Hocuri
58d0fd39b5
clippy
2025-09-03 17:56:12 +02:00
Hocuri
40e3c34f59
refactor: It's not actually necessary for Alice to remember how the message was encrypted
2025-09-03 17:56:12 +02:00
Hocuri
1377a77ea8
refactor: Use the same decode_name() function for the contact name, remove redundant check for grpid.is_some()
...
If grpid is none, the group/brodacast name isn't used, anyways
2025-09-03 17:56:11 +02:00
Hocuri
db32f1142c
Don't include the broadcast's shared secret in the QR code
2025-09-03 17:56:11 +02:00
Hocuri
738f6c1799
feat: Transfer the broadcast secret in an encrypted message rather than directly in the QR code
2025-09-03 17:56:11 +02:00
Hocuri
e1abaebeb5
WIP, untested: Receiving side of passing broadcast secret in a message
2025-09-03 17:56:11 +02:00
Hocuri
0978a46ab6
WIP, untested: Sending side of transferring the secret in member-added message
2025-09-03 17:56:11 +02:00
Hocuri
410048a9e1
Improve TODOs
2025-09-03 17:56:11 +02:00
Hocuri
72336ebb8a
Add benchmark for message decryption
2025-09-03 17:56:11 +02:00
Hocuri
fca8948e4c
Speed up message decryption by not iterating in the s2k algorithm
...
The passphrase has as much entropy as the session key, so, there is no
point in making the computation slow by iterating.
2025-09-03 17:56:11 +02:00
Hocuri
d431f2ebd3
Add benchmark for message decryption
2025-09-03 17:56:10 +02:00
Hocuri
ad0e3179dd
Remove unused and problematic ensure!
...
`secret_keys.is_empty()` only checked whether any secret keys were
passed. This is not helpful, and made decrypting fail in the benchmark.
2025-09-03 17:56:10 +02:00
Hocuri
494ad63a73
feat: Increase secret size to 256 bits of entropy
...
This is for quantumn computers. When trying to break AES, quantumn
computers give a square-root speedup, i.e. the 144 bits of entropy would
take as many queries as breaking 72 bits of entropy on a normal computer. This neglects
e.g. the costs of quantumn circuits and quantumn error correction [1], so,
144 bits entropy would actually have been fine, but in order to be on
the very safe side and so that noone can complain, let's increase it to
256 bits.
[1]: https://csrc.nist.gov/csrc/media/Events/2024/fifth-pqc-standardization-conference/documents/papers/on-practical-cost-of-grover.pdf
2025-09-03 17:56:10 +02:00
Hocuri
13bbcbeb0e
Add some print statements for debugging
2025-09-03 17:56:10 +02:00
Hocuri
a14b53e3ca
fix: Don't show a weird 'vb-request-with-auth' message when a subscriber joins
2025-09-03 17:56:10 +02:00
Hocuri
9474fbff56
fix: Correct member-added info messages
2025-09-03 17:56:10 +02:00
Hocuri
c4001cc3ff
fix: Let Alice send vb-member-added so that the chat is immediately shown on Bob's device
2025-09-03 17:56:08 +02:00
Hocuri
548f5a454c
Add TODO
2025-09-03 17:55:40 +02:00
Hocuri
91110147c3
fix: Actually send broadcast message to recipients, ALL TESTS PASS NOW - fix test_broadcasts_name_and_avatar().
2025-09-03 17:55:40 +02:00
Hocuri
6012595f1a
test: fix test_encrypt_decrypt_broadcast()
2025-09-03 17:55:40 +02:00
Hocuri
504b2d691d
test: fix test_leave_broadcast
2025-09-03 17:55:40 +02:00
Hocuri
7e191f6cf9
fix: Make joining a channel work with multi-device, fix test_leave_broadcast_multidevice
2025-09-03 17:55:40 +02:00
Hocuri
37f6da1cc9
test: Fix one panic in test_broadcasts_name_and_avatar, but there is another one where I couldn't find the problem
2025-09-03 17:55:39 +02:00
Hocuri
df2693f307
test: Fix test_broadcast_multidev
2025-09-03 17:55:39 +02:00
Hocuri
cdd280a2d3
make test_block_broadcast pass
2025-09-03 17:55:39 +02:00
Hocuri
6bb714a6e5
fix: Make syncing of QR tokens work, make test_sync_broadcast pass
2025-09-03 17:55:39 +02:00
Hocuri
b276eda1a2
Make basic multi-device work on joiner side, fix test_only_minimal_data_are_forwarded
2025-09-03 17:55:39 +02:00
Hocuri
9c747b4cb0
fix: make test_broadcast work, return an error when trying to add manually add a contact to a broadcast list, don't have unpromoted broadcast lists, make basic multi-device, inviter side, work
2025-09-03 17:55:39 +02:00
Hocuri
326deab025
Broadcast-securejoin is working!!
2025-09-03 17:55:39 +02:00
Hocuri
24561cd256
test: Add test_send_avatar_in_securejoin
2025-09-03 17:55:39 +02:00
Hocuri
5da7e45b2b
Adapt the rest of the code to the new QR code type
2025-09-03 17:55:39 +02:00
Hocuri
3389e93820
feat: Add broadcast QR type (todo: documentation)
2025-09-03 17:55:39 +02:00
Hocuri
789b923bb8
feat: Store symmetric key non-redundantly in the database
2025-09-03 17:55:39 +02:00
Hocuri
547f750073
Make it compile
2025-09-03 17:55:39 +02:00
Hocuri
382023de11
sync broadcast secret for multidevice
2025-09-03 17:55:39 +02:00
Hocuri
3781a35989
feat: Add create_broadcast_shared_secret()
2025-09-03 17:55:39 +02:00
Hocuri
8653fdbd8e
feat: Save the secret to encrypt and decrypt messages. Next: Send it in a 'member added' message.
2025-09-03 17:55:38 +02:00
Hocuri
47bf4da1fe
WIP: Start with decryption, and a test for it. Next TODO: SQL table migartion.
2025-09-03 17:55:38 +02:00
Hocuri
ec2056f5e2
feat: Symmetric encryption. No decryption, no sharing of the secret, not tested.
2025-09-03 17:55:35 +02:00
link2xt
7ea6ca35d7
feat: do not replace messages with an error on verification failure
2025-09-02 18:29:53 +00:00
link2xt
a9aad497fc
api!: remove deprecated is_protection_broken
2025-09-02 18:29:53 +00:00
link2xt
7da8489635
api!: remove is_profile_verified APIs
...
UIs now display green checkmark in a profile
if the contact is verified.
Chats with key-contacts cannot become unprotected,
so there is no need to check 1:1 chat.
2025-09-02 18:29:53 +00:00
link2xt
683561374d
test: add TestContext.create_chat_id
2025-09-02 18:29:53 +00:00
link2xt
66c9982822
fix: add "Messages are end-to-end encrypted." to non-protected groups
...
The messages are end-to-end encrypted
in encrypted group regardless
of whether the group is protected or not.
2025-09-02 18:29:53 +00:00
link2xt
1b6450b210
feat: do not set "unknown sender for this chat" error
2025-09-02 18:29:53 +00:00
Nico de Haen
143ba6d5e7
fix: Mark message as failed if it can't be send ( #7143 )
...
resolves #6849
2025-09-02 14:32:25 +02:00
bjoern
6b338a923c
feat: warn for outdated versions after 6 months ( #7144 )
...
3 months were proven to be too short some years ago, after that issue,
we went far up to 12 months.
however, 12 months were considered too long after recent discussions :)
so, 6 months seems to be a good compromise.
the warning is still repeated every months and the text is unchanged.
advantage is still that this approach does not require network or
opt-in, and catches really all lazy updaters with few effort, cmp
https://github.com/deltachat/deltachat-desktop/issues/5422
2025-09-01 17:44:54 +02:00
iequidoo
e6ab1e3df5
fix: Update verifier_id if it's "unknown" and new verifier has known verifier
...
Now that the previous commit avoids creating incorrect reverse verification chains, we can do
this. Sure, existing users' dbs aready have verification chains ending with "unknown" roots, but at
least for new users updating `verifier_id` to a known verifier makes sense.
2025-09-01 05:09:19 -03:00
iequidoo
5da6976bf9
fix: Don't verify contacts by others having an unknown verifier
...
If this happens, mark the contact as verified by an unknown contact instead. This avoids introducing
incorrect reverse chains: if the verifier itself has an unknown verifier, it may be `contact_id`
actually (directly or indirectly) on the other device (which is needed for getting "verified by
unknown contact" in the first place).
2025-09-01 05:09:19 -03:00
iequidoo
bd15d90e77
refactor: Check that verifier is verified in turn
2025-09-01 05:09:19 -03:00
