Nabble - Gnu - Screen

Re: String escape for total number of windows?

2009-12-20T21:53:23Z

May be you need to run autogen.sh or your autoconf is too old.

http://www.mail-archive.com/screen-users@.../msg02475.html

On Mon, Dec 21, 2009 at 11:03 AM, John Magolske <listmail@...> wrote:

> * Thor Andreassen <ta@...> [091218 08:36]:
>> On Thu, Dec 10, 2009 at 10:25:11PM -0800, John Magolske wrote:
>> > Is there a string escape sequence that will display the total number
>> > of windows (just a single number, not all window numbers and names)?
>> > ...
>> ...
>> I've attached a patch that copies some of the functionality of the '%w'
>> handling, and should be applied to the newest source on savannah. It
>> seems to produce the correct result, although someone more familiar with
>> the code should check it for any unwanted consequences.
>
> Thanks for the patch, this should be very useful.
>
> For some reason I'm getting an error when running configure:
>
> % git clone git://git.savannah.gnu.org/screen.git
>
> % cd /screen/src
>
> % cp ~/number_of_windows.patch .
>
> % patch -i number_of_windows.patch
> patching file extern.h
> patching file process.c
> Hunk #1 succeeded at 5342 (offset 25 lines).
> patching file screen.c
>
> % autoconf
>
> % ./configure
> ...
> ...
> config.status: error: cannot find input file: config.h.in
>
>
> John
>
>
> --
> John Magolske
> http://B79.net/contact
>
>
> _______________________________________________
> screen-users mailing list
> screen-users@...
> http://lists.gnu.org/mailman/listinfo/screen-users
>

_______________________________________________
screen-users mailing list
screen-users@...
http://lists.gnu.org/mailman/listinfo/screen-users

Re: String escape for total number of windows?

2009-12-20T21:33:08Z

* Thor Andreassen <ta@...> [091218 08:36]:
> On Thu, Dec 10, 2009 at 10:25:11PM -0800, John Magolske wrote:
> > Is there a string escape sequence that will display the total number
> > of windows (just a single number, not all window numbers and names)?
> > ...
> ...
> I've attached a patch that copies some of the functionality of the '%w'
> handling, and should be applied to the newest source on savannah. It
> seems to produce the correct result, although someone more familiar with
> the code should check it for any unwanted consequences.

Thanks for the patch, this should be very useful.

For some reason I'm getting an error when running configure:

% git clone git://git.savannah.gnu.org/screen.git

% cd /screen/src

% cp ~/number_of_windows.patch .

% patch -i number_of_windows.patch
patching file extern.h
patching file process.c
Hunk #1 succeeded at 5342 (offset 25 lines).
patching file screen.c

% autoconf

% ./configure
...
...
config.status: error: cannot find input file: config.h.in

John

--
John Magolske
http://B79.net/contact

_______________________________________________
screen-users mailing list
screen-users@...
http://lists.gnu.org/mailman/listinfo/screen-users

Re: terminal capability problem when login therough screen

2009-12-20T20:36:00Z

On Mon, Dec 21, 2009 at 09:31:38AM +0530, J. Bakshi wrote:
> But #1 actually set the TERM environment in local box. Is not there any
> command option in screen to set TERM just for screen session and not
> globally ? I have tried with "screen -T xterm" but found the cursor is
> not working properly with vim; where setting TERM environment locally
> solved the issue. Any workaround with "screen -T" ?

From the point of view of screen, these two things should be the same.

# first
$ export TERM=xterm
$ screen

# second
$ TERM=xterm screen

You could create an alias to screen that does the second thing.

-- Erik

_______________________________________________
screen-users mailing list
screen-users@...
http://lists.gnu.org/mailman/listinfo/screen-users

Re: terminal capability problem when login therough screen

2009-12-20T20:01:38Z

Erik Osheim wrote:

> On Fri, Dec 18, 2009 at 12:38:58PM -0500, Pia Mikeal wrote:
>
>> I usually just put something like the following in my .profile on remote
>> servers
>>
>> export TERM=vt100
>>
>
> Alternately you can do this:
>
> (within screen on local box)
> user@local$ terminfo > SOMEFILE
> user@local$ scp SOMEFILE user@remote:
>
> (then on remote)
> user@remote$ tic SOMEFILE
>
> This will copy the terminal definition your local box is using onto the
> server. If you only have a small number of servers I like this
> approach--if you have tons it can be tedious to make sure they all have
> the corect TERM setting and I might be more tempted to do something
> like what Pia suggests.
>
> -- Erik
>
>
>
>

Hello Pia and Erik,

Thanks for your responses. I am searching some technique where I don't
need to do any modification on remote server as I have to maintain a no.
of remote server. I got 2 solution.

#1. In local box
----------------------

` ` `
TERM=xterm
` ` `

#2.
------

In debian lenny box the screen version does not have the problem.

But #1 actually set the TERM environment in local box. Is not there any
command option in screen to set TERM just for screen session and not
globally ? I have tried with "screen -T xterm" but found the cursor is
not working properly with vim; where setting TERM environment locally
solved the issue. Any workaround with "screen -T" ?

Thanks

--
জয়দীপ বক্সী

_______________________________________________
screen-users mailing list
screen-users@...
http://lists.gnu.org/mailman/listinfo/screen-users

Re: Ping Packagers

2009-12-20T19:18:04Z

On Sat, Dec 19, 2009 at 1:51 PM, Sadrul Habib Chowdhury
<imadil@...> wrote:

> * Dustin Kirkland had this to say on [09 Dec 2009, 10:37:17 -0800]:
>> On Wed, Dec 9, 2009 at 10:23 AM, Dustin Kirkland <kirkland@...> wrote:
>> > On Wed, Dec 9, 2009 at 10:19 AM, Sadrul Habib Chowdhury
>> > <imadil@...> wrote:
>> >> Hello everyone. Are there any screen packagers on this list? I am
>> >> contemplating a sort of a beta-release somewhat soon-ish (still
>> >> non-trivial amount of work to do, though, so no ETA yet), and was
>> >> thinking if this list is the best place to reach the packagers.
>> >
>> > Hello-
>> >
>> > I maintain (along with some others) the screen package in Ubuntu, and
>> > I do follow this list.
>> >
>> > We (Ubuntu) as well as Debian are carrying a stack of patches against
>> > Screen. Could we see about getting as many of these as possible into
>> > your next upstream release?
>>
>> FYI, you can see the patches we're carrying at:
>> http://bazaar.launchpad.net/~ubuntu-branches/ubuntu/karmic/screen/karmic/files/head%3A/debian/patches/
> [snip]
>
> I looked at a number of them:

Great, thanks!

>> 17manpage_sessionname_warning.dpatch
>
> Applied after slight modification.
> (855b8e7b98894e8f94033945720714c235911262)
>
>> 18manpage_maxwin_limit.dpatch
>
> Applied after removing the note about debian.
> (04005670529085ef8527608c0ec34b06569f0bd8)
>
>> 20defmonitor.dpatch
>
> Already in.
>
>> 21manpage_nethack_activation.dpatch
>
> Applied. (3b19a1d343e72e4fa786e9ab514d80303902a2c1)
>
>> 25allow_symlink_sockdir.dpatch
>
> Applied. (1e450254e75877b6bb5cacedd5466245dbb54576)
>
>> 27doc_sty_noenvpassing.dpatch
>
> Applied. (f6c4b82)
>
>> 28blankerprg_callsemantics.dpatch
>
> Applied in 52efeb8 after some change. Instead of unsetting blankerprg on
> no argument, it will show the currently set blankerprg. To unset, use ''
> as the argument. I have updated the man/info page accordingly.
>
>> 31upstream_cherries.dpatch
>
> Already in.
>
>> 32misc_minor_fixes.dpatch
>
> eh, I like that message :)
>
>> 33increase_max_winmsg_renditions.dpatch
>
> I think 256 is a bit too excessive. I am going to increase it to 128
> instead, which I honestly think should be more than enough.

Hmm, what's the actual cost? Some memory? How much? Would it be
realistic to make this dynamically allocated?

I maintain a project that provides a layer on top of Screen called
Byobu. We make heavy use of the hard status formatting with a series
of several dozen toggle on/off status scripts. With 1920x1080
displays, and an 8 pitch font, there's a lot of room across the bottom
of the screen.

See for an example this screenshot (with almost everything toggled on):
http://rookery.canonical.com/~kirkland/Byobu.png

For each color change, I also use \005{-} to "undo" it, and bring it
back to the native color. This means that each color change costs 2
against that 128 or 256 number. They add up quickly.

I appreciate your bumping it up to 128. We will probably continue to
carry a patch for 256 in Ubuntu. Assuming it's not too expensive, it
would be nice if you would reconsider. Thanks.

> I am also considering increasing the oft requested window limit. That
> will probably require a bit more work than just changing the define.

Definitely would be nice. I'm currently hitting the limit at 40
windows. What are you thinking for a limit? Again, a dynamic malloc
would be nice (haven't looked at the code to judge the feasibility,
though).

>> 40cjk_eastasian.dpatch
>
> Already in.
>
>> 45suppress_remap.dpatch
>
> I think this is incorrect. What keybindings are broken now? I think
> 17b43961982caa5132d98544addc4de3f89103db fixes at least part of the bug
> this patch tries to fix. I need to know if keys other than the end-key
> are broken.

Checking with Loic, on CC.

Loic- the upstream Screen maintainer is helping review the stack of
patches Debian and Ubuntu are carrying and there's a question (above)
about the current applicability of your patch. Care to comment?

>> 50EXP_tilde_expansion.dpatch
>
> Already in.
>
>> 51EXP_session_creation_time.dpatch
>
> Need to take a closer look. The idea sounds nice. But it doesn't look
> like the implementation is very portable. I think it would be simpler to
> add the timestamp of the session creation in the socket name, like we now
> add just the PID.
>
>> 56-source-file-not-found-warning.dpatch
>
> I disagree with this change. I think the error message is appropriate.

Understood. Would you support an additional, new directive that would
source-if-found, but throw no warning if not? Something like
"-source" or "@source" a la Makefile syntax?

>> 58-show-encoding-hardstatus.dpatch
>
> I don't like this patch. I wouldn't want more string escapes unless they
> are really useful. In this case, I am not convinced that showing the
> encoding in the caption is all that useful. This information is available
> in 'info'. Is there a use case where that's not enough?

I'm not sure. An Ubuntu user complained about this, and provided a
patch that seemed to work for me. It didn't bother me as a packager,
so I included it. If it's ill-advised, or evil somehow, I can drop it
and inform the user that upstream rejected it.

>> 59-no-beep-on-write-acl.dpatch
>
> Applied. (84b8ef5)
>
> I can't get the rest of them at the moment, setting an "Internal Server
> Error" from launchpad.

Sorry about that. There was a Launchpad maintenance window recently,
maybe you hit that.

Thanks for going over these!

In the future, what's the currently preferred mechanism for patch
submission? Devel mailing list, bug tracker, or otherwise?

:-Dustin

_______________________________________________
screen-users mailing list
screen-users@...
http://lists.gnu.org/mailman/listinfo/screen-users

Re: Cannot set $TERM in screen -- it always defaults to xterm

2009-12-19T17:30:27Z

>
> What shell are you using, and does the problem persist if you start your
> shell without config files?

That seemed to be the problem -- by .bashrc had export TERM=xterm. I still
don't understand why that would override trying to set TERM directly in screen,
but after I had removed the line from bashrc and toyed around with $TERM
setting, I was able to get 256 color support under xterm.

_______________________________________________
screen-users mailing list
screen-users@...
http://lists.gnu.org/mailman/listinfo/screen-users

Re: Ping Packagers

2009-12-19T16:21:21Z

* Miroslav Lichvar had this to say on [10 Dec 2009, 10:35:00 +0100]:

> On Wed, Dec 09, 2009 at 01:19:00PM -0500, Sadrul Habib Chowdhury wrote:
> > Hello everyone. Are there any screen packagers on this list? I am
> > contemplating a sort of a beta-release somewhat soon-ish (still
> > non-trivial amount of work to do, though, so no ETA yet), and was
> > thinking if this list is the best place to reach the packagers.
>
> Hi,
>
> I'm the maintainer for Fedora. You can see the patches we
> are currently using here:
>
> http://cvs.fedoraproject.org/viewvc/devel/screen/

Thanks for the link. I looked at a number of them:

> screen-4.0.1-args.patch

Applied after some change (7ac593d74dfd2243cd60c5d848547ebd9971a8b0).

I didn't include the change in ansi.c: w_NumArgs should never be
negative. If something is causing is to be set to a negative number, then
that's a bug and we need to figure where/why/how that happens, so the
real bug can be fixed.

> screen-4.0.2-maxstr.patch

It looks like MAXSTR has been increased to 512. It can still be increased
more, of course. Is there a bug-report somewhere that requested for this
change? I would like to know exactly what problems the limitations
caused.

> screen-4.0.3-libs.patch

I plan to take a look at https://savannah.gnu.org/bugs/?27318 first.

> screen-4.0.3-resize.patch

Already in.

> screen-4.0.3-ipv6.patch

Need to take a closer look.

> screen-4.0.3-stropts.patch

I believe e3b1d2c50a5cc84a0985dbf608b9ee47996b7e47 takes care of this?
(i.e. using HAVE_STROPTS_H)

Cheers,
Sadrul

_______________________________________________
screen-users mailing list
screen-users@...
http://lists.gnu.org/mailman/listinfo/screen-users

Re: Ping Packagers

2009-12-19T11:44:13Z

* Dustin Kirkland had this to say on [09 Dec 2009, 10:37:17 -0800]:

> On Wed, Dec 9, 2009 at 10:23 AM, Dustin Kirkland <kirkland@...> wrote:
> > On Wed, Dec 9, 2009 at 10:19 AM, Sadrul Habib Chowdhury
> > <imadil@...> wrote:
> >> Hello everyone. Are there any screen packagers on this list? I am
> >> contemplating a sort of a beta-release somewhat soon-ish (still
> >> non-trivial amount of work to do, though, so no ETA yet), and was
> >> thinking if this list is the best place to reach the packagers.
> >
> > Hello-
> >
> > I maintain (along with some others) the screen package in Ubuntu, and
> > I do follow this list.
> >
> > We (Ubuntu) as well as Debian are carrying a stack of patches against
> > Screen. Could we see about getting as many of these as possible into
> > your next upstream release?
>
> FYI, you can see the patches we're carrying at:
> http://bazaar.launchpad.net/~ubuntu-branches/ubuntu/karmic/screen/karmic/files/head%3A/debian/patches/

[snip]

I looked at a number of them:

> 17manpage_sessionname_warning.dpatch

Applied after slight modification.
(855b8e7b98894e8f94033945720714c235911262)

> 18manpage_maxwin_limit.dpatch

Applied after removing the note about debian.
(04005670529085ef8527608c0ec34b06569f0bd8)

> 20defmonitor.dpatch

Already in.

> 21manpage_nethack_activation.dpatch

Applied. (3b19a1d343e72e4fa786e9ab514d80303902a2c1)

> 25allow_symlink_sockdir.dpatch

Applied. (1e450254e75877b6bb5cacedd5466245dbb54576)

> 27doc_sty_noenvpassing.dpatch

Applied. (f6c4b82)

> 28blankerprg_callsemantics.dpatch

Applied in 52efeb8 after some change. Instead of unsetting blankerprg on
no argument, it will show the currently set blankerprg. To unset, use ''
as the argument. I have updated the man/info page accordingly.

> 31upstream_cherries.dpatch

Already in.

> 32misc_minor_fixes.dpatch

eh, I like that message :)

> 33increase_max_winmsg_renditions.dpatch

I think 256 is a bit too excessive. I am going to increase it to 128
instead, which I honestly think should be more than enough.

I am also considering increasing the oft requested window limit. That
will probably require a bit more work than just changing the define.

> 40cjk_eastasian.dpatch

Already in.

> 45suppress_remap.dpatch

I think this is incorrect. What keybindings are broken now? I think
17b43961982caa5132d98544addc4de3f89103db fixes at least part of the bug
this patch tries to fix. I need to know if keys other than the end-key
are broken.

> 50EXP_tilde_expansion.dpatch

Already in.

> 51EXP_session_creation_time.dpatch

Need to take a closer look. The idea sounds nice. But it doesn't look
like the implementation is very portable. I think it would be simpler to
add the timestamp of the session creation in the socket name, like we now
add just the PID.

> 56-source-file-not-found-warning.dpatch

I disagree with this change. I think the error message is appropriate.

> 58-show-encoding-hardstatus.dpatch

I don't like this patch. I wouldn't want more string escapes unless they
are really useful. In this case, I am not convinced that showing the
encoding in the caption is all that useful. This information is available
in 'info'. Is there a use case where that's not enough?

> 59-no-beep-on-write-acl.dpatch

Applied. (84b8ef5)

I can't get the rest of them at the moment, getting an "Internal Server
Error" from launchpad.

I plan to start looking at the fedora patches next, and the rest of the
debian/ubuntu patches.

Cheers,
Sadrul

_______________________________________________
screen-users mailing list
screen-users@...
http://lists.gnu.org/mailman/listinfo/screen-users

Re: String escape for total number of windows?

2009-12-19T04:08:23Z

Thanks for the patch. I was looking for this functionality.

2009/12/18, Thor Andreassen <ta@...>:

> On Thu, Dec 10, 2009 at 10:25:11PM -0800, John Magolske wrote:
>> Is there a string escape sequence that will display the total number
>> of windows (just a single number, not all window numbers and names)?
>
> There does not seem to be.
>
>> For example, if I were in the 7th window and 23 windows were open,
>> I'd like to see something like
>>
>> 7 / 23 : vim ~/.mutt/temp/mutt-1000-25786-33
>>
>> by setting hardstatus to something like
>>
>> hardstatus alwayslastline "%n / %x : %t"
>>
>> Of course %x is not an escape for anything, I'm just using it here as
>> a placeholder...there doesn't seem to be an escape for total number of
>> windows. Might it be possible to achieve this by truncating the output
>> of %w or some such thing?
>
> Looking at the code, at least with my eyes, truncation seems difficult
> and will probably end up as a hack.
>
> I've attached a patch that copies some of the functionality of the '%w'
> handling, and should be applied to the newest source on savannah. It
> seems to produce the correct result, although someone more familiar with
> the code should check it for any unwanted consequences.
>
> [...]
>
> --
> HTH
> Thor
>

--
Artur Skonecki
http://adb.cba.pl

_______________________________________________
screen-users mailing list
screen-users@...
http://lists.gnu.org/mailman/listinfo/screen-users

Re: Detached screens disappear - help debugging?

2009-12-18T18:05:52Z

May be this another possible reason then.

http://www4.informatik.uni-erlangen.de/~jnweiger/screen-faq.html

Q: If I leave screen too long detached it mysteriously disappears,
although I feel that this is not a flaw in the program but rather my
system being 'cleaned' of stray jobs.
A: Check the output of "screen -ls". If there is no session, your
shells inside screen possibly have set an autologout. This kills the
processes after a certain time of inactivity. Such a feature is often
installed per default in the users account, as it helps to keep
intruders off on terminals where you forgot to log out. But it is not
a good idea for shells running under screen.

http://www.ivorde.ro/Secure_Unix_shell_account_with_auto_logout_after_seconds_of_inactivity__Idle_TMOUT-42.html

Regards
Pandu

On Sat, Dec 19, 2009 at 1:38 AM, Leinartas, Michael
<MICHAEL.LEINARTAS@...> wrote:

> My screen sockets are actually in /var/run/screen/S-<user>/
>
> I can't find anything that would be clearing it (checked crons, logrotate, etc). When these screens disappear, both the SCREEN process and the socket are gone. It seems almost like screen shuts itself down for some reason. If it were shutting down uncleanly (say a crash or a kill -9) it'd leave behind a socket, right?
>
>
> -----Original Message-----
> From: Pandurangan R S [mailto:pandurangan.r.s@...]
> Sent: Wednesday, December 16, 2009 9:29 PM
> To: Leinartas, Michael
> Cc: screen-users@...
> Subject: Re: Detached screens disappear - help debugging?
>
>>> After detaching, my screen session will disappear some time after.
>
> Is the screen process dead (check using ps) or you are not just able
> to attach to the screen.
>
> If the screen process is alive and "screen -ls" tells "No sockets found"
> then it may the case that your sockdir (usually /tmp/uscreens/<user>)
> is getting cleaned up!
>
> My admin was wiping off the entire /tmp when diskspace was low and
> this was causing me a similar issue.
>
> Regards
> Pandu
>
> On Wed, Dec 16, 2009 at 11:06 AM, Leinartas, Michael
> <MICHAEL.LEINARTAS@...> wrote:
>> I've had a problem with screen on my home computer for quite some time
>> that's been driving me crazy. After detaching, my screen session will
>> disappear some time after. Sometimes an hour later it's there, sometimes
>> it's not. It's nearly always gone if I leave it overnight. I never have
>> any problems if I'm attached - if I say, leave a screen session on the
>> console open for 3 days, it'll be fine. It also doesn't seem to matter how
>> many terminals are in the session or whether they're idle at a prompt or
>> running something.
>>
>> So where can I start figuring out what's causing these sessions to die? It
>> doesn't seem to be anything obvious like an oom killer (nothing in dmesg).
>> I've checked the system crons in /etc/cron.* and it doesn't look like
>> there's anything wiping stuff in /var/run or anything.
>>
>> Some system details incase it matters to anyone:
>> Screen 4.0.3
>> Debian squeeze (same problem when I was on lenny) amd64 arch
>> linux-2.6.32 (vanilla)
>> AMD quad core (Phenom II)
>> 4gb ram
>>
>>
>> I'd appreciate any help on this because it's driving me nuts.
>>
>> michael
>>
>> _______________________________________________
>> screen-users mailing list
>> screen-users@...
>> http://lists.gnu.org/mailman/listinfo/screen-users
>>
>>
>

_______________________________________________
screen-users mailing list
screen-users@...
http://lists.gnu.org/mailman/listinfo/screen-users

Re: hstatus?

2009-12-18T15:54:46Z

On Mon, Dec 14, 2009 at 12:59:24PM -0500, Aaron Davies wrote:

> Screen 4.00.02, running on RHEL AS 4.5 (and possibly further
> customized by my employer, I'm not sure) sets the PuTTY window title
> to
>
> [screen %n: %t]
>
> (Those are screen escapes, e.g. a literal title might be "[screen 0:
> -ksh]".)
>
> Any change I make to hstatus is *appended* to this, rather than
> replacing it.

Does changing hardstatus instead work better?

> Where does it come from, and how can I change it?

Don't know, I can't reproduce said behavior here.

[...]

--
HTH
Thor

_______________________________________________
screen-users mailing list
screen-users@...
http://lists.gnu.org/mailman/listinfo/screen-users

RE: Detached screens disappear - help debugging?

2009-12-18T12:08:55Z

My screen sockets are actually in /var/run/screen/S-<user>/

I can't find anything that would be clearing it (checked crons, logrotate, etc). When these screens disappear, both the SCREEN process and the socket are gone. It seems almost like screen shuts itself down for some reason. If it were shutting down uncleanly (say a crash or a kill -9) it'd leave behind a socket, right?

-----Original Message-----
From: Pandurangan R S [mailto:pandurangan.r.s@...]
Sent: Wednesday, December 16, 2009 9:29 PM
To: Leinartas, Michael
Cc: screen-users@...
Subject: Re: Detached screens disappear - help debugging?

>> After detaching, my screen session will disappear some time after.

Is the screen process dead (check using ps) or you are not just able
to attach to the screen.

If the screen process is alive and "screen -ls" tells "No sockets found"
then it may the case that your sockdir (usually /tmp/uscreens/<user>)
is getting cleaned up!

My admin was wiping off the entire /tmp when diskspace was low and
this was causing me a similar issue.

Regards
Pandu

On Wed, Dec 16, 2009 at 11:06 AM, Leinartas, Michael
<MICHAEL.LEINARTAS@...> wrote:

> I've had a problem with screen on my home computer for quite some time
> that's been driving me crazy. After detaching, my screen session will
> disappear some time after. Sometimes an hour later it's there, sometimes
> it's not. It's nearly always gone if I leave it overnight. I never have
> any problems if I'm attached - if I say, leave a screen session on the
> console open for 3 days, it'll be fine. It also doesn't seem to matter how
> many terminals are in the session or whether they're idle at a prompt or
> running something.
>
> So where can I start figuring out what's causing these sessions to die? It
> doesn't seem to be anything obvious like an oom killer (nothing in dmesg).
> I've checked the system crons in /etc/cron.* and it doesn't look like
> there's anything wiping stuff in /var/run or anything.
>
> Some system details incase it matters to anyone:
> Screen 4.0.3
> Debian squeeze (same problem when I was on lenny) amd64 arch
> linux-2.6.32 (vanilla)
> AMD quad core (Phenom II)
> 4gb ram
>
>
> I'd appreciate any help on this because it's driving me nuts.
>
> michael
>
> _______________________________________________
> screen-users mailing list
> screen-users@...
> http://lists.gnu.org/mailman/listinfo/screen-users
>
>

_______________________________________________
screen-users mailing list
screen-users@...
http://lists.gnu.org/mailman/listinfo/screen-users

Re: terminal capability problem when login therough screen

2009-12-18T10:22:12Z

On Fri, Dec 18, 2009 at 12:38:58PM -0500, Pia Mikeal wrote:
> I usually just put something like the following in my .profile on remote
> servers
>
> export TERM=vt100

Alternately you can do this:

(within screen on local box)
user@local$ terminfo > SOMEFILE
user@local$ scp SOMEFILE user@remote:

(then on remote)
user@remote$ tic SOMEFILE

This will copy the terminal definition your local box is using onto the
server. If you only have a small number of servers I like this
approach--if you have tons it can be tedious to make sure they all have
the corect TERM setting and I might be more tempted to do something
like what Pia suggests.

-- Erik

_______________________________________________
screen-users mailing list
screen-users@...
http://lists.gnu.org/mailman/listinfo/screen-users

Re: terminal capability problem when login therough screen

2009-12-18T09:38:58Z

I usually just put something like the following in my .profile on remote
servers

export TERM=vt100

because your problem is that the remote box does not know about the
terminal type that your local screen sets it to. The reason I do this on
a case by case basis rather than putting it in the local .screenrc is
because I usually want to take advantage of the terminal settings I have
in my local screen, but some of my servers only support vt100

HTH,

Pia

On Thu, 17 Dec 2009, J. Bakshi wrote:

> Hello list,
>
> A strange issue which I encounter during ssh through screen. If I ssh to
> my remote servers ( one debian and one suse) I have no problem at all to
> use vim. But if I *ssh through screen* then the vim on remote debian
> box reports
>
> ` ` `
> E558: Terminal entry not found in terminfo
> 'screen.rxvt' not known. Available builtin terminals are:
> builtin_riscos
> builtin_amiga
> builtin_beos-ansi
> builtin_ansi
> builtin_pcansi
> builtin_win32
> builtin_vt320
> builtin_vt52
> builtin_xterm
> builtin_iris-ansi
> builtin_debug
> builtin_dumb
> defaulting to 'ansi'
>
> ` `
>
> where the remote suse box reports
>
> ` ` `
> E437: terminal capability "cm" required
> Press ENTER or type command to continue
> ` ` `
>
> Can I add some code in my .screenrc to resolve this problem ?
> Please suggest.
> Thanks
>
>
>
> _______________________________________________
> screen-users mailing list
> screen-users@...
> http://lists.gnu.org/mailman/listinfo/screen-users
>

Pia Mikeal
System Administrator
Markey Center for Structural Biology
Purdue University
pmikeal@...
(765)496-6781

_______________________________________________
screen-users mailing list
screen-users@...
http://lists.gnu.org/mailman/listinfo/screen-users

Re: String escape for total number of windows?

2009-12-18T07:38:04Z

On Thu, Dec 10, 2009 at 10:25:11PM -0800, John Magolske wrote:
> Is there a string escape sequence that will display the total number
> of windows (just a single number, not all window numbers and names)?

There does not seem to be.

> For example, if I were in the 7th window and 23 windows were open,
> I'd like to see something like
>
> 7 / 23 : vim ~/.mutt/temp/mutt-1000-25786-33
>
> by setting hardstatus to something like
>
> hardstatus alwayslastline "%n / %x : %t"
>
> Of course %x is not an escape for anything, I'm just using it here as
> a placeholder...there doesn't seem to be an escape for total number of
> windows. Might it be possible to achieve this by truncating the output
> of %w or some such thing?

Looking at the code, at least with my eyes, truncation seems difficult
and will probably end up as a hack.

I've attached a patch that copies some of the functionality of the '%w'
handling, and should be applied to the newest source on savannah. It
seems to produce the correct result, although someone more familiar with
the code should check it for any unwanted consequences.

[...]

--
HTH
Thor

diff --git a/src/extern.h b/src/extern.h index 0eb811a..a0c0249 100644 --- a/src/extern.h +++ b/src/extern.h @@ -207,6 +207,7 @@ extern int IsNumColon __P((char *, int, char *, int)); extern void ShowWindows __P((int)); extern char *AddWindows __P((char *, int, int, int)); extern char *AddWindowFlags __P((char *, int, struct win *)); +extern int NumberOfWindows __P((int)); extern char *AddOtherUsers __P((char *, int, struct win *)); extern int WindowByNoN __P((char *)); extern struct win *FindNiceWindow __P((struct win *, char *)); diff --git a/src/process.c b/src/process.c index 0e835f0..eb625dc 100644 --- a/src/process.c +++ b/src/process.c @@ -5317,6 +5317,26 @@ int where; return ss; } +int +NumberOfWindows(flags) +int flags; +{ + int number_of_windows = 1; + struct win **pp, *p; + + for (pp = wtab; pp < wtab + MAXWIN; pp++) + { + if ((p = *pp) == 0) + continue; + if ((flags & 1) && display && p == D_fore) + continue; + if (D_fore && D_fore->w_group != p->w_group) + continue; + number_of_windows++; + } + return number_of_windows; +} + char * AddWindowFlags(buf, len, p) char *buf; diff --git a/src/screen.c b/src/screen.c index edec34a..38b512a 100644 --- a/src/screen.c +++ b/src/screen.c @@ -2955,6 +2955,13 @@ int rec; numpad++; } break; + case 'x': + { + int number_of_windows = NumberOfWindows((*s == 'w' ? 0 : 1) | (longflg ? 0 : 2) | (plusflg ? 4 : 0) | (minusflg ? 8 : 0)); + sprintf(p, "%d", number_of_windows); + p += strlen(p) - 1; + } + break; case 'n': s++; /* FALLTHROUGH */

_______________________________________________
screen-users mailing list
screen-users@...
http://lists.gnu.org/mailman/listinfo/screen-users

Re: search in copy/scrollback mode when using vertical split

2009-12-18T05:40:31Z

* Jacob Alexandersson had this to say on [18 Dec 2009, 14:28:42 +0100]:

> On Thu, Dec 17, 2009 at 6:16 PM, Sadrul Habib Chowdhury <imadil@...>wrote:
>
> > * Jacob Alexandersson had this to say on [15 Dec 2009, 10:02:18 +0100]:
> > > Hello all,
> > >
> > > I'm using screen from the ubuntu repositories which has the
> > vertical-screen
> > > patch. I just noticed that when I've got the screen in vertical-split
> > mode,
> > > the search function (/ and ?) doesn't work properly. The searches find
> > > nothing even though matches are there [snip]
> >
> > Does screen tell you that 'Pattern not found' even when there are
> > matches? Or does it not do anything at all?
> >
>
> It does _not_ tell me "Pattern not found"; it says nothing. When I hit "n",
> it will find the next match.... but it doesnt work like that when I'm not in
> "split mode"; then it matches right away after doing / <search> .....
>
> The cursor never moves (even if there is a match) after doing / <search> ...
> even when there is a match.

Yes, this is the same behaviour as I had described in my mail. It has
been fixed (commit 71cd5c06c6).

Thanks for the bug report.

Cheers,
Sadrul

_______________________________________________
screen-users mailing list
screen-users@...
http://lists.gnu.org/mailman/listinfo/screen-users

Re: search in copy/scrollback mode when using vertical split

2009-12-18T05:28:42Z

On Thu, Dec 17, 2009 at 6:16 PM, Sadrul Habib Chowdhury <imadil@...> wrote:

* Jacob Alexandersson had this to say on [15 Dec 2009, 10:02:18 +0100]:

> Hello all,
>
> I'm using screen from the ubuntu repositories which has the vertical-screen
> patch. I just noticed that when I've got the screen in vertical-split mode,
> the search function (/ and ?) doesn't work properly. The searches find

> nothing even though matches are there [snip]

Does screen tell you that 'Pattern not found' even when there are
matches? Or does it not do anything at all?

It does _not_ tell me "Pattern not found"; it says nothing. When I hit "n", it will find the next match.... but it doesnt work like that when I'm not in "split mode"; then it matches right away after doing / <search> .....

The cursor never moves (even if there is a match) after doing / <search> ... even when there is a match.

> Anyone seen this? Any workaround?

I see that screen does not update the cursor properly for the first match
in vertical mode (although the cursor is moved there, as can be seen by
using the cursor-movement keys). I am investigating why this happens.
For the subsequent searches (pressing 'n'/'N'), the cursor is updated
properly and everything works as it should.

Note that I am using the development version of screen (from
https://savannah.gnu.org/git/?group=screen).

Cheers,
Sadrul

_______________________________________________
screen-users mailing list
screen-users@...
http://lists.gnu.org/mailman/listinfo/screen-users

Re: Cannot set $TERM in screen -- it always defaults to xterm

2009-12-18T01:36:37Z

On Tue, Dec 15, 2009 at 05:15:22AM +0000, Leo Alekseyev wrote:
> On Ubuntu 9.10, GNU screen ignores any attempt to set the TERM variable --
> either by setting term="xterm-256color" in .screenrc, or using :term command in
> screen itself, or using a -T command line switch. Whatever I do, $TERM is stuck
> at xterm. Somehow screen is able to display 256 colors under gnome-terminal, but
> under xterm it defaults to 8. I tried reinstalling from the repositories and
> compiling from source, and I still see this behavior. Can someone explain what
> is going on and how to fix it? TIA.

What shell are you using, and does the problem persist if you start your
shell without config files?

--
Thor

_______________________________________________
screen-users mailing list
screen-users@...
http://lists.gnu.org/mailman/listinfo/screen-users

Re: search in copy/scrollback mode when using vertical split

2009-12-17T10:31:21Z

* Sadrul Habib Chowdhury had this to say on [17 Dec 2009, 12:16:50 -0500]:

> * Jacob Alexandersson had this to say on [15 Dec 2009, 10:02:18 +0100]:
> > Hello all,
> >
> > I'm using screen from the ubuntu repositories which has the vertical-screen
> > patch. I just noticed that when I've got the screen in vertical-split mode,
> > the search function (/ and ?) doesn't work properly. The searches find
> > nothing even though matches are there [snip]
>
> Does screen tell you that 'Pattern not found' even when there are
> matches? Or does it not do anything at all?
>
> > Anyone seen this? Any workaround?
>
> I see that screen does not update the cursor properly for the first match
> in vertical mode (although the cursor is moved there, as can be seen by
> using the cursor-movement keys).

First, this particular glitch shows up with any kind of splits, i.e. it
will happen even if you have horizontal splits only.

Second, I have pushed a fix for this (revision 71cd5c06c).

Cheers,
Sadrul

_______________________________________________
screen-users mailing list
screen-users@...
http://lists.gnu.org/mailman/listinfo/screen-users

Re: search in copy/scrollback mode when using vertical split

2009-12-17T09:16:50Z

* Jacob Alexandersson had this to say on [15 Dec 2009, 10:02:18 +0100]:
> Hello all,
>
> I'm using screen from the ubuntu repositories which has the vertical-screen
> patch. I just noticed that when I've got the screen in vertical-split mode,
> the search function (/ and ?) doesn't work properly. The searches find
> nothing even though matches are there [snip]

Does screen tell you that 'Pattern not found' even when there are
matches? Or does it not do anything at all?

> Anyone seen this? Any workaround?

I see that screen does not update the cursor properly for the first match
in vertical mode (although the cursor is moved there, as can be seen by
using the cursor-movement keys). I am investigating why this happens.
For the subsequent searches (pressing 'n'/'N'), the cursor is updated
properly and everything works as it should.

Note that I am using the development version of screen (from
https://savannah.gnu.org/git/?group=screen).

Cheers,
Sadrul

_______________________________________________
screen-users mailing list
screen-users@...
http://lists.gnu.org/mailman/listinfo/screen-users

Re: Copy-scrollback mode -- VIsual feedback in window caption?

2009-12-17T08:10:10Z

On Thursday 17 December 2009 3:24:20 am Chris Jones wrote:

> On Thu, Dec 17, 2009 at 01:28:50AM EST, Alexander Steinert wrote:
> > Chris Jones, Tue 2009-12-15 23:45 CET:
> > > Maybe the new feature was not necessary after all, since it's quite
> > > possible to acquire the habit of hitting '^A m' when coming back to
> > > one's screen session, as in "now, let's see.. where was I"..
> >
> > ^G is much more helpful, IMHO, assuming that you don't have an
> > application running where that could be "harmful".
>
> I agree.. although I can't think of any applications I run on a regular
> basis taking 'g' as a command, apart from ELinks, where it pops up a
> dialog prompting you for a URL, it feels like an accident to happen.
>
> Unless you go through all your favorite applications and remap 'g' to
> displaying a message such as "no.. you are not in screen's copy mode",
> naturally :-)
>

Doing '^A [' should either get you into copy mode (which can be promptly
exited with '[') or, if already on copy mode, tells you "Must be on a window
layer".

_______________________________________________
screen-users mailing list
screen-users@...
http://lists.gnu.org/mailman/listinfo/screen-users

Re: terminal capability problem when login therough screen

2009-12-17T01:25:05Z

J. Bakshi wrote:

I get one clue. No problem if I use urxvt as terminal but mrxvt
produces the above error.

--
জয়দীপ বক্সী

_______________________________________________
screen-users mailing list
screen-users@...
http://lists.gnu.org/mailman/listinfo/screen-users

terminal capability problem when login therough screen

2009-12-17T01:02:01Z

Hello list,

A strange issue which I encounter during ssh through screen. If I ssh to
my remote servers ( one debian and one suse) I have no problem at all to
use vim. But if I *ssh through screen* then the vim on remote debian
box reports

` ` `
E558: Terminal entry not found in terminfo
'screen.rxvt' not known. Available builtin terminals are:
builtin_riscos
builtin_amiga
builtin_beos-ansi
builtin_ansi
builtin_pcansi
builtin_win32
builtin_vt320
builtin_vt52
builtin_xterm
builtin_iris-ansi
builtin_debug
builtin_dumb
defaulting to 'ansi'

` `

where the remote suse box reports

` ` `
E437: terminal capability "cm" required
Press ENTER or type command to continue
` ` `

Can I add some code in my .screenrc to resolve this problem ?
Please suggest.
Thanks

_______________________________________________
screen-users mailing list
screen-users@...
http://lists.gnu.org/mailman/listinfo/screen-users

terminal capability problem when login therough screen

2009-12-17T00:56:40Z

Hello list,

A strange issue which I encounter during ssh through screen. If I ssh to my remote servers ( one debian and one suse) I have no problem at all to use vim. But if I *ssh through screen* then the vim on remote debian box reports

` ` `
E558: Terminal entry not found in terminfo
'screen.rxvt' not known. Available builtin terminals are:
builtin_riscos
builtin_amiga
builtin_beos-ansi
builtin_ansi
builtin_pcansi
builtin_win32
builtin_vt320
builtin_vt52
builtin_xterm
builtin_iris-ansi
builtin_debug
builtin_dumb
defaulting to 'ansi'

` `

where the remote suse box reports

` ` `
E437: terminal capability "cm" required
Press ENTER or type command to continue
` ` `

Can I add some code in my .screenrc to resolve this problem ?
Please suggest.
Thanks

Re: Copy-scrollback mode -- VIsual feedback in window caption?

2009-12-17T00:24:20Z

On Thu, Dec 17, 2009 at 01:28:50AM EST, Alexander Steinert wrote:
> Chris Jones, Tue 2009-12-15 23:45 CET:

> > Maybe the new feature was not necessary after all, since it's quite
> > possible to acquire the habit of hitting '^A m' when coming back to
> > one's screen session, as in "now, let's see.. where was I"..
>
> Note that lastmsg does not necessarily refer to the window that is
> currently selected.

Ah, good point..!

I wondered about that after posting, but I was already in bed and by the
morrow I had forgotten about it.

> ^G is much more helpful, IMHO, assuming that you don't have an
> application running where that could be "harmful".

I agree.. although I can't think of any applications I run on a regular
basis taking 'g' as a command, apart from ELinks, where it pops up a
dialog prompting you for a URL, it feels like an accident to happen.

Unless you go through all your favorite applications and remap 'g' to
displaying a message such as "no.. you are not in screen's copy mode",
naturally :-)

CJ

_______________________________________________
screen-users mailing list
screen-users@...
http://lists.gnu.org/mailman/listinfo/screen-users

Re: Copy-scrollback mode -- VIsual feedback in window caption?

2009-12-16T22:28:50Z

Chris Jones, Tue 2009-12-15 23:45 CET:
> Maybe the new feature was not necessary after all, since it's quite
> possible to acquire the habit of hitting '^A m' when coming back to
> one's screen session, as in "now, let's see.. where was I"..

Note that lastmsg does not necessarily refer to the window that is
currently selected.

^G is much more helpful, IMHO, assuming that you don't have an
application running where that could be "harmful".

Stony

_______________________________________________
screen-users mailing list
screen-users@...
http://lists.gnu.org/mailman/listinfo/screen-users

Re: Detached screens disappear - help debugging?

2009-12-16T19:29:17Z

>> After detaching, my screen session will disappear some time after.

Is the screen process dead (check using ps) or you are not just able
to attach to the screen.

If the screen process is alive and "screen -ls" tells "No sockets found"
then it may the case that your sockdir (usually /tmp/uscreens/<user>)
is getting cleaned up!

My admin was wiping off the entire /tmp when diskspace was low and
this was causing me a similar issue.

Regards
Pandu

On Wed, Dec 16, 2009 at 11:06 AM, Leinartas, Michael
<MICHAEL.LEINARTAS@...> wrote:

_______________________________________________
screen-users mailing list
screen-users@...
http://lists.gnu.org/mailman/listinfo/screen-users

Detached screens disappear - help debugging?

2009-12-15T21:36:18Z

Detached screens disappear - help debugging? I've had a problem with screen on my home computer for quite some time that's been driving me crazy. After detaching, my screen session will disappear some time after. Sometimes an hour later it's there, sometimes it's not. It's nearly always gone if I leave it overnight. I never have any problems if I'm attached - if I say, leave a screen session on the console open for 3 days, it'll be fine. It also doesn't seem to matter how many terminals are in the session or whether they're idle at a prompt or running something.

So where can I start figuring out what's causing these sessions to die? It doesn't seem to be anything obvious like an oom killer (nothing in dmesg). I've checked the system crons in /etc/cron.* and it doesn't look like there's anything wiping stuff in /var/run or anything.

Some system details incase it matters to anyone:
Screen 4.0.3
Debian squeeze (same problem when I was on lenny) amd64 arch
linux-2.6.32 (vanilla)
AMD quad core (Phenom II)
4gb ram

I'd appreciate any help on this because it's driving me nuts.

michael

_______________________________________________
screen-users mailing list
screen-users@...
http://lists.gnu.org/mailman/listinfo/screen-users

Re: Copy-scrollback mode -- VIsual feedback in window caption?

2009-12-15T14:45:33Z

On Tue, Dec 15, 2009 at 05:13:57PM EST, Sadrul Habib Chowdhury wrote:
> * Chris Jones had this to say on [15 Dec 2009, 16:22:07 -0500]:

[..]

> > Although what I had in mind was something that would display permanently
> > and that would refresh itself automatically when the cursor is moved,
> > similar to Vim's 'statusline' items 'l' and 'v'.
>
> There's nothing in the existing code that will make that happen. From the
> earlier mail:
>
> A vim-like caption could indeed be useful. It can be done by, for
> example, interpreting the escape sequences differently when in
> copy/mode. But I really really don't want that to happen. I think
> it'd be much better to leave that kind of thing to scripts (when the
> support for that gets a bit more mature).

Yes, I noticed that.

Maybe the new feature was not necessary after all, since it's quite
possible to acquire the habit of hitting '^A m' when coming back to
one's screen session, as in "now, let's see.. where was I"..

CJ

_______________________________________________
screen-users mailing list
screen-users@...
http://lists.gnu.org/mailman/listinfo/screen-users

Re: Copy-scrollback mode -- VIsual feedback in window caption?

2009-12-15T14:06:09Z

* Chris Jones had this to say on [15 Dec 2009, 16:22:07 -0500]:

> On Tue, Dec 15, 2009 at 04:10:42PM EST, Sadrul Habib Chowdhury wrote:
> > * Chris Jones had this to say on [15 Dec 2009, 15:08:57 -0500]:
> > [snip]
>
> > >
> > > Without additional changes to the code, is there any way I could
> > > display the x,y coordinates of the cursor position as the transient
> > > message does?
> > >
> > > Something like:
> > >
> > > ---------------------------------------------------------------------
> > > Copy-mode Column xxx, Line yyy
> > > ---------------------------------------------------------------------
> > >
> > > Useful in contexts where the cursor is difficult to spot. I
> > > sometimes have this problem on some web pages with the ELinks web
> > > browser.
> > >
> > > Naturally, what I mean is the x,y within the page that's displayed
> > > in the view port, not relative to the entire scroll-back buffer.
>
> > Pressing Ctrl+G in copy-mode will show that.
>
> Oh, very useful, and vim-like as well.. very nice.
>
> Although what I had in mind was something that would display permanently
> and that would refresh itself automatically when the cursor is moved,
> similar to Vim's 'statusline' items 'l' and 'v'.

There's nothing in the existing code that will make that happen. From the
earlier mail:

A vim-like caption could indeed be useful. It can be done by, for
example, interpreting the escape sequences differently when in
copy/mode. But I really really don't want that to happen. I think
it'd be much better to leave that kind of thing to scripts (when the
support for that gets a bit more mature).

Cheers,
Sadrul

_______________________________________________
screen-users mailing list
screen-users@...
http://lists.gnu.org/mailman/listinfo/screen-users

Re: Copy-scrollback mode -- VIsual feedback in window caption?

2009-12-15T13:22:07Z

On Tue, Dec 15, 2009 at 04:10:42PM EST, Sadrul Habib Chowdhury wrote:
> * Chris Jones had this to say on [15 Dec 2009, 15:08:57 -0500]:
> [snip]

> >
> > Without additional changes to the code, is there any way I could
> > display the x,y coordinates of the cursor position as the transient
> > message does?
> >
> > Something like:
> >
> > ---------------------------------------------------------------------
> > Copy-mode Column xxx, Line yyy
> > ---------------------------------------------------------------------
> >
> > Useful in contexts where the cursor is difficult to spot. I
> > sometimes have this problem on some web pages with the ELinks web
> > browser.
> >
> > Naturally, what I mean is the x,y within the page that's displayed
> > in the view port, not relative to the entire scroll-back buffer.

> Pressing Ctrl+G in copy-mode will show that.

Oh, very useful, and vim-like as well.. very nice.

Although what I had in mind was something that would display permanently
and that would refresh itself automatically when the cursor is moved,
similar to Vim's 'statusline' items 'l' and 'v'.

Since I already use Ctrl-G as my screen escape key, I need to do a
Ctrl-g g, of course.

> Also, thanks for testing the patch.

You're welcome.

CJ

_______________________________________________
screen-users mailing list
screen-users@...
http://lists.gnu.org/mailman/listinfo/screen-users

Re: Copy-scrollback mode -- VIsual feedback in window caption?

2009-12-15T13:02:44Z

* Chris Jones had this to say on [15 Dec 2009, 15:08:57 -0500]:
[snip]

>
> Without additional changes to the code, is there any way I could display
> the x,y coordinates of the cursor position as the transient message
> does?
>
> Something like:
>
> ---------------------------------------------------------------------
> Copy-mode Column xxx, Line yyy
> ---------------------------------------------------------------------
>
> Useful in contexts where the cursor is difficult to spot. I sometimes
> have this problem on some web pages with the ELinks web browser.
>
> Naturally, what I mean is the x,y within the page that's displayed in
> the view port, not relative to the entire scroll-back buffer.

Pressing Ctrl+G in copy-mode will show that.

Also, thanks for testing the patch.

Cheers,
Sadrul

_______________________________________________
screen-users mailing list
screen-users@...
http://lists.gnu.org/mailman/listinfo/screen-users

Re: Copy-scrollback mode -- VIsual feedback in window caption?

2009-12-15T12:08:57Z

On Tue, Dec 15, 2009 at 12:19:09PM EST, Sadrul Habib Chowdhury wrote:

[..]

> I haven't committed the changes yet, still testing (patch attached,
> should someone want to try it out). Opinions?

Does a very nice job like so:

caption always "%?%P%{+u kw} copy-mode %{+u wk} %:%{+u wk} %?"
hardstatus alwayslastline "[%H]%L=%=%-w%45L>(%n-%t)%+w%=%-13=[%d/%m %c]"

The caption duplicates the beginning of the 'copy-mode' message.

Anyway, since my caption line was an empty that I use it as a separator
it was easy to fit into my setup.

The only possible problem with this approach might be for users who
already have a very long caption/status line who would end up with
something really huge.

Without additional changes to the code, is there any way I could display
the x,y coordinates of the cursor position as the transient message
does?

Something like:

---------------------------------------------------------------------
Copy-mode Column xxx, Line yyy
---------------------------------------------------------------------

Useful in contexts where the cursor is difficult to spot. I sometimes
have this problem on some web pages with the ELinks web browser.

Naturally, what I mean is the x,y within the page that's displayed in
the view port, not relative to the entire scroll-back buffer.

In any event, thanks much for the patch, works for me.

Thanks,

CJ

_______________________________________________
screen-users mailing list
screen-users@...
http://lists.gnu.org/mailman/listinfo/screen-users

Unable to set $TERM in screen (always defaults to xterm)

2009-12-15T09:34:26Z

On Ubuntu 9.10, GNU screen ignores any attempt to set the TERM
variable -- either by setting term="xterm-256color" in .screenrc, or
using :term command in screen itself, or using a -T command line
switch. Whatever I do, $TERM is stuck at xterm. Somehow screen is able
to display 256 colors under gnome-terminal, but under xterm it
defaults to 8. I tried reinstalling from the repositories and
compiling from source, and I still see this behavior. Can someone
explain what is going on and how to fix it? TIA.

_______________________________________________
screen-users mailing list
screen-users@...
http://lists.gnu.org/mailman/listinfo/screen-users

Re: Copy-scrollback mode -- VIsual feedback in window caption?

2009-12-15T09:14:31Z

* Hugo Heden had this to say on [15 Dec 2009, 11:03:39 +0100]:
[snip]

> > > Not sure this sounds like an obscure use case.. But I would love it if
> > > I could configure some visual feedback for when being in
> > > copy-scroll-back-mode.)
> >
> > Otherwise, if you really feel visual feedback is a good idea, I guess a
> > permanent blinking message in reverse video in the current status area
> > would probably do the trick and cover all the bases, even for thos who
> > do no use a caption/status line (?)
> >
>
> Yes, that sounds reasonable, though I *think* it'd be enough with something
> that isn't actually blinking.

Something configurable sounds like a good idea to me. That way, users who
don't want to don't need to adapt to a change.

So, I decided to go with adding cruft to the caption string (sigh*). I
have added 'P' flag, which will set '%?' to true if the current window is
in copy/paste mode. So, if you want to simply change the color etc. of
your caption, prepend '%?%P%{XXX}%?' to your caption and you're set ('XXX'
stands for the usual attribute/color modifier). If, on the other hand,
you want to change the caption string in copy mode, change your caption
string to '%?%PCaption for copy mode%:Your usual caption%?'

* I sigh because (1) I really didn't want to add more cruft to the
already-too-complicated caption string, and (2) this fix doesn't help
someone who doesn't have a caption string.

> Something Vim like perhaps (as suggested by
> Dustin Kirkland elsewhere in this thread)? (Though I'm not sure what the
> "line number" would be helpful for.. Should the lines be counted from the
> top or the bottom of the scrollback buffer?)

A vim-like caption could indeed be useful. It can be done by, for
example, interpreting the escape sequences differently when in copy/mode.
But I really really don't want that to happen. I think it'd be much
better to leave that kind of thing to scripts (when the support for that
gets a bit more mature).

I haven't committed the changes yet, still testing (patch attached,
should someone want to try it out). Opinions?

Cheers,
Sadrul

diff --git a/src/mark.c b/src/mark.c
index ce94293..83eaed1 100644
--- a/src/mark.c
+++ b/src/mark.c
@@ -1035,6 +1035,7 @@ processchar:
LAY_CALL_UP(LRefreshAll(flayer, 0));
}
ExitOverlayPage();
+ WindowChanged(fore, 'P');
if (append_mode)
LMsg(0, "Appended %d characters to buffer",
newcopylen);
@@ -1277,6 +1278,7 @@ MarkAbort()
rem(markdata->x1, markdata->y1, markdata->cx, markdata->cy, redisp, (char *)0, yend);
}
ExitOverlayPage();
+ WindowChanged(fore, 'P');
}

diff --git a/src/process.c b/src/process.c
index 2874e7e..0e835f0 100644
--- a/src/process.c
+++ b/src/process.c
@@ -2231,6 +2231,7 @@ int key;
break;
}
MarkRoutine();
+ WindowChanged(fore, 'P');
break;
case RC_HISTORY:
{
diff --git a/src/screen.c b/src/screen.c
index e8f4cac..10819c9 100644
--- a/src/screen.c
+++ b/src/screen.c
@@ -118,6 +118,8 @@ int VBellWait, MsgWait, MsgMinWait, SilenceWait;
extern struct acluser *users;
extern struct display *displays, *display;

+extern struct LayFuncs MarkLf;
+

extern int visual_bell;
#ifdef COPY_PASTE
@@ -2804,6 +2806,16 @@ int rec;
if (minusflg)
qmflag = 1;
break;
+ case 'P':
+ p--;
+ if (display && ev && ev != &D_hstatusev) /* Hack */
+ {
+ /* Is the layer in the current canvas in copy mode? */
+ struct canvas *cv = (struct canvas *)ev->data;
+ if (ev == &cv->c_captev && cv->c_layer->l_layfn == &MarkLf)
+ qmflag = 1;
+ }
+ break;
case '>':
truncpos = p - winmsg_buf;
truncper = num > 100 ? 100 : num;

_______________________________________________
screen-users mailing list
screen-users@...
http://lists.gnu.org/mailman/listinfo/screen-users