[hashcash] Re: Some of my thoughts
- From: "Eric S. Johansson" <esj@xxxxxxxxxx>
- To: hashcash@xxxxxxxxxxxxx
- Date: Fri, 28 Jul 2006 10:53:12 -0400
Mate Soos wrote:
2) Concerning reputation: I wrote my Diploma thesis on trust model
...
my thoughts around hash cash and reputation are fairly simple.
Reputation is something that is earned to trust. By repeated exposure,
you can develop a sense of how trustworthy a data source is. For
example, in a Web 2.0 application, if someone keep submitting articles
or notes that are generally voted as unacceptable, the reputation of
that entity falls. You can communicate that reputation via increasingly
large stamp requirements before submitting. This slows down the rate of
submissions and if you make it a foreground process, you can clearly
communicate to a human that they have a bad reputation.
as you can see, Hash cash does not describe reputation, it is a proxy
for the value of reputation. This feedback property can be used in many
circumstances. In the context of anti-spam, you can use it in place of
the blacklist by making it increasingly difficult to get through a
content filter.
At the same time, you can use reputation to reward. Faster access,
easier passage. Again, in the antispam context, you can reduce the cost
of postage to sites that are clean.
not perfect but good enough to try.
3) I personally believe that the idea of hashcash is brilliant. It is
...
stamp), but failed, I am not a really good programmer. At least I tried.
hey, at least you tried. also it may not be your fault. tb plugins are
*hard* to get right.
4) Regarding distributed and DNS-based distribution of the
hardness(lenght of 0's) of the stamp: Doing it the DNS way is simple.
Once hashcash gains momentum, there will be other people who solve this
problem (companies, maybe, or just commited individuals). I do not think
we should be concerned about this. If people realise that hashcash has a
huge potential, they will solve this problem.
it turns out the stamp server by name probably isn't all that difficult.
I just need to figure out how I'm going to store the per user data
(again). But fetching it and presenting it, that's not difficult. But
yes, knowing the DNS route is the shortest path and easiest.
Unfortunately, making the hop to the next generation may be difficult
and I'm not sure I want to put that hurdle in place.
I know that what I wrote sounds harsh. However, I believe there should
be groups of people doing individual tasks, not one group doing all
tasks. Our mission is to propagate hashcash and show its viability. Once
it is shown to be working, others will take care of the rest.
so, know Python?
---eric
Other related posts: