08 Jun 15 16:45, you wrote to me:   
      
    ac> 07 Jun 15 14:06, you wrote to me:   
      
    ac>>>> It should also be broadly useful for things like viewing CP437   
    ac>>>> "ANSI art", running nethack with the IBMgraphics option or   
    ac>>>> running EPIC with scripts that use CP437 artwork."   
      
    ml>>> i'm going over to the STATS echo now to see how my graphs are   
    ml>>> coming out ;)   
      
    ml>> i had to switch golded back to xlatlocalset cp437 and it works! the   
    ml>> graphs are showing the proper block characters and the message bodies   
    ml>> with double and single line frames are good...   
      
    ac> Cool.   
      
    ml>> i find it interesting how he's accomplishing this ""feat of majik""   
    ml>> too ;)   
      
    ac> Well you could look at the source code,   
      
   i did... that's why i found it interesting ;)   
      
    ac> but essentially cp437.c runs the output of the program through the   
    ac> iconv library to convert CP437 to UTF-8. iconv supports other   
    ac> charsets, eg. CP850 (see iconv -l), so you could mod the source and   
    ac> probably support other translations. Only one at a time though. Just   
    ac> like using codepages in DOS.   
      
   yeah, i thought about that...   
      
    ml>> perhaps golded should be using the same basic method by default on   
    ml>> *nix and having it available on winwhatever... it should be available   
    ml>> on OS/2 as i have programs that use it there already... if golded can   
    ml>> go this way, all those translation tables might be able to be thrown   
    ml>> out ;)   
      
    ac> GoldED has "experimental" support for iconv already (make ICONV=1) but it   
    ac> doesn't build in FreeBSD so I haven't tried it, and I know it's not going   
    ac> to fix the UTF-8 problems with GoldED. Using cp437.c works fine.   
      
   ahhh... i've tried the wide ncurses thing but maybe i did it wrong... the   
   instructions i saw had it before the make... are both wide curses and iconv   
   needed at the same time? my idea is/was that golded would read the   
   characterset from the control lines and use that to feed to iconv as the   
   "from" with the localset as the "to" and pump the message body through it to   
   display the result afterward... if one changed force changed the character set   
   of the message, golded would reprocess through iconv using the new forced   
   character set as the "from"...   
      
   when writing, the follow would go the other way, of course... the end result   
   being the proper glyphs displayed and the proper bytes stored in the message   
   base with the proper character set control line...   
      
   i'll try to play with it later... can't really do that while reading and using   
   the cp437 wrapper ;)   
      
   )\/(ark   
      
   ... Velveeta on Wonder Bread with a Bud Lite!   
   ---   
    * Origin:  (1:3634/12.73)