-
Notifications
You must be signed in to change notification settings - Fork 6
/
TODO
executable file
·69 lines (61 loc) · 3.13 KB
/
TODO
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
TODO
- Way More of REMAIL
- replace templates/help.msg with something not stolen from dizum,
and update it
- figure out a way to refresh mlist.txt, rlist.txt and pubring.{mix,asc}
- Let setup.py be rerun without having to git checkout the modified files
- In mixKey.decode we assume a version of 2 if it's not provided. This is...
probably ok.
- mixKey.decode doesn't know the difference between CNm and CNp modes in the
pubring.mix string. Neither does the author.
- In mixKeystore loadKeyRing there's an inconsistency where if we load a
private key but never the public key we don't have the shortname and other
properties. It'd be nice to throw an exception on access of those in that
case...
- mixconfig.getCapString may not be accurate
- Can't have a single quote in your passphrase for the key.
- Untested Structure Sections:
PacketType.FinalHopPartialMessage
UserDataType.Gzip
UserDataType.Email
- Tons of Unimplemented Features:
Pool Delays
Partial Messages / Message Fragments
Parsing Dummy Messages
Posting to Newsgroups
Doesn't reply with an explanation if it recieves a non-mix email
- Ensure remailer talks TLS over the SMTP links whenever possible
- Normalize logging. Make it a option chosen on setup
NSA logging - log all messages in and out
Debug logging - log messages that cause errors
Minimal logging - log almost nothing
- Right now any message with a destination of null: gets ignored. No one
would mix a legit and a dummy recipient together... right?
- lamson has very, very rudimentary support for skipping over .onion MX records
(because of frell). It'd be nice to fully support delivering to .onion
- destination-block file writing is not Thread-safe.
- Message Database is not Thread-safe.
- Message Database does not expire old messages, although it has the
information to do so.
- Message Database could be made more efficient with a tree structure or
something
- Maximum Message Size
- Allow configuration as a middleman node that whitelists addresses
- If in middleman mode, and recieve a final hop message, pick a remailer
to random-hop it to.
- Make header block use regexes, and copy Dizum's
LONG TERM REMAILER REFRESHES
These may break older mixes. They should only be done if there are serious
advantages of keeping the old protocol mostly-intact instead of moving to a
standard encrypted-wire-protocol of OpenPGP
- Newer Cryptographic Algorithms
RSA 2048 or ECC instead of RSA 1024
AES-256 instead of TDES
- Update its assumption that it is the only mixmaster implementation
rem.c : 175 : bufileft(content, "mixmaster")
- mixmaster seems to use the Message ID as a timestamp, which is inconsistent
with the spec. mixfaster treats it as truly random.
- Give the option to send a confirmation message for destination-block address
- Require verification for destination-block addresses added to the list
- Have a shared database for all remailers of destination-block addresses.
Hash them to provide anonymity.