Hashcash.IO Updates

About spam, invites and dogecoins

First of all – thank you all for so much of feedback!┬áSeeing so much interest in what I am doing is truly inspiring!

I received quite a few emails asking about different aspects of this project and I would like to share my answers and future plans in this post.

So what about invites?

Let me get back a little bit. I have a few content-based projects out there and have always struggled with spam. Even the famous and sometimes hard to read reCAPTCHA did not always help. From what I observed so far, it looks like posting spam is semi-automated nowadays. I even found a website – antigate.com – which provides an API to real humans reading CAPTCHAs. Think mechanical turk, but just for reading captchas (and way cheaper.)

On the one hand, dealing with reCAPTCHA when Google did not recognize you is terrible – it often requires 3-5 attempts to finally decipher what is written. When Google ‘recognizes’ you – it is much easier – they just pretend it is CAPTCHA when in reality they use you to solve their business problem – decoding street numbers – you can read more about it on Techcrunch. So for human spammers it becomes very easy to do their job.

Anyway, I really disliked CAPTCHAs, all of them, and especially reCAPTCHA. And after finding out about antigate.com I even had an idea of creating service where visitors will be presented with CAPTCHAs taken from antigate.com and submit results back to antigate.com to get paid :) And about that time I learned about bitcoin, hashcash.org and proof-of-work concept.

My idea was to bring what hashcash.org offered for email spam to web applications. I.e. make it slightly expensive for a browser to submit an actual form by calculating proof of work. Except I wanted this proof of work to do something meaningful. I entertained idea of mining bitcoins with browsers, but by that time GPUs already took over it and in-browser mining wasn’t an option anymore. Later a new altcoin was introduced – Litecoin based on scrypt algorithm, which was harder to parallelize on the GPU. That’s where I started looking into implementing miner into client-side Javascript.

Long story short – I used emscripten to compile scrypt version of cpuminer into asm.js code, wrap in Web Workers and some API, and here we are – the very first prototype of hashcash.io. It had no front page, no login form, just rudimentary D3-based dashboard to easily see what is going on. After trying it out on few of my websites 100% of spam and bot-registered accounts were eliminated. After that I spent some time customizing Bootstrap theme, adding some examples and key generation form on the home page of hashcash.io, as well as adding login form to make sure only I had access to the crappy dashboard for now :) That’s where the project stands right now.

So to answer the question about invites – there is still no reason to have one – there is nothing valuable behind login form yet.

Dogecoin

Litecoin is boring. Dogecoin is fun. At some point I switched proof-of-work to one used in dogecoin network using p2pool-rav. As of right now an average mining speed of the whole pool is probably 20 hashes/sec, which is nowhere enough to generate a single share. Which leads me to the next topic:

Get paid for visitors/bots solving proof-of-work

So the grand vision of this project is to eventually make it possible for a website admin to get paid for each proof-of-work solved on his site. So even if spammers eventually start solving these, at least website admin will get paid. But for this to happen the pool needs to acquire a critical mass of websites using it to successfully generate shares in a reasonable amount of time. Right now this is not possible. And this is another reason there is nothing behind login form – there are no revenue as of now to share with website admins.

If you think this concept is cool, please spread the word about it. Right now I am trying to support painless integration with Drupal, WordPress CMSes via plugins, there is Django plugin and there is also jQuery plugin for custom integration. I need more feedback from developers out there. And I need more sites using it to generate shares.

As soon as the project generates a single dogecoin at address DMGQ5Ah5D7FSBL2uKiugwHQneGdugnvZfP, I will start working on the next phase – dashboard and revenue sharing.

Thank you very much for your attention. And if you have any questions – you can reach me at [email protected]

One thought on “Hashcash.IO Updates”

  1. The idea seems pure genius, if browser mining can generate anything, of course. Wonder of other mining algorithms would work better? Groestl seems the most effective when CPU mining, too bad the coin itself is worthless.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>