While testing this bug report in Gwibber
https://bugs.launchpad.net/gwibber/+bug/397297 I found that using
urllib2.Request() with a long authorization would cause it to error and
not authenticate. I'm not sure if it is because the header wasn't being
passed or any other issue.

The header would fail when the value part of the authorization string
was 89 characters.

I was able to resolve the issue by using urllib2.HTTPBasicAuthHandler()
and urllib2.install_opener() to install it, instead of passing the
Authorization header directly to urllib2.Request().

Please let me know if I can provide more information to test the issue.

There is a OLD Known issue that base64.encodestring will insert in a line-break in the authorization header. See Issue1574068
I fear, you might have done something like this.

base64string = base64.encodestring('%s:%s' %(username,passwd))[:-1]
request.add_header("Authorization","Basic %s" % base64string)
urllib2.urlopen(request)

You might have encountered this problem of Request not working with long authorization headers.
Instead, if you do

base64string = base64.b64encode('%s:%s' %(username,passwd))[:-1]

and then add the string to your headers, it might work properly. It does so for me.

(alternatively you can set base64.MAXBINSIZE=1000000, but the first solution is much cleaner.

I am closing this bug as Invalid. If you feel, it is valid and you used base64.b64encode for encoding the Basic Authentication header. feel free to reopen the bug, with a snippet of code attached.