Brian, This explains things. Now I can update my resume... Thanks a lot! Anton. At 13:19 18/07/00 -0400, you wrote: > > I can read this file: > > > > read http://www.melbourne.net/antonr/resume.html > > > > but not this one: > > > > read http://www.melbourne.net/antonr/r�sum�.html > > ** User Error: URL error: http://www.melbourne.net/antonr/r�sum�.html. > > ** Where: read http://www.melbourne.net/antonr/r�sum�.html > > ><snip> > >An unescaped � is not valid in URLs, it should be encoded as %E9 >http://www.melbourne.net/antonr/r�sum�.html fails for me in IE5.5, but works >in NT4.7 >http://www.melbourne.net/antonr/r%E9sum%E9.html works in both. > >More details can be found in RFCs 1738 and 2396. > >In REBOL, (at least in REBOL/View 0.9.8.3.1) you must prevent the %E9 >encoding from being decoded too soon. A literal %E9 in a URL gets decoded >into � immediately; therefore in literal URLs you must escape the % by using >%25, or hide the url inside a string. >read http://www.melbourne.net/antonr/r%E9sum%E9.html doesn't work, but >read to-url "http://www.melbourne.net/antonr/r%E9sum%E9.html" does, and so >does >read http://www.melbourne.net/antonr/r%25E9sum%25E9.html. > >C programmers will recognize this as being similar to backslash encoding in >literal strings, where \n means a newline character, \t means tab, so \\ >must be used to mean a single \. > >Non-C-programmers will probably just roll their eyes. > >- Brian
