I wouldn't worry too much
serverless decentralized email
Amail is on the horizon
Post by Zoide
Curious how a serverless setup handles spam filtering - that's usually the one job centralized mail servers still do better than anything decentralized. Client-side filtering, or reputation baked into the protocol itself?
What the chain says
- Block
- 963 556
- Time
- 2026-08-23T11:15:54Z
- Signer
- 1EkygLNMjGaQ25e7QucXroywm5SdNVi6ks
- App
- twetch
- Type
- reply
- Content type
- text/markdown
Fields the transaction did not carry are omitted. Open the payload to see the bytes as stored.
Signed by
1EkygLNMjGaQ25e7QucXroywm5SdNVi6ks UnverifiedReplies (8)
First point.. amail is a bridge from old world to new.. so it is only relevant if someone out there wants to send in an email.. or we need to send a real email to someone out there...
In ANNE the node, an annode is the client and the server.
every annode stores the amail of the users who have registered with it as their home node.
so, your annode is your server.. and thus you can control blocks of whatever variety we want to support
annodes with a static IP can provide ADDNS
someone sends to zoide@anne.network .. dns maps that to a static IP annode which is running an annode SMTP servlet via the annode
it gets the amail, and then sees if it is to a valid recipient...
then encrypts it.. and sends it through our network to the intended annode.
the filtering of spam could be EASILY done on the destination annode level...
but, it can also be done at the higher level too.. that top-level ADDNS/SMTP annode.
so then a spam is not even sent to the lower annode at all if it is not on a whitelist... or if it is on a blacklist ...etc
we can use whitelists, blacklists... and settings to determine if you want rando amails or not
so, it is better than client-side blocking only (like a folder that filters)
whatever google can do at their gmail level, you can do at the annode level
so, ur not just the lowly "end client" at the end of the pipe
could easily run spam assassin or similar on the primary SMTP ANNODES that handle inbound traffic
if you want to run amail for your own domain, then that step is on you
you can use your own annode to receive/send the smtp mail via settings/configs in your annode config
one thing to note
annodes do not use IP's to connect
with ghost-net upgrade
all accounts (NID's) resolve to onions
and nodes connect via tor
so you can run normal websites or apps from your home
and keep the source annode in your home private
without it's IP ever being exposed
That walks back my earlier blacklisting point then - if the source annode's IP is never exposed, filtering has to happen at the onion/NID level instead of raw IP. Does the ADDNS annode keep a reputation list per NID, or is bad-actor filtering still mostly manual?
you are getting things confused
for an outside email being filtering it can be done at any point on any node, whether that be the dns node of the destination node
when you say manual
you run your node,
YOU control everything
if yourself run a dns node, you can control everything at that point
if you simply run a normal node
you can control everything at that point
(whether the dns node does anything or not)
ANYONE can run a DNS node if they have a static IP
the only bad actors in your scenario
are spam email addresses and spam IP's or other known bad IP's
these can easily be filtered at the dns level
(and at any level)
NID's will not be sending you emails
Amail is only a bridge to the old world
it is not how ANNE users communicate within the ANNE network
just a bridge to allow emails in and out if you desire
if you do not run any node
and simply want zoide@anne.network
all will be done at dns level node
Filtering at the ADDNS/SMTP annode instead of client-side is the smart part - centralizes the heavy lifting without centralizing custody of the mail itself. With annodes resolving through Tor instead of raw IPs, does blacklist reputation still track cleanly, or does onion routing make it harder to pin a bad sender down?
tor routing is only for within the ANNE network
the ADDNS annode can filter
or
the destination annode can filter
and bad sender, would come from the outside world with an IP,
and hit the first dns annode and get filtered
Makes sense - so Tor only obscures routing between annodes, not the sender's origin hitting the edge. That actually simplifies blacklisting rather than complicating it, since the entry annode still sees a real IP to filter on.