Clarify some more things

README.md

@@ -1,19 +1,20 @@
 # WaybackProxy
 
-WaybackProxy is a HTTP proxy that sends all requests through the [Internet Archive Wayback Machine](http://web.archive.org) and [OoCities](http://www.oocities.org), returning the original retro-browser-friendly markup.
+WaybackProxy is a retro-friendly HTTP proxy which retrieves pages from the [Internet Archive Wayback Machine](http://web.archive.org) or [OoCities](http://www.oocities.org) and delivers them in their original form, without toolbars, scripts and other extraneous content that may confuse retro browsers.
 
 ![1999 Google viewed on Internet Explorer 4.0 on Windows 95](http://i.imgur.com/tXsLc6O.png)
 
 ## Setup
 
 1. Edit `config.py` to your liking
-2. Start `waybackproxy.py`
+2. Start `waybackproxy.py` (Python 3 is required)
 3. Set up your retro browser:
 	* If your browser supports proxy auto-configuration, set the auto-configuration URL to `http://ip:port/proxy.pac` where `ip` is the IP of the system running WaybackProxy and `port` is the proxy's port (8888 by default).
 	* If proxy auto-configuration is not supported or fails to work, set the browser to use an HTTP proxy at that IP and port instead.
-	* Transparent proxying is also supported for advanced users. No configuration to WaybackProxy itself is required. In transparent mode, client machines must be pointed at a dummy DNS server so they can find the proxy; `dnsmasq -A "/#/ip"` is a good choice.
+	* Transparent proxying is also supported for advanced users, with no configuration to WaybackProxy itself required.
+		* The easiest way to set up a transparent WaybackProxy is to run it on port 80 ([this cannot be done on Linux without security implications](https://unix.stackexchange.com/questions/87348/capabilities-for-a-script-on-linux)\), set up a fake DNS server - such as `dnsmasq -A "/#/ip"` where `ip` is the IP of the system running WaybackProxy - to redirect all requests to the proxy, and point client machines at that DNS server.
 4. Try it out! You can edit most settings that are in `config.py` by browsing to http://web.archive.org while on the proxy, although you must edit `config.py` to make them permanent.
-5. Press Ctrl+C to stop
+5. Press Ctrl+C to stop the proxy
 
 ## Known issues and limitations
 
@@ -23,10 +24,11 @@ WaybackProxy is a HTTP proxy that sends all requests through the [Internet Archi
   * Strange 404 errors caused by bad server responses or incorrect URL capitalization at archival time;
   * Infinite redirect loops;
   * Server errors when it's having a bad day.
-* WaybackProxy will work around some redirection scripts (example: `http://example.com/redirect?to=http://...`) which are not archived by the Wayback Machine, but the destination URLs might not be archived as well.
+* WaybackProxy will work around some redirection scripts (example: `http://example.com/redirect?to=http://...`) which are not archived by the Wayback Machine, but the destination URLs are sometimes not archived either.
 * WaybackProxy is not a generic proxy. The POST and CONNECT methods are not implemented.
+* Transparent proxying mode requires HTTP/1.1 and therefore cannot be used with some really old (pre-1996) browsers. Use standard mode with such browsers.
 
 ## Other links
 
-* [Donate to the Internet Archive](https://archive.org/donate/), they need your help to keep the Wayback Machine and its petabytes upon petabytes of data available to everyone at no cost.
-* [Check out 86Box](https://86box.github.io/), the emulator I use for testing WaybackProxy on older browsers.
+* [Donate to the Internet Archive](https://archive.org/donate/), they need your help to keep the Wayback Machine and its petabytes upon petabytes of data available to everyone for free with no ads.
+* [Check out 86Box](https://86box.github.io/), the emulator I use for testing WaybackProxy on older browsers.

config.py

@@ -1,8 +1,9 @@
-# Listen port for the HTTP proxy
+# Listen port for the HTTP proxy.
 LISTEN_PORT = 8888
 
-# Date to get pages from Wayback (YYYY, YYYYMM or YYYYMMDD)
-DATE = '1998'
+# Date to get pages from Wayback. YYYYMMDD, YYYYMM and YYYY formats are
+# accepted, the more specific the better.
+DATE = '20011025' # <- Windows XP release date in case you're wondering
 
 # Allow the client to load pages and assets up to X days after DATE.
 # Set to None to disable this restriction.
@@ -14,8 +15,8 @@ GEOCITIES_FIX = True
 # Use the Wayback-tampered URL as a shortcut when loading images.
 # May result in faster loads, but all images will point to 
 # http://web.archive.org/... as a result. Set this value to 2 to enable an
-# experimental mode using username:passwords on top of the original URLs
-# instead (not supported by Internet Explorer and some other browsers).
+# experimental mode using authentication on top of the original URLs instead
+# (which is not supported by Internet Explorer and some other browsers).
 QUICK_IMAGES = True
 
 # Allow the Content-Type header to contain an encoding. Some old browsers
@@ -23,5 +24,5 @@ QUICK_IMAGES = True
 # False if you're using one of them.
 CONTENT_TYPE_ENCODING = True
 
-# Don't print log entries
+# Disables logging if set to True.
 SILENT = False

waybackproxy.py

@@ -131,9 +131,9 @@ class Handler(socketserver.BaseRequestHandler):
 
 			if e.code in (403, 404, 412):
 				# 403, 404 or tolerance exceeded => heuristically determine the static URL for some redirect scripts
-				match = re.search('''[^/]/((?:http(?:%3A|:)(?:%2F|/)|www(?:[0-9]+)?\.(?:[^/%]+))(?:%2F|/).+)''', archived_url, re.IGNORECASE)
+				match = re.search('''[^/]/((?:http(?:%3A|:)(?:%2F|/)|www(?:[0-9]+)?\.(?:[^/%]+))(?:%2F|/).+)''', archived_url, re.I)
 				if not match:
-					match = re.search('''(?:\?|&)(?:[^=]+)=((?:http(?:%3A|:)(?:%2F|/)|www(?:[0-9]+)?\.(?:[^/%]+))?(?:%2F|/)[^&]+)''', archived_url, re.IGNORECASE)
+					match = re.search('''(?:\?|&)(?:[^=]+)=((?:http(?:%3A|:)(?:%2F|/)|www(?:[0-9]+)?\.(?:[^/%]+))?(?:%2F|/)[^&]+)''', archived_url, re.I)
 				if match:
 					print(match.groups())
 					# we found it
@@ -260,7 +260,7 @@ class Handler(socketserver.BaseRequestHandler):
 					# date codes. This taints the HTML with web.archive.org
 					# URLs. QUICK_IMAGES=2 uses the original URLs with an added
 					# username:password, which taints less but is not supported
-					# by all browsers - IE6 notably kills the whole page if it
+					# by all browsers - IE notably kills the whole page if it
 					# sees an iframe pointing to an invalid URL.
 					data = re.sub(b'(?:(?:http(?:s)?:)?//web.archive.org)?/web/([0-9]+)([a-z]+_)/([^:]+)://',
 						QUICK_IMAGES == 2 and b'\\3://\\1:\\2@' or b'http://web.archive.org/web/\\1\\2/\\3://', data)