 Previous Page |
Next Page  |
Duke's activities in computing in the humanities in general and in CALL in particular were institutionalized in the Humanities Computing Facility and the Duke University Computer-Assisted Language Learning (DUCALL) Project, with Prof. Frank Borchardt as Director and principal investigator. DUCALL currently employs a full-time staff of eight people. The remarks here on our experience in implementing Unicode may be helpful to other small software development teams in the same position.
DUCALL began full-scale development of a Microsoft Windows version of CALIS around 1989. Version 1.0 of CALIS for Windows 3.0, or "WinCALIS," was released in 1991. Like some other Windows programs with DOS roots (e.g., Borland Paradox), it made a lackluster attempt to maintain character code compatibility with the DOS version of CALIS, but as support was added for new languages and characters (such as the ANSI/Latin1 accented upper-case letters and other symbols not available in DOS), text ended up being a mongrel of codes which were neither purely ANSI nor purely DOS-IBM. Since the codes used for the extended character sets were from the beginning sui generis in DOS CALIS, character coding in WinCALIS 1.0 was a real coding morass. This situation was slightly improved in WinCALIS 1.1, by embracing the purely ANSI TrueType fonts for the West European Latin1 languages, which became available in Windows 3.1. But WinCALIS support for other non-European languages and scripts, ranging from Russian and Greek to Vietnamese and Japanese kana, continued to be based on ad hoc character codes.
Development of a Unicode-compliant new version of WinCALIS, version 2.0, began in the spring of 1991. The stimulus was provided by the addition of double-byte Chinese-language support in WinCALIS, which came about through merging WinCALIS with a Chinese word processor for DOS developed at Duke since 1983, called the "Duke Chinese Typist."
| Previous Page |
Next Page |