On Thu, Sep 25, 2008 at 8:23 AM, André Warnier <aw@xxxxxxxxxx> wrote: > In that respect, yes you should. The client should not have to know on > which platform the server is running, and send a fundamentally different URL > depending. You don't have to know the platform. You have to know the filename byte-by-byte (typically by following a proper link). > I did the following experiment : on a Linux Debian Apache 2.2.4 server, I copied the file "joaquín.gif" in the document root. I don't have too much experience here, but was the filename morphed into some other codeset during the copy (mount option, editor, etc)? > In any case, that the Windows incarnation of Apache 2.x would return a 403 code (instead of the less inappropriate 404) is still another issue. > The URL requested is not illegal, nor is this document submitted to any kind of permission or authorization. IMO The 403 is returned in a path where errors imply a high likelyhood of someone actively trying to fool the server -- I don't think a 403 is too inappropriate here. I don't know if any other status code would be appealing enough to motivate trapping this error early (i.e. don't bother trying to map to a file if there's invalid utf-8 somewhere in the URL on windows). -- Eric Covener covener@xxxxxxxxx --------------------------------------------------------------------- The official User-To-User support forum of the Apache HTTP Server Project. See <URL:http://httpd.apache.org/userslist.html> for more info. To unsubscribe, e-mail: users-unsubscribe@xxxxxxxxxxxxxxxx " from the digest: users-digest-unsubscribe@xxxxxxxxxxxxxxxx For additional commands, e-mail: users-help@xxxxxxxxxxxxxxxx
|