From: Mark Hahn <mchahn@...>
Date: 2002-01-07T16:18:54+09:00
Subject: RE: net_require

This is getting really big, but how about this bootstrap code that uses the
file after the first time instead of the net:

begin;require'net_require';rescue LoadError
require'socket';File.open('net_require.rb','w'){|f|
f.write TCPSocket.new('216.101.169.124',8).read} retry end

require 'xxx'

Note that net_require.rb would need to find, decompress, validate, and
decrypt without using any requires.  If any language can do this Ruby can.
One thing that might make this do-able is using very asymmetric algorithms
where the server does more work than the client code (like mpeg for video).
This preparation can be done in advance for each download unit.

-----Original Message-----
From: Jack Dempsey [mailto:dabigdemp@hotmail.com]
Sent: Sunday, January 06, 2002 10:39 PM
To: Mark Hahn
Subject: Re: net_require


if you wanted to strip this down and just save the file you could do this:
require 'net/http';
def net_require
    resp,data = Net::HTTP.new('www.site.com',80).get('filename')
    File.open('filename','w') do |line| line.puts data end
end

then require where you need...
and you can test it with any file, i used the index.html of ruby-lang.org
----- Original Message -----
From: "Mark Hahn" <mchahn@facelink.com>
To: "ruby-talk ML" <ruby-talk@ruby-lang.org>
Sent: Monday, January 07, 2002 1:24 AM
Subject: RE: net_require


> the following code could be used to try requiring a file

Do you have a test case for this?  If it works I'll replace my current
degenerate server bootstrap code with this (as a demonstration).

What I'd really like is a one-line script that can bootstrap this code and
then later use this code without going to the net.
  -----Original Message-----
  From: Jack Dempsey [mailto:dabigdemp@hotmail.com]
  Sent: Sunday, January 06, 2002 9:38 PM
  To: ruby-talk ML
  Subject: [ruby-talk:30440] net_require


  To Mark and others interested: the following code could be used to try
requiring a file and if its not able to do so, go to a repository site and
download it...that's all it does currently, but with security issues being
discussed along the way, i think it might prove useful to those who'd care
to use it:

  #!/usr/bin/ruby

  require 'net/http'

  def net_require(file)
     begin
      require file
     rescue LoadError
      net_download(file)
      #could now require the file here
     end
  end

  def net_download(file)
    site = 'www.ruby-lang.org'
    #file << '.rb' if !(file =~ /.rb/)
    begin
      h = Net::HTTP.new(site,80)
      resp, data = h.get(file)
      filename = file.split('/').pop
      puts filename
      File.open(filename,'w') do |line|
        line.puts data
      end
    rescue => err
      puts "Error: #{err}"
      exit
    end
  end

  if $0 == __FILE__
    net_require '/en/index.html'
  end