First I should mention that I don't encourage anyone to do/try this.
This is a theoretical question. (Besides, it requires skills that most
people don't possess.)
Let's say Web master Joe Doe has a crappy, spammy site without any
useful content and filled with ads. He wants to boost his PR, but
obviously he is having a hard time to get backlinks. So he decides to
write a program to function much like a search engine crawler, only
whenever it finds a form, he submits the URL to his Web site to it and
continues.
Basically:
1) Start the program at some large site with many links to many
different sites.
2) Make it note all URLs on the first page.
3) Go to all these pages in order. Put all visited URLs in a list, so
that it only visits each once. (Preferaly store this in a database.)
4) Regexp all URLs on this new page too.
5) Go to 3.
Also, of course, whenever it finds a <form> element on any page, it
stops for a while and sends the information to it.
This wouldn't be trivial to code, but surely not impossible. The most
tricky part as I see it is to know what to send in forms; some forms
require you to enter a username, a full name, a telephone number and/or
an e-mail address. Most forms also check these values server-side.
Now, let's assume that Joe manages to code together a working solution.
It will check for common form elements and input dummy values such as
"Joe", "Chris", etc. in the "name" field, "random@example.com" in
e-mail, etc., and naturally his spam message in the first textarea,
which is very likely to be the content to be added (guest book,
anonymous comments on a blog, etc.).
Joe is proud of his work and lets the program run for a few weeks on
his Unix server at home (not using proxies). Whenever it is out of URLs
(if that can happen), it stops and notifies him so that he can add a
new site for the program to start from.
What do you think will be the results of this experiment, assuming his
program isn't fundamentally broken?
1) Thanks to CAPTCHAs (those bastards!), "registered users only"
policies and other obstacles, Joe's bot doesn't gain him many
backlinks. It was mostly wasted effort.
... OR...
2) Due to the complexity of the WWW, and the zillions of poorly written
sites out there, Joe gets hundreds of thousands of backlinks, each
boosting his PR minimally, but together giving him several PR points
(Google).
So... What do you think? :)
