-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Bruno Haible wrote: > When a user is getting an error message from 'msgfmt', and the user is a > translator who has never written any code in any programming language, > why should the error message she shall see be influenced by the syntax > of programming languages?
Why shouldn't it be? It shouldn't be influenced by programming languages, just to be influenced by programming languages, obviously: but when a quoting mechanism popularized by various programming languages happens to make _sense_, why not? > When a user is burning a CD from a description of its TOC stored in a file, > and this TOC file has a syntax error, why should the error message use > double-quotes and backslash as syntax characters? Because we need to use _something_ for syntax characters, and '\"' is far more understandable than '%22'. Or do you really think that people should have to pick up their ASCII and/or Unicode tables to understand an error message? > There are maybe 10 million users of our system and our programs. 9.5 millions > at least know URLs, and maybe 1 million knows what %20 means. Maybe 1 million know what %20 means (I personally think that may be generous). Maybe 9.5 millions know what URLs mean. How many will know what %22, %25, etc means? I'm guessing not many at all. > How many out > of the 10 million are C programmers? That's an irrelevant question. The relevant question is, How many people know what \" means? My expectation is, virtually 100%. Certainly a lot more than the 10% you hypothesize as knowing what %20 means (let alone the other characters). - -- Micah J. Cowan Programmer, musician, typesetting enthusiast, gamer... http://micah.cowan.name/ -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.6 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD4DBQFHtfVp7M8hyUobTrERAqBrAJjRMZXLm4CRN5yCfcuhcM5HgMKoAJ4omfs3 uDc7h3h70qFSMAORHBebfw== =KIEk -----END PGP SIGNATURE-----
