese) can not be automated, neither at the > server nor at the client. A long time ago I used to use NEmacs (Nihongo Japanese Emacs) with Wnn (input method). Wnn could also do gyaku-henkan (reverse conversion), which I used to convert Kanji to Kana. From Kana, it is a small step to Romaji. I agree, though, that Kanji->Kana cannot be done 100% deterministically. Erik
Received on Wednesday, 11 December 1996 12:10:36 UTC