On Feb 18, 2010, at 9:15 AM, Karl Vollmer wrote: > I'd would do the entire site. UTF-8 has a base of standard ASCII so all > english characters would still render correctly even if the visitor does not > have UTF-8 installed on their system. They would lose the non-ASCII > characters, but if they don't have UTF-8 it's doubtful they have the fonts > for non-ASCII characters anyway. > > google.com is UTF-8, if it's good enough for google.... :)
Ok. So what we're specifically talking about is this, right: Index: includes/http_header.inc =================================================================== --- includes/http_header.inc (revision 1172) +++ includes/http_header.inc (working copy) @@ -1,7 +1,7 @@ <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN" "http://www.w3.org/TR/html4/strict.dtd"> <HTML> <HEAD> - <META http-equiv="Content-Type" content="text/html; charset=ISO-8859-1"> + <META http-equiv="Content-Type" content="text/html; charset=UTF-8"> <?php if (isset($meta_desc)) { print("<META name=\"description\" content=\"$meta_desc\">\n"); But even with this change, it's still safest to use the &# encoding of the Hebrew characters in mirrors.txt, right? Or does this change allow me to put back the Hebrew characters in mirrors.txt? Which way is preferred, or are they more-or-less equivalent? As a side note: I'm wondering how you come up with the &#... encoding for a random set of non-english characters. Are there UTF-8 -> &#encoding convertors or something? > As for MySQL - you don't appear to be using it, as otherwise we'd all have to > have MySQL databases... :) FWIW, we use a postgres DB for our regression testing database (see www.open-mpi.org/mtt/), but we don't allow that stuff to be mirrored. There's waaaay too much data accumulated each night, and it would also significantly raise the bar on the requirements for mirroring the Open MPI web site. Not desirable. :-) -- Jeff Squyres jsquy...@cisco.com For corporate legal information go to: http://www.cisco.com/web/about/doing_business/legal/cri/