summaryrefslogtreecommitdiff
path: root/src/bytestream-factory.c
Commit message (Collapse)AuthorAgeFilesLines
* remove unused assignationsGuillaume Desmottes2014-06-031-2/+0
|
* bytestream_factory_negotiate_stream(): don't pass stream id to callbacksWill Thompson2012-12-061-1/+1
| | | | None of them use it, so…
* bytestream_factory_negotiate_stream(): remove return valueWill Thompson2012-12-061-6/+2
| | | | It's always TRUE, so it's unneeded.
* bytestream_factory_negotiate_stream(): use send_iq_asyncWill Thompson2012-12-061-22/+18
| | | | | This clarifies that this function can never fail, which can be the next thing we clean up…
* bytestream-factory: stop NIHing TpWeakRefWill Thompson2012-12-061-34/+14
| | | | | Step one in cleaning up gabble_bytestream_factory_negotiate_stream() is to simplify its context structure.
* Use meta-headers for everythingSimon McVittie2012-05-301-1/+2
| | | | | Reviewed-by: Xavier Claessens <xavier.claessens@collabora.co.uk> Bug: https://bugs.freedesktop.org/show_bug.cgi?id=49384
* Stop using tp_handle_ref/unrefXavier Claessens2012-05-101-5/+0
|
* Update Wocky snapshot for One Big Header.Will Thompson2012-02-061-1/+1
| | | | https://bugs.freedesktop.org/show_bug.cgi?id=27489
* Replace trivial uses of lm_message_node_get_child_with_namespaceWill Thompson2012-01-311-18/+17
| | | | | These are the cases where it was relatively quickly apparent to me that we only needed a direct child, not a deeper descendent.
* bytestream-factory: move to WockyNodeIterWill Thompson2012-01-311-34/+25
|
* Remove vestigial Loudmouth headers and build systemWill Thompson2012-01-271-1/+0
|
* Remove LmHandlerResultWill Thompson2012-01-271-6/+3
| | | | | | | | | | | | | | | | | | | | | | | | The callback type for _gabble_connection_send_with_reply returned one of these. All instances, bar two, returned LM_HANDLER_RESULT_REMOVE_MESSAGE, which makes sense, because if you send an IQ you had better be prepared to handle the reply. When I removed lm_connection_send_with_reply(), I stopped _gabble_connection_send_with_reply() paying attention to the return value, with no apparent ill effects. The two which returned LM_HANDLER_RESULT_ALLOW_MORE_HANDLERS were: • the disco pipeline's reply handler, in the case where the request is no longer in the pipeline (such as in a timeout). I checked, and as far as I can tell nothing is trying to scrape any information out of zombie disco replies. • the request pipeline's reply handler, in the same case. I couldn't be bothered to check for this, but I don't think it will pose any ill effects because normally the handler is the highest-priority callback, so anyone trying to catch these would not catch the ones which were not cancelled… which seems unlikely.
* BytestreamFactory: stop using LmMessageHandlerWill Thompson2012-01-271-117/+101
|
* Delete LmMessage{,Type,SubType} typedefsWill Thompson2012-01-271-36/+36
|
* Remove lm_message_get_[sub_]type.Will Thompson2012-01-271-7/+17
| | | | | Wocky is a little less convenient here, but wocky_stanza_extract_errors() comes to the rescue.
* Remove lm_message_build[_with_subtype]Will Thompson2012-01-271-26/+26
| | | | | | | | | | The build specs aren't *quite* compatible. lm_message_build()'s '(' accepted two arguments: tag name, and content (usually ""). So: '(', "foo", "" becomes '(', "foo" '(', "foo", "bar" becomes '(', "foo", '$', "bar" I didn't actually do this with sed as a result.
* Expunge lm_message_ref and lm_message_unref.Will Thompson2012-01-261-1/+1
| | | | And also a couple of uses of lm_message_get_node().
* Merge branch 'master' into BYE-BYE-LOUDMOUTHWill Thompson2012-01-261-6/+9
|\ | | | | | | | | | | | | | | | | | | | | | | | | Conflicts: src/connection.c src/connection.h src/ft-manager.c src/jingle-content.c src/jingle-factory.c src/jingle-session.c src/message-util.c src/muc-channel.c src/muc-channel.h
| * Add configure option to disable file transfer support.Andre Moreira Magalhaes (andrunko)2011-12-201-0/+2
| |
| * Remove gabble_decode_jid in favour of wocky_decode_jid.Andre Moreira Magalhaes (andrunko)2011-11-171-2/+3
| |
| * Use _unref instead of _free _destroy when possible.unrefXavier Claessens2011-11-161-4/+4
| | | | | | | | | | | | | | | | | | | | | | Replace g_(ptr_)array_free (foo, TRUE) and g_hash_table_destroy with respectively g_(ptr_)array_unref (foo) and g_hash_table_unref. I used this command to generate this patch: for f in `find -name "*.c"`; do sed -i $f -re 's/g_ptr_array_free \(([^ ,]+), TRUE\)/g_ptr_array_unref \(\1\)/'; done See Danielle's blog for explanation of possible bug _free can do: http://blogs.gnome.org/danni/2011/11/16/mistakes-with-g_value_set_boxed/
* | Remove GabbleXmppError.Will Thompson2011-05-031-40/+43
| | | | | | | | There is a FIXME buried in here … but the tests pass.
* | stop using LmMessageNode wrapperJonny Lamb2011-05-021-60/+60
| | | | | | | | | | | | WockyNode is the future. Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>
* | use the new xmlns and xml:lang build tagsJonny Lamb2011-05-021-10/+10
|/ | | | Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>
* Rename to WockyStanza, WockyNode and related changesSjoerd Simons2010-04-201-41/+49
|
* Remove the proxy from the potential proxies list if the result is invalidSjoerd Simons2010-04-061-2/+2
|
* Use g_resolver_lookup_by_name instead of getaddrinfo in GibberTcpTransportSjoerd Simons2010-04-061-8/+13
|
* Don't accept an open ibb iq if we didn't accept the bytestream yetSjoerd Simons2009-12-181-0/+12
|
* Make sure we only call the callback if it's still aliveSjoerd Simons2009-12-181-4/+3
|
* plug various leaksDafydd Harries2009-12-181-1/+1
|
* remove useless \nGuillaume Desmottes2009-12-071-1/+0
|
* fix typoGuillaume Desmottes2009-12-071-1/+1
|
* remove proxy from the list if we receive an IQ errorGuillaume Desmottes2009-12-071-0/+24
|
* use a timestamp instead of a timer to manage the lifetime of the proxies listGuillaume Desmottes2009-12-071-30/+11
|
* proxies_disco_cb: log a debug msg if disco failedGuillaume Desmottes2009-12-071-1/+4
|
* Always recycle the socks5_potential_proxies listGuillaume Desmottes2009-12-071-16/+5
| | | | | | We ensure that we won't query twice the same proxy by clamping the number of proxy needed to the length of the list of the potential proxies. This add the nice side effect of prioritizing a different proxy every time.
* proxies_disco_cb: Replace the old list instead of adding elements to itGuillaume Desmottes2009-12-071-5/+8
|
* add_proxy_to_list: free the oldest proxy once removedGuillaume Desmottes2009-12-071-0/+1
|
* remove double ';'Guillaume Desmottes2009-12-071-1/+1
|
* mark the proxies list has expired after 6 hoursGuillaume Desmottes2009-12-071-0/+40
|
* query proxies.telepathy.im if there is no fallback serverGuillaume Desmottes2009-12-071-2/+68
| | | | This server returns a list of public proxies that can be used as fallback.
* remove the sock5 proxies timer; instead request a new proxy if a reply is wrongGuillaume Desmottes2009-12-071-39/+12
|
* factor out query_proxiesGuillaume Desmottes2009-12-071-25/+35
|
* don't recycle proxies list if that doesn't make senseGuillaume Desmottes2009-12-071-2/+14
|
* recycle the list of potential proxiesGuillaume Desmottes2009-12-071-6/+18
| | | | | This avoid to get stuck with the latest proxies added in the cache once we queried all the potential proxies.
* gabble_bytestream_factory_get_socks5_proxies: stop randomizing proxiesGuillaume Desmottes2009-12-071-4/+0
| | | | The newer proxies have to be put first in the SOCKS5 stanza.
* streamhost are identified by their hostGuillaume Desmottes2009-12-071-11/+2
| | | | | | The SOCKS5 reply from the other peer only contains the jid of the streamhost. We can't have more than one proxy using the same jid as then we won't be able to know which one is used by the peer.
* remove oldest proxy from the cacheGuillaume Desmottes2009-12-071-0/+15
|
* move existing proxy to the head of the list when they are re-discoveredGuillaume Desmottes2009-12-071-6/+13
|
* gabble_bytestream_factory_query_socks5_proxies: always request at least one ↵Guillaume Desmottes2009-12-071-5/+8
| | | | | | proxy This ensure to always have at least one fresh proxy in the cache.
View Lorry Controller Status