BTW, can you try to nail this down somewhat more, e.g. by using
numerical ip addresses for URL and/or proxy ?

samg samg wrote:
> Hello everybody,
>
> This little program has some problems ...
> _______________________________________________
> import pycurl
>
> def debug(t, b):
> if t<3:
> print "debug(%d): %s" % (t, b)
>
> def body(buf):
> pass
>
> c = pycurl.Curl()
> c.setopt(pycurl.URL, 'http://XXXXX')
> c.setopt(pycurl.VERBOSE, 1)
> c.setopt(pycurl.WRITEFUNCTION, body)
> c.setopt(pycurl.DEBUGFUNCTION, debug)
> c.perform()
> c.close()
> _______________________________________________
>
>
> I made my tests with 2 pyCurl versions :
> - pycurl-ssl-7.13.0.win32-py2.4.exe : libcurl/7.13.0
> OpenSSL/0.9.7e zlib/1.2.2 c-ares/1.2.1 libidn/0.5.13
> - pycurl-7.13.0.win32-py2.4.exe :libcurl/7.13.0
>
> With the 2 versions of pyCurl, all is ok when i try to
> get a page from an intranet site :
> debug(0): About to connect() to XXXXX port 80
> debug(0): Trying x.x.x.x...
> debug(0): connected
> debug(0): Connected to XXXXX (x.x.x.x) port 80
> debug(2): GET / HTTP/1.1
> debug(1): HTTP/1.1 200 OK
>
>
> When i get a page from an internet site by a proxy,
> it's ok with the No-SSL version :
> debug(0): About to connect() to proxy port 8080
> debug(0): Trying x.x.x.x...
> debug(0): connected
> debug(0): Connected to proxy (x.x.x.x) port 8080
> debug(0): Proxy auth using Basic with user 'XXXXX'
> debug(2): GET
> http://www.python.org/pics/PyBanner049.gif HTTP/1.1
> debug(1): HTTP/1.1 200 OK
>
> But with the SSL version :
> debug(0): Could not resolve host: www.python.org
> (Domain name not found)
> debug(0): Closing connection #0
>
>
> What's the problem ? (I suspect c-ares)
>
> For information, the OS is windows 2000 and I have no
> problem with the curl program (with or without SSL)
>
> Regards
>
>
>
>
>
>
> Découvrez le nouveau Yahoo! Mail : 250 Mo d'espace de stockage pour vos mails !
> Créez votre Yahoo! Mail sur http://fr.mail.yahoo.com/
> _______________________________________________
> http://cool.haxx.se/mailman/listinfo/curl-and-python
>

-- 
Markus Oberhumer, <markus@oberhumer.com>, http://www.oberhumer.com/
_______________________________________________
http://cool.haxx.se/mailman/listinfo/curl-and-python