h a l f b a k e r yBreakfast of runners-up.
add, search, annotate, link, view, overview, recent, by name, random
news, help, about, links, report a problem
browse anonymously,
or get an account
and write.
register,
|
|
|
When you subscribe you add (in confidence) the email addresses of people that you know and the user ids of those people you know that already subscribe to the site.
If someone sends on this list sends you an email it appears at the top of your email list in the 1 degree of separation section. If
someone on the list of one of the people on your list send you an email it appears below that in the 2 degrees of separation section (and so on).
Anybody not connected to you appears in the unsorted list below the sorted emails.
Of course, while I'm dreaming you should have a 'generally considered to be spam' bucket which you can move spam to with the click of a button. All emails sent to any user from a 'generally considered to be spam' email address will go to the spam bucket unless they appear on 'friend' lists.
[If this idea gets interest and if you're very very lucky I may post the 'Degrees of separation plumber and builder recommendation guide'. But only if you're very very lucky.]
(?) Procmail Scoring
http://fsinfo.cs.un...man/procmailsc.html
[prometheus, Dec 13 2001, last modified Oct 05 2004]
Mutt Scoring
http://www.mutt.org...anual-3.html#ss3.22
[prometheus, Dec 13 2001, last modified Oct 05 2004]
Bayesian prioritization research article
http://www.nytimes....articles/17lab.html
Eric Horvitz is thinking somewhat along your lines, st3f. [bristolz, Dec 13 2001, last modified Oct 05 2004]
PGP web of trust
http://www.rubin.ch.../weboftrust.en.html
written for the smart newbie [hello_c, Dec 13 2001, last modified Oct 05 2004]
(?) Webs of trust in general
http://www.firstmon...ues/issue3_6/khare/
Much more serious & general article. [hello_c, Dec 13 2001, last modified Oct 05 2004]
US Patent 9,100,221: Systems for messaging senders and recipients of an electronic message
https://www.google.com/patents/US9100221
the patent that cites this idea [notexactly, Nov 23 2015]
Please log in.
If you're not logged in,
you can see what this page
looks like, but you will
not be able to add anything.
Destination URL.
E.g., https://www.coffee.com/
Description (displayed with the short name and URL.)
|
| |
This doesn't help solve the problem that some of the worst crap I get sent is from people I know pretty well. But I like the principle of it prioritising mail from friends. Nice. |
|
| |
Fine if we share an ISP/web mail service, but otherwise very nasty. That would require all email servers to communicate among each other to check where a sender comes in your given list. If I send you a message, your mail server could check my name against your list, and if it finds it, then that's simple, flag it as one degree of separation. But what if it doesn't find it? Does your mail server (or email software) then have to contact everyone on your friends list to see if I'm included? And if none of them find my name, do they have to contact everyone on *their* list? |
|
| |
There are ways to manage the distribution of the list, involving a protocol for mail servers to notify each others of changes in people's lists and keep track of who is a second or third degree from you, but the data traffic required would be voluminous. |
|
| |
Of course we all could use the same mail provider, but that seems unlikely because many companies won't surrender power over their email easily. This could work in a limited way (e.g. on *shudder* AOL), but as a whole-internet solution, not. |
|
| |
Pretty neat idea. How would this work in conjunction
with other sorting methods (rules and such)? |
|
| |
[pottedstu] Couldn't yr problem be at least partially solved by examining which, if any, email addresses are CC'd in any mails received. Admittedly this would only work with the kind of situation where someone decides to send/forward something to several people in their address book, but as most people who send me mail do this on some occasions, it seems workable. |
|
| |
pottedstu, alx: Don't worry about large scale implementations of this system. As this data is worth large anount of money in the marketing community it's dangerous to share it between mail servers - this would be a one domain operation only. The filtering would work for one degree of separation beyond the people who had accounts here. |
|
| |
PotatoStew: Rules? Where we're going we don't needs rules. Sorry. Implement as necessary. e.g. In my main inbox show all bits of mail which are less than 5 degrees of separation. |
|
| |
Other possibilities:
Several lists of people - friends, business, listservers that I occasionally read could go to different inboxes on the same account.
You could put a confidence percentage on people based on how likely they will have bulk emailers in their friends list. You could then sort by percentage rater that how many degrees of separation. This is my imagination. You can do anything. |
|
| |
Too complicated. One folder for spam, one folder from
personal addies not rendered to be from a spam server
(whatever) is fine for me and most people, I think. Blah @
spam anyway. |
|
| |
It's not just about spam. It's about overload, too (link, horvitz). |
|
| |
Sign your mail & use the PGP web of trust? |
|
| |
The plumber thing sounds good though. Find a plumber recommended by your friend, or failing that by a friend of a friend, or failing that by a friend of a friend of a... |
|
| |
Anything beyond 2 degrees is indistinguishable to me. |
|
| |
Mostly baked I think - Homail has an junk mail protection option (called 'exclusive') whereby any email originating from someone who's address is not in your address list (or 'safe' list) goes into the junk folder. ie it's a 1st degree filter. |
|
| |
Now, if it kept track of the cc addresses on those non-junk messages and treated them as non-junk too (-alx's idea) it would almost be a 2nd degree filter. |
|
| |
The granted patent application 9,100,221 cites this HB idea as prior art. |
|
| |
Wow. That's the most citations I've ever seen in one place,
by a large margin. 397 (by my count) other patents, and 318
non-patent citations including this idea. |
|
| |
Edit: Actually 382 patent citations, not 397. I used a more
accurate counting method this time. Conveniently, the total
is now 700. |
|
| |
Wow... I thought you were joking until I read the patent! |
|
| |