Note, by the way, that on other machines "clear" *used* to do the
right thing for me; we had a long thread about it a while back;
http://www.mail-archive.com/tmux-users@lists.sourceforge.net/msg01608.html
This is a completely new system, and I don't know what changed.
It's Fedora 15, running tmux 1.4.
Suggestions welcome for how to test/debug this, as usual.
-Robin
On Sat, Sep 24, 2011 at 06:15:29PM -0700, Robin Lee Powell wrote:
> On my current Fedora machine, no, it doesn't.
>
> If I run "clear", with alternate-screen off, it simply destroys
> what's on the screen; it doesn't go into tmux's backscroll/history
> or anything, it's just *gone*.
>
> I want this fixed so very, very badly. ;_;
>
> It annoys me so much that I may have to switch back to GNU screen,
> where I never had this problem, despite it being a worse program in
> basically every other respect.
>
> -Robin
>
> On Mon, Aug 22, 2011 at 10:28:09PM +0100, Nicholas Marriott wrote:
> > Don't we do this already? Clear screen scrolls lines into the history?
> > Or do you want something else? It isn't very obvious.
> >
> >
> >
> > On Mon, Aug 22, 2011 at 10:22:16PM +0200, Fabian Groffen wrote:
> > > Hi list,
> > >
> > > In my quest to get rid of the alternative buffer feature of todays
> > > terminals, I've configured tmux with:
> > >
> > > set -g terminal-overrides 'xterm*:smcup@:rmcup@'
> > > set -g alternate-screen off
> > >
> > > This basically does everything I want, except that compared to a
> > > non-tmux session (with terminfo settings to remove smcup and rmcup
> > > capabilities) a "clear" overwrites the existing contents of the
> > > terminal. E.g. when I run Vim in tmux, I lose the information on my
> > > screen, whereas outside it isn't. The \E[2J sequence triggers a full
> > > clear screen, and the attached patch changes the behaviour for that
> > > sequence in the tty_cmd_clearscreen function not to overwrite, but
> > > rather add empty lines to make the terminal clean. This is only done
> > > when alternate-screen is set to off, when this behaviour makes sense.
> > >
> > > Please consider the attached patch.
> > >
> > >
> > > --
> > > Fabian Groffen
> > > Gentoo on a different level
> >
> > > # HG changeset patch
> > > # User Fabian Groffen <grob...@gentoo.org>
> > > # Date 1314042382 -7200
> > > # Node ID 1b455a85ed6ca5e6aeea23a7152bffcc902966b3
> > > # Parent 051fdc014dc5bc730e4b0eb380af8b09938dd9b8
> > > tty_cmd_clearscreen: don't overwrite the buffer if alternate-screen=off
> > >
> > > If the alternative-screen isn't used, overwriting the buffer contents
> > > loses the information that the user probably wanted to retain (hence no
> > > alternate-screen). Instead, write enough newlines to clear the screen
> > > and reset the cursor, such that we retain all information that was on
> > > the screen before the clear.
> > >
> > > For example: having ls output and starting Vim won't overwrite the ls
> > > output, but show the output including the vim invocation.
> > >
> > > diff -r 051fdc014dc5 -r 1b455a85ed6c tty.c
> > > --- a/tty.c Mon Aug 22 20:14:55 2011 +0200
> > > +++ b/tty.c Mon Aug 22 21:46:22 2011 +0200
> > > @@ -899,24 +899,32 @@
> > >
> > > tty_reset(tty);
> > >
> > > - tty_region_pane(tty, ctx, 0, screen_size_y(s) - 1);
> > > - tty_cursor_pane(tty, ctx, 0, 0);
> > > + if (options_get_number(&wp->window->options, "alternate-screen")) {
> > > + tty_region_pane(tty, ctx, 0, screen_size_y(s) - 1);
> > > + tty_cursor_pane(tty, ctx, 0, 0);
> > >
> > > - if (wp->xoff == 0 && screen_size_x(s) >= tty->sx &&
> > > - tty_term_has(tty->term, TTYC_EL)) {
> > > - for (i = 0; i < screen_size_y(s); i++) {
> > > - tty_putcode(tty, TTYC_EL);
> > > - if (i != screen_size_y(s) - 1) {
> > > - tty_emulate_repeat(tty, TTYC_CUD, TTYC_CUD1, 1);
> > > - tty->cy++;
> > > + if (wp->xoff == 0 && screen_size_x(s) >= tty->sx &&
> > > + tty_term_has(tty->term, TTYC_EL)) {
> > > + for (i = 0; i < screen_size_y(s); i++) {
> > > + tty_putcode(tty, TTYC_EL);
> > > + if (i != screen_size_y(s) - 1) {
> > > + tty_emulate_repeat(tty, TTYC_CUD,
> > > TTYC_CUD1, 1);
> > > + tty->cy++;
> > > + }
> > > + }
> > > + } else {
> > > + for (j = 0; j < screen_size_y(s); j++) {
> > > + tty_cursor_pane(tty, ctx, 0, j);
> > > + for (i = 0; i < screen_size_x(s); i++)
> > > + tty_putc(tty, ' ');
> > > }
> > > }
> > > } else {
> > > - for (j = 0; j < screen_size_y(s); j++) {
> > > - tty_cursor_pane(tty, ctx, 0, j);
> > > - for (i = 0; i < screen_size_x(s); i++)
> > > - tty_putc(tty, ' ');
> > > - }
> > > + /* make sure we don't overwrite buffer contents, move down */
> > > + for (j = 0; j < screen_size_y(s); j++)
> > > + tty_putc(tty, '\n');
> > > + tty_region_pane(tty, ctx, 0, screen_size_y(s) - 1);
> > > + tty_cursor_pane(tty, ctx, 0, 0);
> > > }
> > > }
> > >
> >
> > > ------------------------------------------------------------------------------
> > > uberSVN's rich system and user administration capabilities and model
> > > configuration take the hassle out of deploying and managing Subversion
> > > and
> > > the tools developers use with it. Learn more about uberSVN and get a free
> > > download at: http://p.sf.net/sfu/wandisco-dev2dev
> >
> > > _______________________________________________
> > > tmux-users mailing list
> > > tmux-users@lists.sourceforge.net
> > > https://lists.sourceforge.net/lists/listinfo/tmux-users
> >
> >
> > ------------------------------------------------------------------------------
> > uberSVN's rich system and user administration capabilities and model
> > configuration take the hassle out of deploying and managing Subversion and
> > the tools developers use with it. Learn more about uberSVN and get a free
> > download at: http://p.sf.net/sfu/wandisco-dev2dev
> > _______________________________________________
> > tmux-users mailing list
> > tmux-users@lists.sourceforge.net
> > https://lists.sourceforge.net/lists/listinfo/tmux-users
------------------------------------------------------------------------------
All of the data generated in your IT infrastructure is seriously valuable.
Why? It contains a definitive record of application performance, security
threats, fraudulent activity, and more. Splunk takes this data and makes
sense of it. IT sense. And common sense.
http://p.sf.net/sfu/splunk-d2dcopy2
_______________________________________________
tmux-users mailing list
tmux-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/tmux-users
