>I dont know if this url
>http://www.server.org/gestran/images/../../../../videoclub_2.zip
>is rtf compliant, but wget and IE translate it to
>http://www.server.org/videoclub_2.zip
>which is a right behaviour and the file gets downloaded so the user
get

No, that's WRONG behavior. If you feed garbage data into a program, and
it
tries to make sense of it, it may or may not give you the desired
result,
but one thing is sure, it does not return the CORRECT result, which
should be an error message.

If the URL is wrong, how do you know the "corrected" one it uses
is really the one you wanted? If software makes arbitrary changes to
the URL,
the user will never learn that the URL is wrong.

Should cURL include a "spell checker" too, and correct spelling of
words in URLs?

Internet Explorer is a terrible example to use, since it often sends
people off
to irrelevant nonsense sites if they make a typo when entering their
URL.
Unexpected behavior is always bad.

                                       
Received on 2004-10-18