In 0.9.x we fixed an issue, because OMEMO devices should be defined in "item"
with id "current". This should work, but it won't work if there is no "current".
If there is no "current" we will just use the first item.
Issue #1384
Profanity requests the device list and keys for contacts, also when the user
hasn't generated the OMEMO key. If the user has no OMEMO key, there is no need
to request OMEMO information.
Issue: #1332
We call omemo_init() when starting profanity and should have an
omemo_close() at exit.
For now we free the fingerprint autocompleter in there.
Fixes valgrind:
```
==13226== 24 bytes in 1 blocks are definitely lost in loss record 2,855
of 6,958
==13226== at 0x483677F: malloc (in
/usr/lib64/valgrind/vgpreload_memcheck-amd64-linux.so)
==13226== by 0x48AD39: autocomplete_new (autocomplete.c:57)
==13226== by 0x4AB89F: omemo_init (omemo.c:127)
==13226== by 0x42C283: _init (profanity.c:206)
==13226== by 0x42BFF3: prof_run (profanity.c:98)
==13226== by 0x4B25E6: main (main.c:172)
```
Regards https://github.com/profanity-im/profanity/issues/1131
Probably missing copy of body to plain in carbon and privmessage.
Only covers the incoming message path because goal is OMEMO decryption
of untrusted message.
Cover some of the log functions but not all.
This is actually not needed because result will always be set.
hartmann had problems building profanity on OpenBSD and he proposed this
patch.
Let's use it to make the compiler happy there.
There is 3 policy:
- manual: OMEMO session are only started manually
- automatic: OMEMO session are only started if they have been started
manually before
- always: OMEMO session are always started unless they have been ended
manually before
Closes#1040 and fixes#1052
If OMEMO wasn't loaded, removing key would create empty identity.txt
file. Then at load time it would segfault profanity.
Another commit should fix load of malformed identity.txt.
When decrypting first message with prekey, libsignal wants to remove
used prekey from storage. Return value on success should be 0.
We used to return number of deleted keys. Thus libsignal was considering
we failed to remove the key and we were ignoring plaintext.
