Path: utzoo!utgpu!water!watmath!clyde!att!osu-cis!killer!ames!lll-tis!lll-winken!uunet!mcvax!ukc!reading!onion!cf-cm!cybaswan!iiit-sh From: iiit-sh@cybaswan.UUCP (s.hosgood) Newsgroups: comp.sys.ibm.pc Subject: Re: US PC programmers still live in a 7-bit world! Summary: Ascii gives problems even in Britain.. Message-ID: <29@cybaswan.UUCP> Date: 8 Jul 88 21:14:58 GMT References: <1988Jun22.223158.1366@LTH.Se> <699@omen.UUCP> <1288@odyssee.UUCP> Lines: 55 The world needs an 8-bit international character set. The 'national characters' in ascii give problems to us Brits too, though we've got the 'right' alphabetic characters, we run into the fact that the pound sign often appears on our terminals instead of the dollar sign, or sometimes (rarely) instead of the hash mark. The nuisance is that the treatment is inconsistent, and frequently printers on a system disagree with half the VDUs! Sometimes you type a document with a price in pounds on one machine only to get it in dollars by magic in the final output. The conversion is done with no reference to the current exchange rate of course! Hex constants in assembler etc look silly if the leading dollar gets printed as a pound. 'Include' lines in C sometimes come out as '(pound) include'. The problems are so silly that no-ones come up with a systematic cure. You can get used to anything, but should you have to? I look forward to the day 8-bit codes take over. Everyone will have the full set of codes, and many more like the 'degrees' symbol and Greek letters etc etc. However, even if you allow for the letters of French, German, Icelandic and many others, somewhere, someone will miss out. The Celtic languages often need vowels with circumflexes. Esperanto needs quite a few other letters with circumflexes, and the 'u' needs an upside-down one! You would need the letters in upper and lower case too.. To be fair to everyone (who uses the Roman or Greek characters), when a newer character set is accepted, it ought to have the 'more popular' characters built in (i.e. for the use of the bigger natural languages), but surely it ought also to enshrine a standard way for the slightly off-beat characters to be sent as a digraph?? If you go and read the esperanto group, most of the arguments at the moment are about the various ways of representing thier circumflex characters. Just about everyone does it differently. They all involve digrams however, which are hard to read. P.S. Can you reprogram the character set in an IBM-PC to put in the characters you want?? Rumour has it that INT $1F points to a table of characters with codes $80 - $FF. It seems to be ignored on my clone-XT (monochrome adaptor fitted). I want to show those Esperanto digrams in a readable way! ---------------------------------------------------------------------- Steve Hosgood BSc, Image Processing and Systems Engineer, Institute for Industrial Information Techology, Innovation Centre, University of Wales, Swansea SA2 8PP Phone (0792) 295213 JANET: iiit-sh@uk.ac.pyr.swan Fax (0792) 295532 Telex 48149 My views are not necessarily those of my employers!