[hashcash] hashcash idea

I haven't been deeply into the conversation, so sorry if my answer is
totally out of context!

If I understand, we make a kind of stamp that we can use into a message, but
we don't build the stamp based on the message.

Wouldn't it be nicer if we calculate the MD5 of (the title, sender, receiver
and mime parts of the message)?
Then we choose how much time we want to spend to calculate the HashCash.
(With someone I really like, I will spend 1 minute of processing power to
calculate my HashCash (So a faster computer is better, but remove the
problem of computer that has more and more power).)
Based on this md5 we create a HashCash token that we add in the header of
the mail.

We can include this processing in many open source mail processors. (Ex:
Mozilla)
It will then be used by spam inspector and others spam mails processor as an
alternative way to detect if the mail is a spam.

If we send this mail through a web interface (hotmail), we can provide a
java applet that calculates the HashCash. (It can even be done directly in
some clients that support it.).
It can also be a way to sign other messages, like forums, website creation
so (replace the "enter this number" that we can see in a lot of site.)

Best regards, sorry if this mail is seen as a spam by some people ;-)
Gab

-----Original Message-----
From: hashcash-bounce@xxxxxxxxxxxxx [mailto:hashcash-bounce@xxxxxxxxxxxxx]
On Behalf Of Martin Hans
Sent: vendredi, 19. mars 2004 11:17
To: hashcash@xxxxxxxxxxxxx
Subject: [hashcash] hashcash stamp server

Hi,
I have been quietly following the discussions on this list for a few weeks
and 
have been wondering how to integrate hashcash with my favorite MUA, kmail.

What I think we need is a stamp server, that runs locally or on a remote 
machine and where the MUA can order stamps. I think the following commands 
should be in the protocol:

ORDER ressource
means "the user is writing an email message to <ressource> right now, so a 
stamp should be put into production. Returns a <stamp-id>.

GET <stamp-id>|<ressource>
means "I need a stamp for ressource or the stamp I ordered earlier right now

because the email is ready to be sent." Blocks until the stamp is ready and 
then returns it.

Client libraries for something like this would be easy to write and would
make 
it possible to integrate in any open source email client.

For having stamp management like the one discussed in this thread: 
http://news.gmane.org/find-root.php?message_id=%3c20040306205338.GB3028%
40longshot.toehold.com%3e we would also need the server to maintain a list
of 
the most often needed stamps. The stamp server could prioritize in the 
following order:

1. answer GET requests
2. process stamps requested by ORDER
3. make sure we have one stamp for today for the top-n of email recepients
4. make sure we have two stamps for today for the top-n of email recepients
5. etc.
6. make sure we have one stamp for tomorrow for the top-n of email
recepients
7. make sure we have two stamps for tomorrow for the top-n of email
recepients
8. etc.

Maybe there should also be a way to tell the stamp server that an email has 
been received that the user might want to send an answer to soon. But with 
the amount of emails many of us receive, this would maybe be too much.

Is there a way yet to announce to someone that he has been whitelisted and 
does not need to produce stamps for me anymore.

Just my to cents.

Martin

Other related posts: