What is it?
-----------

I read about 30 to 40 comics daily from the net. That is along with
the comics in the newspaper. I wrote an Emacs lisp program to download
them all for convenient viewing.

However, when I go away on vacation, I wanted something to download
the comics daily and save them. I had problems getting Emacs to
startup and run reliably, so I quickly wrote the first get-comics.


First get-comics
----------------

The first get-comics was a multi-threaded app that was driven by an
xml file to download comics. I wrote it multi-threaded because I
needed it fast. This was basically a prototype. It worked, but it
tended to swamp the proxy server I used at the time.


get-comics 2
------------

The second version of get-comics is a single threaded version. Rather
than spawning a thread per connection, it handles all the connections
in a single thread using the `select' call. I find this version much
more reliable than the multi-threaded version. It also has lower
resource usage.

This is the current version.

get-comics 3
------------

The third version of get-comics used libcurl. I found that libcurl
hangs too often. Since the second version has been reliable for years,
I basically dropped this version in favour of the second version.


JSON Parsers
------------

I was using JSON at work and much preferred it to XML. I decided to
convert get-comics to use JSON.

get-comics went through three JSON parsers. I started with JSON
lib. But it was an extra library dependency which is a pain on distros
like Ubuntu that do not provide include files by default.

So I decided to go with a source code version. I only needed a parser,
not a full JSON library. I decided on js0n. I really like js0n. It is
very small and fits nicely with the comics layout. Unfortunately, it
is not portable to windows.

So I finally decided on JSON_parser.c. It is portable. And it has the
big plus that it tells you the exact character in the file the syntax
error was on.


Documentation
-------------

The documentation for get-comics is either in man format
(get-comics.1) or an HTML version (get-comics.html)

There is a sample JSON file that I use: comics.json.