rees@apollo.uucp (Jim Rees) (03/05/88)
I've been playing around with 8-bit fonts, specifically Latin-1 fonts. The uucp transport layer is 8-bit transparent, and so is readnews. If you make one small change to inews, you can use 8-bit fonts too (assuming your system supports them). It's kind of neat to read through news articles and see all the accent marks on French names. How widespread and how standard is Latin-1 for expression of international characters? Someone from Europe want to comment? Rn apprently strips the high bit somewhere. I haven't delved into it yet. But here's the fix for inews (your line numbers will vary): % diff -c inews.c.bak inews.c *** inews.c.bak Fri Dec 18 06:10:05 1987 --- inews.c Fri Mar 4 14:03:20 1988 *************** *** 1230,1240 **** #endif /* !NFSCLIENT */ if (mode != PROC && tty && STRCMP(bfr, ".\n") == 0) break; for (cp = bfr; c = toascii(*cp); cp++) { ! if (isprint(c) || isspace(c) || c == '\b') ! putc(c, tmpfp); ! if (c == '\n') ! linecount++; } if (bfr[0] == '>') linserted++; --- 1230,1245 ---- #endif /* !NFSCLIENT */ if (mode != PROC && tty && STRCMP(bfr, ".\n") == 0) break; + #define islatin1(c) ((unsigned)(c)>=0xa0) for (cp = bfr; c = toascii(*cp); cp++) { ! if (islatin1(*cp)) ! putc(*cp, tmpfp); ! else { ! if (isprint(c) || isspace(c) || c == '\b') ! putc(c, tmpfp); ! if (c == '\n') ! linecount++; ! } } if (bfr[0] == '>') linserted++;
marius@rhi.hi.is (Marius Olafsson) (03/07/88)
From article <3aa84fac.b8ab@apollo.uucp>, by rees@apollo.uucp (Jim Rees): > How widespread and how standard is Latin-1 for expression of international > characters? Someone from Europe want to comment? We use ISO8859/1 (Latin1) for our native language domestic postings (we need to use 20 extra letters from 8859). This 8-bit functionality of the news software is best done simply by substituting an ISO8859 version of the 'ctype' macros, and then the changes to well written reader programs are minimal (as long as they do not use the 8th bit for something). (I can provide the patches to vnews/rn/vn (with modified ctype macros) in the unlikely event anyone is interested.) It is my understanding that the RFC's dealing with the exchange of Usenet messages specify that the text must be 7-bit ascii, so we make sure that none of this leaves the country - Is this true? -- Marius Olafsson marius@rhi.hi.is University of Iceland {mcvax,enea}!hafro!rhi!marius
rees@apollo.uucp (Jim Rees) (03/09/88)
It is my understanding that the RFC's dealing with the exchange of Usenet messages specify that the text must be 7-bit ascii, so we make sure that none of this leaves the country - Is this true? Yes, but the standard simply codifies standard practice. Changes have historically been made in the protocol first, and the document last. So I think the right thing is to go ahead and post articles to the net with the new character set, and let the software catch up. It doesn't do any harm to any news installation that I know of, although the high bit will normally get stripped on the first hop, rendering the article useless. Personal to Marius: Please send me your ctype macros and rn patches. I sent you mail but have doubts about routing to your '.is' domain.
ct@dde.uucp (Claus Tondering) (03/11/88)
In article <3aa84fac.b8ab@apollo.uucp> rees@apollo.uucp (Jim Rees) writes: >How widespread and how standard is Latin-1 for expression of international >characters? Someone from Europe want to comment? Our company makes a M680x0-based multi-CPU computer. We use Latin-1 (ISO 8859/1) on it and that's really great! It is a big improvement over ASCII and various national variations of ASCII. -- Claus Tondering Dansk Data Elektronik A/S, Herlev, Denmark E-mail: ct@dde.uucp or ...!uunet!mcvax!diku!dde!ct
fortin@zap.UUCP (Denis Fortin) (03/22/88)
In article <3aa84fac.b8ab@apollo.uucp> rees@apollo.uucp (Jim Rees) writes: >I've been playing around with 8-bit fonts, specifically Latin-1 fonts. >The uucp transport layer is 8-bit transparent, and so is readnews. >If you make one small change to inews, you can use 8-bit fonts too >(assuming your system supports them). >It's kind of neat to read through >news articles and see all the accent marks on French names. ' Hmmm. It's more than neat... For people up here in Quebec, it's actually quite useful and interesting. >How widespread and how standard is Latin-1 for expression of international >characters? Someone from Europe want to comment? Well, this is not Europe, but a group of people around here are starting to toy with the idea of exchanging news and mail locally in ISO-Latin-1. A special header (X-Character-Set: ISO-Latin-1 ??? Any other suggestions???) would be used and the news would be translated to whatever character set the local machine uses when it is displayed (for example, on an IBM PC-type of machine or a DEC VT-220, you can represent some of the ISO-Latin-1 characters). Not being able to put accents in texts is a real pain when you're typing texts in French: a bit like typing English without descenders (imagine: it would be ouite oainful to have to oet bu without descenders :-). q p g y (Oh, by the way, Microsoft Windows uses ISO-Latin-1 internally.) >Rn apprently strips the high bit somewhere. I haven't delved into it yet. Sigh. Hopefully there's a quick fix. > But here's the fix for inews (your line numbers will vary): Thanks. -- Denis Fortin | fortin@zap.UUCP CAE Electronics Ltd | philabs!micomvax!zap!fortin The opinions expressed above are my own | fortin%zap.uucp@uunet.uu.net
weemba@garnet.berkeley.edu (Matthew P Wiener) (03/23/88)
>>Rn apprently strips the high bit somewhere. I haven't delved into it yet. > >Sigh. Hopefully there's a quick fix. This may or may not count as a fix or as quick. GNU Emacs carries all 8 bits, and hence so does Gnews, an rrn-emulator written in ELisp. (The rn code exists but is untested.) I don't know how you can get it up in Canada; it's sitting on ucbvax for anonymous FTP. FSF is waiting for the rn-code to be tested and debugged, and then they'll distribute Gnews for me. ucbvax!garnet!weemba Matthew P Wiener/Brahms Gang/Berkeley CA 94720