Scripts
There will be a major core application, kind of a runbase and the scripts who will run on top of it.
The core app takes care of all communications, security and authentication. The scripts will do the web interaction part.
Scripts have a very easy code structure and will be very easy to mashup. We count on having 3 scripts ready with our first beta public release. One for digg (log in and vote on a story), one for delicious (log in and add a tag) and a surprise one. We were counting on StambleUpon also, but stumbled onto some problems making ours script structure emulate their proprietary toolbar. Further dig script releases will ad simple random comments as well.
A typical script will:
- load url (with or without parameters)
- fill necessary get / post forms for various reasons, including log in.
- click on some random links, wait some random time and exit.
Scripts will also permit simple conditional loops to test estates and/or variables. Capchas will not be a bigger issue, since we have our users login on file (will not fail any login) and we have full access to our user cookie.
Members are encouraged to mashup and develop their own scripts.
However, we are sorry to say this, but we can not leave it completely open because we would most certainly be the perfect target for spammers.
So how can you still contribute? Easy, untrusted scripts will run on a very limited base (either time or number of hits is very limited), but it be more than enough for you to do your tests.
When you fill comfortable your script is ready to be released you just have to send it us, so that we can run our tests on it, hash it and authenticate it. From that day on, everyone would be able to use the script freely. (The unchanged version)
Don't start having some ideas! It's useless to say that tampered and patched base apps will be kicked off the network.
And also useless to say, code base will never be open source. Our team has discussed this a lot and have finally agreed on having it this way. We're sorry but previous bad experiences from similar tools forced us into this. This was born as an open source project, but the risks of having a parallel network of spammers were too big to even consider it a possibility.
I hope you can understand our motivations.