lynx-dev relative URL handling bug

[Vlad Harchev]

 Hi,

 Just found a page on some site, it's URL was something like
http://www.blah.com/foo.html
The page contained broken relative references of the form:
        ../ftp/bar.html
(broken since the page they were contained has no directory part).
Following these links made lynx visit 
        http://ftp/bar.html
I.e. the site name was trimmed when computing absolute URL.

Dunno how other browsers treat such bugs, but I guess they just drop "../" if
there are no directories in the base url.

 It seems it would be easy to fix it.

 PS: I'm using the old lynx, so that may be fixed in current one.

 Best regards,
  -Vlad


; To UNSUBSCRIBE: Send "unsubscribe lynx-dev" to address@hidden