>===== Original Message From Ron Jeffries <ronjeffries / acm.org> =====
>I was looking for how to do access to passworded pages with http.get,
>and found a reference here. Thanks! The code looks like:
>
>
>enc = 'Basic ' + encode64(username:password) #place real values here
>resp, data = h.get(url, 'Authorization' => enc)

I usually just use urls of the form:

http://name:password / www.site.com/whatever/else.html

This is part of the standard definition of what a URL can
look like, and I have yet to run across a common tool that
does not support it properly.

>Now I have a meta-question: where could this kind of thing be looked
>up?

I have a meta-meta-question.  Once it is decided where this
kind of thing should be looked up, how do we keep it up to
date? :-)

This brings me to another of those projects that I think
would be great to do some day, but I don't see the time for
emerging in my future.  You know those IRC bots that sit on
IRC channels and can be trained to answer simple questions?
They aren't very smart, but they are good for FAQ type
questions, and they are good for pointing people at useful
documents.  They tend to be most useful for people at the
start of their learning curves, and virtually useless once
you have mastered the subject.

What would it take to implement an IRC bot that could be
populated with a merged dump of several others, and then
run as an interactive utility?  The idea is that you install
packages X-Y-Z, then merge in their help databases, and now
you have a command line utility that has a good chance of
answering most of the really basic questions on any of them.
Useful places to take a data dump from would include IRC
channels and help desks.

Cheers,
Ben

-------------------------------------------
The Fastest Browser on Earth now for FREE!!
Download Opera 5 for Windows now! Get it at
http://www.opera.com/download/
-------------------------------------------