The command is 'rendition'. Details in man-page. The defaults are
currently set in a way to make sure that the new settings are
noticeable. This changeset is preferred over either of the patches in
savannah #18382, because it allows specifying the renditions for both
bell and monitor windows, and is flexible enough that new renditions
can be added if desired.
The tilde-expansion didn't happen with the following screenrc command:
screen 0 screen -S name -c ~/file
With this change, proper tilde-expansion happens for the nested screen.
On OS X, screen fails to compile because WindowDied is
unconditionally declared to take an int, but is called
with an argument of type union wait. This wraps the
declaration and two associated calls.
Signed-off-by: William Pursell <bill.pursell@gmail.com>
"
This feature is useful when using a very slow remote shell session (SSH
over a bad GPRS connection). By locally preparing a command you can
enter a command and send it as bulk (without typos ;-)).
"
The gain from this patch is that, instead of doing ^a:stuff "command",
the user would be doing ^aXcommand. This is a relatively pretty minor
convenience, but I suppose it could be useful if the user needs to do
this frequently.
Some commands don't work correctly if screen is started in daemon mode
and not attached/detached at least once. This happens because without
any attaches, no window is considered as the foreground window when
remotely executing a command without a display. In such a case, treat
the last created window as the foreground window, as suggested in
#10074.
Fixes crash from commands such as setenv, readreg, 'layout select' etc.
when starting with 'screen -md'. Detailed report at #24152, debian bug
#496750.
Detailed bug report by YONETANI Tomokazu in #24058.
I am not entirely sure *why* this change fixes the bug, but it
apparently does (and it looks like a correct fix too, actually).
If there is a session with the exact name specified, then use that one
instead of matching other sessions with a similar name. Detailed bug
report by Marius Mikucionis
(http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=361274)
If a session has a password set, then it should be necessary for the
user to enter the password before a remote detach. This change achieves
that behaviour.
A small change in the messaging system is necessary, thus bumping the
message version, for the first time since it has been introduced, it
seems. But I don't think it's going to break anything.
Discussion at #24029
The blank layer, which is used by the windowlist, is not resized when
the canvas is resized, which caused the bug. This change should fix the
issue.
Bug reported by Jean-Yves Levesque in
<111f7ec40808050840t521e4d9bmcd54353087c81fec@mail.gmail.com>
