Re: reported a iconv improvement and now feel bad

[Dan Jacobson]

P> Your suggestion is quite reasonable, but I suspect that your bet isn't
P> quite correct.  (The best way to settle the bet is to write the patch.  :-)
>> 
>> how about it reporting what argument it ran into trouble?  Certainly
>> it uses them one by one?

Paul> I'm afraid not.  Try running the command "man iconv_open" for more info.

OK, I did that, and it was over my head.  Somewhere deep in these
processes there is a point where we run into the first invalid iconv
-f or -t argument.  Seems like some pretty inflexible program writing
if we can't somehow get that information back to the user from right
there, no matter how deep.  And, since it is a fatal error, stop there
too, instead of after doing several things I suppose.
-- 
http://jidanni.org/ Taiwan(04)25854780