The documentation for resize.c:CheckScreenSize says to modify all
windows only when change_flag is 2. But ChangeScreenSize goes ahead and
resizes every window regardless of the value of change_flag. This caused
SIGWINCH to be sent to windows not in focus. This change fixes the issue.
Closes#23771.
The bug existed if compiled with multiuser support. Thanks to Maarten Billemont
and Geraint Edwards for reporting the bug and testing the fix. (#23817)
To: jw@cs.fau.de
Subject: Re: vertical split
Date: Tue, 20 Dec 2005 07:21:51 +0000
Juergen Weigert wrote:
>Bill,
>Let me ask you the license question now.
>
>I have now seen 3 patches of yours:
>
> OneCanvas.diff
> fFtT.patch
> vert_split.diff
>
>Do you agree to distribute your code under the GNU General Public License?
>Just as screen is distributable?
>
>
Absolutely. I just wanted to be able to use fF, but now that I'm digging around in
the code I'm having so much fun...But I'm not sure that any of those 3 are really
releasable. There were all pretty much slapped together fairly quickly, as I'm trying
to get my head around the screen code. Is there a screen patch naming convention?
the fFtT.patch file is almost releasable as is, but I think some of my comments got
left out of that file. If by "distribute...under the GPL" you mean incoroporate it
into the source on Savannah, then by all means yes, I'm perfectly happy with the GPL.
attacher.c: use attach_fd, use HAVE_SETRESUID
ansi.c: DoCSI now supports CSI 11t, 7t, 21t, 8t was only 8t before.
configure.in: quoting of SOCKDIR, use SVR4 instead of __SVR4, use sun instead of __sun.
Check for both setresuid() and setreuid(). Prefer sockets over pipes.
help.c: MAPKEYS term[] support. Use Layer2Window().
socket.c: SendAttachMsg() added.
process.c: bind -k added. KMAP_KEYS support.
screen.c: use attach_fd.
teln.c: TO_LINEMODE added, symmetric debug code.
termcap.c: ktab[]
tty.sh: use tcflow TCOON
tty.c.dist dated 1994 ignored.
utmp.c: no longer cast u->ut_(x)time into time_t.
