| Commit message (Collapse) | Author | Age | Lines |
... | |
|
|
|
|
|
|
| |
The previous value (7) was FL | AL | BL, and the block role was replaced
with "Hide" in msnp21, so the server was rejecting the parts of the ADL
that had it. And since adding blocked contacts isn't very useful anyway,
this is like silencing an annoying warning.
|
|\ |
|
| |
| |
| |
| | |
packages will be installable.
|
| |
| |
| |
| |
| |
| |
| | |
Also fix a leak in msn_ns_callback while i'm at it.
Also fix a potential null deref when req->reply_body is null and
BITLBEE_DEBUG is enabled, but i don't even know if this one can happen.
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
CID 18634: 'Logically dead code' in jabber_get_info
CID 18638: 'Dereference after null check' in oauth2_access_token_done
CID 18691: 'Copy into fixed size buffer' in bee_irc_user_new
CID 20274: Leak in bee_irc_chat_invite
CID 20297, CID 20283: Leaks in crypt_main
Some the base64 leaks there weren't detected, needs modeling.
|
| |
| |
| |
| | |
Fixes trac ticket 1196
|
| |
| |
| |
| |
| |
| | |
Mostly to be able to test twitter_message_len externally against the
twitter-text conformance tests (the current version definitely fails -
it doesn't do utf8 normalization)
|
|\| |
|
| | |
|
| |
| |
| |
| |
| | |
Also don't send modes - clients are supposed to request that if they
need it.
|
| | |
|
|/ |
|
|\ |
|
| |
| |
| |
| |
| | |
Difficult because there's no bee_user struct pointing at the user themselves
so instead just fake one for very limited use.
|
| |
| |
| |
| | |
Also, fix bug in parsing of entities in DMs.
|
| | |
|
|/ |
|
|
|
|
|
|
|
|
| |
This is enough to log in with their usernames, make 'chat add' based
groupchat joins slightly more smooth, and see mention names as nicks.
All the MUC list stuff is left out intentionally since that's not as
stable as I wish.
|
|
|
|
|
|
|
| |
Third time's the charm.
Previous attempt fixed something and broke something else. Whatever.
Definitely valgrind clean this time.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Another take on the subprotocols idea that, IMO, was a failure.
Unlike the other implementation, this one doesn't touch gtalk/facebook
accounts, it just adds another copy of the "jabber" prpl called "hipchat".
And, based on the protocol name:
- sets JFLAG_HIPCHAT to jabber_data
- sets the default value of the "server" setting
- only includes the oauth setting for jabber-type accounts
This is slightly more "hardcoded" but honestly facebook and gtalk are
just as hardcoded as this.
Copying the prpl is needed because the meaning of the usernames is
completely different (there's no srv lookup stuff either)
|
|
|
|
|
|
| |
The servers send them for other reasons nowadays, and all
non-MPOP clients are blocked, so OTH is highly unlikely
to mean "someone else logged in with your account"
|
| |
|
|
|
|
| |
I avoided doing this before to keep the diff small. But meh.
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
|
| |
- bump protocol version
- update CVR parameters
- don't send BLP, send ADL right after getting contact list
- NFY payload size parsing
- MSNP21 style ADL
|
| |
|
|
|
|
| |
Fixes the test suite. I guess it's useful for something.
|
|
|
|
|
|
|
|
|
| |
Since this removes invitation.h, do "make distclean" to fix build errors
MSNFTP is a file transfer method used by early msn messenger versions,
and has been replaced by MSNP2P probably 10 years ago.
This code was disabled/commented out in bitlbee
|
|
|
|
| |
Also split underscore_dedupe from nick_dedupe.
|
|
|
|
| |
It's not 2003 anymore
|
| |
|
|
|
|
|
|
|
|
| |
If the from="..." of the message that includes a subject refers to us,
that buddy object won't have an ext_jid set, and passing that to
strchr() results in pain.
This happens with recent versions of an xmpp server called "lets-chat".
|
|\ |
|
| |
| |
| |
| | |
Added date to doc/CHANGES and fixed "make tar".
|
| | |
|
| | |
|
|/
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
- irc_im.c:
- bee_irc_user_msg: strdup leaks when otr swallows messages
- bee_irc_user_action_response: GString leak in all ctcp replies
- otr.c:
- call g_slist_free() on the list of the otr_policy setting
- otr_filter_msg_in: call otrl_tlv_free() if "tlvs" are returned
- otr_filter_msg_out: don't g_strdup() if the message should be ignored
- log_otr_message: g_strdup_vprintf() leaks always
- nogaim.c:
- imcb_ask_auth/imcb_ask_add: leaks in g_strdup_printf()
- imcb_ask_add leaks imcb_ask_cb_data if the user already exists
- add imcb_ask_cb_free() to correctly free its data
- msn_util.c: add msn_buddy_ask_free(), ditto
- storage_xml.c: pass_cr/password if base64_decode or arc_decode fail
- ssl_gnutls.c: conn->hostname leak in error conditions, like invalid certs
- jabber_util.c: jabber_buddy_by_ext_jid() leaks jid if it's not an ext jid
|
|
|
|
|
| |
Purple just doesn't work with daemon mode anyway, but it's better to
show the intended error message than to crash while showing it.
|
|
|
|
|
|
|
| |
- Added support for PURPLE_REQUEST_INPUT
- Changed memory management to do the free() of data through
purple_request_close(), letting purple know that the request was
answered, and fixing use-after-free issues with it
|