On Fri, Nov 29, 2013 at 21:25:39 -0500, Ben Boeckel wrote:
> I don't think block-until-there's-a-reader is possible (I think SIGPIPE
> is tossed into the mix right away).
Actually, the comment in uzbl is:
/* We don't really need to write to the file, but if we open the file as
* 'r' we w
On Fri, Nov 29, 2013 at 21:21:41 -0500, Timothy Brown wrote:
> You have it right. Perhaps the current behavior is the intended
> behavior; but what i’d like is for save-buffer to write to the fifo
> but either show an error message (“no readers” or the like) or block,
> in the worst case, until I
You have it right. Perhaps the current behavior is the intended behavior; but
what i’d like is for save-buffer to write to the fifo but either show an error
message (“no readers” or the like) or block, in the worst case, until I read
from the fifo. I haven’t taken a deep look yet, but i’m gue
On Thu, 28 Nov, 2013 at 14:23:39 GMT, Nicholas Marriott wrote:
> I don't know why it doesn't work, give me a set of commands to test?
Would it fail if the pipe has no readers when tmux writes to it?
--Ben
--
Rapidly tro
I don't know why it doesn't work, give me a set of commands to test?
On Tue, Nov 26, 2013 at 10:30:18AM -0500, Timothy Brown wrote:
> Way back in 2010, Tiago Cunha wrote this patch:
>
> Update of /cvsroot/tmux/tmux
> In directory sfp-cvsdas-2.v30.ch3.sourceforge.com:/tmp/cvs-serv10668
>
> M
Way back in 2010, Tiago Cunha wrote this patch:
Update of /cvsroot/tmux/tmux
In directory sfp-cvsdas-2.v30.ch3.sourceforge.com:/tmp/cvs-serv10668
Modified Files:
cmd-load-buffer.c
Log Message:
Sync OpenBSD patchset 652:
In load-buffer, read until EOF rather than using stat() and
