On Wed, Oct 25, 2000 at 12:31:04AM +0200, Kaweh Kazemi wrote:
>then i call the CGI script again (again without parameters, this time
>without my CGI exception catcher)
>and i get following traceback:
>      File "C:\Program Files\Python\quixote\publish.py", line 174, in
>traverse_url
>        if original_path[-1] != '/':
>    IndexError: string index out of range

OK.  This is a minor bug; PATH_INFO is an empty string, so accessing
the first character fails.  What should Quixote do with an empty
string?  Perhaps the most reasonable behaviour is to assume that /q/
is equivalent to /q/index.

>  *) could you have a look at the attached example. i am trying to
>  print the attributes of the response and request objects as a
>  test. but the output seems not to be as expected (try to refresh
>  the output, and there is no output every second time, also the
>  HTML tags are not interpreted). am i missing something?

Hmm... trying your example, it works fine for me and I can't see any
obvious problems in the code.  Are you using MS IE?  IE is notorious
for ignoring the content-type header and just guessing based on the
extension of the URL, which might explain the HTML tags not being
interpreted.

--amk