lp:~j-bormolini/ctrlproxy/main

Branch merges

Propose for merging

No branches dependent on this one.

Related source package recipes

No recipes using this branch. (?)

Create packaging recipe

Related snap packages

Related bugs

Related blueprints

1925. By Joe Bormolini on 2011-06-11

merge to try get bzr in sync...

1924. By Joe Bormolini on 2010-10-05

make CONNECT switch the client state better

Now simulate a disconnect and reconnect of the client so the network
metadata gets resent properly. This fixes an issue for irssi when
connecting in proxy mode (which uses CONNECT) where the name of the
network shows the default network.

maybe one day we can switch servers on the fly without disconnecting :)

1923. By Joe Bormolini on 2010-10-05

auto-away: don't clobber already-set away message

1922. By Joe Bormolini on 2010-10-05

rework isupport/005 to send 001 through 005

I think it's cleaner to redo the whole irc welcome process when the server
actually connects instead of just sending 005.

1921. By Joe Bormolini on 2010-10-05

fix issue with long nicknames and replication

Linestack saves the config instead of the actual server state. When we
reconnect we replicate the stored state (that was never updated to
reflect the true server state!).

This causes issues if the stored state does not match the actual state
(i.e. different nick, different channels, etc...).

Workaround issue by sending the network external state instead of
generating a new state based off of the linestack state.

1920. By Joe Bormolini on 2010-08-25

remove hardcoded "ctrlproxy" server names

1919. By Joe Bormolini on 2010-08-15

fix auto-away crash when invalid network specified

Introduced by me, I think - lose_client() was trying to send a message
to a network that never existed. Maybe we should only register the auto
away hook if the server connects in the first place, but this is simpler

1918. By Joe Bormolini on 2010-08-12

better support disabling auto away idle time

only register the log_data filter if we have a valid idle time set;
don't print a warning if max_idle_time == 0 (i.e. disabled idle time).

some things to consider for the future:
- per-network idle time setting (and anti-idle). the way it works now is
  if any client wakes from auto-away, then all connected network wake
  from away as well (even if no clients are attached and even if the
  original away was not from an auto-away timeout).
- disable check_time if we are already away, enable it on return
- wait longer than 1000ms between calls to check_time because
  max_idle_time is specified in increments of 60000ms
- why is 30 seconds too short for an idle time, anyway?

1917. By Joe Bormolini on 2010-08-10

abort early in network_send_line if not connected

I fixed a specific crash elsewhere, but this this might be a better way
of handling it...

1916. By Joe Bormolini on 2010-08-10

remove references to get_my_hostname from libirc