Re: 21.4.10-mule codings problem, ~ appears

Christoph> Stephen, if that is so[1], you should forward them to Christoph> me. OK. Christoph> Well, it might be similar, but not very[2], and it Christoph> would not prevent the ~ for the character `alpha'. If it's a non-Latin character set (ie, private to x-symbol), that's quite possible. latin-unity has not been extended to deal with non-Latin characters. This isn't very nice to Greeks, Middle Easterners, and Cyrillic users, but it is reasonably safe for the Asians (who generally use some variety of ISO-2022 if they're not already using Unicode). Christoph> [2] assume the default encoding is latin-9 and a file Christoph> containing latin-2 chars: Christoph> * with latin-unity: XEmacs would still store a latin-9 Christoph> encoded file, latin-unity would remap a latin-2 zcaron Christoph> to a latin-9 zcaron, i.e., store a byte with value 184 Christoph> and XEmacs would fail to store a zacute (with a single Christoph> byte) Go ahead, try it. Set latin-unity-preapproved-coding-systems to (buffer-default), buffer-file-coding-system to iso-8859-15, and insert those two characters in an otherwise ASCII buffer, and save. You should get (I do): ------------------------------------------------------------------------ Choose a coding system to save buffer tmp. All preapproved coding systems (buffer-default=iso-8859-15) fail to appropriately encode some of the characters present in the buffer. Please pick a coding system. The following are recommended because they can encode any character in the buffer: iso-8859-2 utf-8 iso-2022-jp iso-2022-7 ctext escape-quoted Note that if you select a coding system that can not encode some characters in your buffer, those characters will be changed to an arbitrary replacement character, by default `~', on output. Page down for more information on coding systems: utf-8, iso-2022-7, and ctext support all characters safely. iso-2022-7 and ctext are ISO 2022 conforming coding systems for 7-bit and 8-bit environments respectively. Be careful, there is a lot of software that does not understand them. utf-8 (Unicode) may also be unsupported in some environments, but they are becoming fewer all the time. utf-8 is recommended if usable (except for some users of Asian ideographs who need to mix languages). In Mule, most iso-* coding systems are capable of encoding all characters. However, characters outside of the normal range for the coding system require use of ISO 2022 extension techniques and is likely to be unsupported by other software, including software that supports iso-2022-7 or ctext. For a list of coding systems, quit this command and invoke `list-coding-systems'. ------------------------------------------------------------------------ Excessively verbose, but you shouldn't lose data. -- Institute of Policy and Planning Sciences http://turnbull.sk.tsukuba.ac.jp University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN Ask not how you can "do" free software business; ask what your business can "do for" free software.