http_tail - tail -f inside your browser

a.k.a. A cautionary tale about how you should think about the absurdity of
       the problem before hacking away

-

Overview

  http_tail aims to solve a simple problem: you want to monitor a log and
  for some arcane reason, someone (who could be you) can SSH to the server and start
  a daemon while you cannot.

  See? Absurd. But hey, "browser-based tail -f"! Also, you can bookmark
  the address so you can monitor the log whenever - and from wherever - you are!
  Actually, "showing the log to someone who can't ssh to your box" seems
  like a good use case. Who knows.

How to use

  To start the service, run:

    python http_tail.py <file> <host> <port>
 
    where:
    
    <file> is the path to the file that will be watched
    <host> is the host name. IT MUST BE THE SAME AS THE ONE YOU'LL USE TO ACCESS
           THE SERVICE, BECAUSE THE VIEW PAGE USES XMLHTTPREQUEST AND IT DOES
           NOT ALLOW CROSS-ORIGIN REQUESTS. BE WARNED.
    <port> the port the service will listen on
  
  After starting the service, point your browser to:

    http://<host>:<port>/view

  Type the file name in the input box and click "Start tail -f"

  e.g:

    python http_tail.py http_tail.py 127.0.0.1 8000

  Starts the service that can be accessed at:

    http://127.0.0.1:8000/view

  Have fun! (I guess)

Known issues:

  - Only updates that add new lines are detected. The server and the view must be updated to
    include the number of columns of the last line
  - Python's BaseHTTPServer isn't exactly a performance workhorse
  - Autoscroll does not work with Firefox
  - This has not been tested on IE, Safari and Opera. Chances are I'll never even try to
    test it on IE. If you still use IE, YOU have an issue