summaryrefslogtreecommitdiff
AgeCommit message (Collapse)Author
2014-12-25Add ipv6-, tcp- and proxy-mode to config.def.hFRIGN
There's no reason to prevent the user from setting this stuff at compile time.
2014-12-23Update LICENSEsin
2014-12-23Properly align tables as wellsin
2014-12-23Align globals and local variablessin
2014-12-23Fix indentation and alignmentsin
2014-12-16Update READMEsin
2014-12-16Update README with mailing list infosin
2014-12-11Simplify commentssin
2014-12-11Indent with tabs and align with spacessin
2014-12-09Refine config.def.h a bitFRIGN
2014-12-09Add ability to specify save file as command line argumentFRIGN
Also, get closer to the STS (Single Tox Standard) by calling the the save file "save file" and not "data file". Additionally, augment the manual section about encrypting and decrypting the save file.
2014-12-08Rework some of the log-warningsFRIGN
2014-12-08Proceed with cleanupFRIGN
2014-12-08Guard FD_APPEND with a do { } while(0)sin
2014-12-08Begin cleanup according to style(9)FRIGN
2014-12-07Bump version to 0.2sin
2014-12-07Re-prompt passphrase when creating a new encrypted saveFRIGN
This is more consistent with the de-facto behaviour of password-setting prompts.
2014-12-07Remove SEE ALSO section from manpagesin
2014-12-05Update LICENSE and AUTHORS section in manpagesin
2014-12-05Fix wording in manpagesin
2014-12-05Add -E and -e to enable and disable data file encryptionsin
2014-12-01Do not handle SIGABRT - let it terminate the process immediatelysin
2014-12-01Underline references to filessin
2014-12-01Add info about `id' file to manpagesin
2014-12-01Update manpagesin
2014-12-01Increase connection delay to 4 secondssin
2014-12-01Do the datasave() as early as possible in shutdown()sin
2014-12-01Save data file on receipt of a deadly signalsin
2014-12-01Update TODOsin
2014-11-30Update list of nodessin
2014-11-30Connect to nodes in a random order to minimize load on any given nodesin
2014-11-30Update TODOsin
2014-11-29Adapt to latest cleanup in APIFRIGN
2014-11-28Fix stylesin
2014-11-27Don't lock up on blocking call_outFRIGN
Since the av-refactor, we keep a call active as long as there is a listener or a sender. In case we don't have a listener, don't lock up in the sending function and rather just fall-through in case the write blocks. We can't do that as easily with the file-data-code, because data- integrity is very important there.
2014-11-26Fit ratox to new toxav-APIFRIGN
mannol has done a great job re-working the toxav-api[0] also based on our suggestions for a stricter callback-approach. Previously, toxav-callbacks could hit any time, leaving us with dramatic timing-issues and instability in certain situations. The new toxav_do()-call enables the client to control when to fire the callbacks. toxav_do_interval() makes sure no time is lost in the meantime. [0]: https://github.com/irungentoo/toxcore/pull/1163
2014-11-01Don't assume mono-audio in defaultsFRIGN
2014-10-24Use (none, pending, active) instead of (0, 1, 2)sin
2014-10-24Rename .ratox.data to .ratox.tox to comply with STSsin
https://github.com/Tox/Tox-STS/blob/master/STS.md
2014-10-23Add double spacing between sections in READMEsin
2014-10-23Add screencasting examplesin
2014-10-23Add an examples section in the READMEsin
2014-10-18Refactor logic in main-loop for pending calls even moreFRIGN
and implement fifoopen in other places.
2014-10-18Don't forget to queue call_out while incoming ringingFRIGN
Else the call would be rejected immediately in case you receive a call and have not yet opened call_out for reading.
2014-10-17Indentation fixessin
2014-10-17Refactor call-state managementFRIGN
Checking call_out only and stopping the call as soon as call_out is gone is not desirable behaviour. It now keeps up the running call until both call_out and call_in have been "closed", no matter how or by whom it was started. This is done by introducing av.state, which is a bit field. While at it, I also incorporated the transmission-state into it. toxav_hangup() will now only be called from the one centralized flag-checking-loop before select, not in the callbacks themselves. This will also get rid of some race-conditions (more or less). Previously, call_out was constantly polled as long as the user was online. Change this to only attempt to open call_out when the call is initialized. This lowers CPU-usage dramatically.
2014-10-16Remove ugly use of O_DIRECTORYsin
Oops kfreebsd builds break. Aw well.
2014-10-16CleanupFRIGN
AV-logging, logic and check for < 0 instead of == -1 (just to make sure).
2014-10-15Remember to cool down the transfer for pending buffers as wellsin
This change halves CPU usage on my system.
2014-10-15Only complain if errno is not EWOULDBLOCKsin