Hi, I want to thank Prof Ripley for his very helpful suggestions re internet proxies. This is great -- very useful! Marwan
-------------------------------------------------------------------------------- ---------------------------------------------- Marwan Khawaja <http://webfaculty.aub.edu.lb/~mk36 if you have MS Explorer -------------------------------------------------------------------------------- ---------------------------------------------- From: [EMAIL PROTECTED] Date: Wed, 22 Jan 2003 19:07:42 +0000 (GMT) To: [EMAIL PROTECTED] Subject: [R] Using Internet proxies There have been quite a few questions recently about this, so I have tried to gather experience. I set up a proxy using Apache2 (a very common server) behind our firewall and tried various authentication approaches. One comment: all the methods return error messages when they fail. Please don't report `it doesn't work' without the full details. 1) For a proxy that authenticates at most by hostname/IP address, and for Windows users, just set up IE6 to work, and use the --internet2 command-line flag. This worked for me (despite various claims here). 2) For a proxy that authenticates at most by hostname/IP address, the internal download.file method works, provided you set the environment variable http_proxy or HTTP_PROXY correctly, e.g. in ~/.Renviron. This is almost of the same form as used by wget, but is less tolerant, and note that setting `no_proxy' disables the proxy for all sites (unlike for wget). 3) If you have a proxy that demands that you enter a username/password combination, you can use the internal download method in R-devel: see ?download.file. 4) Installing wget and using the options(download.file.method="wget") provides a highly tunable approach. For Windows users, wget is still available on http://www.stats.ox.ac.uk/pub/Rtools. Note that none of these methods support proxies that want more advanced methods of authentication, e.g. Digest under HTTP/1.1. If there are any such proxies, please can a user provide us with a proven method. -- Brian D. Ripley, [EMAIL PROTECTED] Professor of Applied Statistics, http://www.stats.ox.ac.uk/~ripley/ University of Oxford, Tel: +44 1865 272861 (self) 1 South Parks Road, +44 1865 272866 (PA) Oxford OX1 3TG, UK Fax: +44 1865 272595 -------------------------------------------------------------------------------- ---------------------------------------------- Marwan Khawaja <http://webfaculty.aub.edu.lb/~mk36 if you have MS Explorer -------------------------------------------------------------------------------- ----------------------------------------------
<<attachment: winmail.dat>>