Skip to main content

optionally run several commands when starting tmux


I have a ~/.tmux/dev file which looks like this:


selectp -t 0    # select the first pane
splitw -h -p 50 # split it into two halves

selectp -t 1 # select the new, second (1) pane
splitw -v -p 50 # split it into two halves

selectp -t 0 # go back to the first pane
send-keys 'vim .' Enter

If I run tmux source-file ~/.tmux/dev from inside a tmux session, it works fine.


I want to create a bash script that opens tmux and automatically runs those commands.


I tried running


tmux new 'tmux source-file ~/.tmux/dev'

but I only see two panes; the pane with vim is missing.


I have read How do I get tmux to open up a set of panes without manually entering them?, but before installing a tool for managing tmux configs, I would like to grok what's happening here.


Thanks in advance for any light you can shed on the subject.



Answer



This is what is happening:




  1. tmux new 'tmux source-file ~/.tmux/dev'


    The new commands creates a new session with a single window that has a single pane. The command tmux source-file ~/.tmux/dev runs in this new pane.



    • So, you have a new session N (where N is some number), with

    • a single window N:0 (or whatever you have base-index set to), with

    • a single pane N:0.0 (or whatever you have base-pane-index set to),

    • running the command tmux source-file ~/.tmux/dev.




  2. The source-file command is processed.



    1. The extra panes are added.

    2. Pane 0 is (re)selected.

    3. The send-keys command then “types” vim . + Enter at pane 0.
      This input is ignored because this pane is just running the tmux client that sent the source-file command.



  3. The tmux client exits, thus closing pane 0.


So the unexpected bit is that pane 0 (i.e. N:0.0) is running (only) the source-file command which ignores your “typed” command. This pane never runs an interactive shell that could interpret the “typed” command.


There are at least a couple of ways you can fix this:




  • Start ~/.tmux/dev with new-window so that pane 0 is running your “default command” (i.e. probably an interactive shell).


    This method has the benefit of not assuming that your current pane is running an interactive shell, and also not assuming that the current pane is 0 (i.e. what happens if you run your original series of commands against a pane that is part of an already split window?). It means you can safely bind source-file ~/.tmux/dev to a key that you can run in any context (since it creates a new window for all of its panes). From the shell you can run either tmux source-file ~/.tmux/dev (to create a new window in the current session), or your original tmux new 'tmux source-file ~/.tmux/dev' to create a new session.


    A minor drawback to this method is that when you run tmux new 'tmux source-file ~/.tmux/dev', the initial window will still run the client that sends source-file and exit fairly quickly. This means that your “main window” (the one with the splits) will be one higher than your base-index and a future new window will be placed before the “main window”. You could fix this by using something like this:


    tmux new 'tmux move-window -t 99 \; source-file ~/.tmux/dev'

    It moves the (ephemeral) initial window to a high index so that the new-window in ~/.tmux/dev will end up at base-index.




  • Use (e.g.) tmux new 'tmux source-file ~/.tmux/dev ; zsh -l' so that the pane ends up running an interactive shell after the source-file command finishes.


    The ugly bit about this is that you end up “hard coding” your preferred shell into this command. Also, the send-keys input (vim . + Enter) is technically sent before the shell starts; this is probably okay, but may not always be completely reliable.


    You could avoid “hard coding” your shell by querying tmux for the default-command (or if that is not set, default-shell (or if that is not set, using SHELL)), but that may be more work than you really want to do.




Comments

Popular Posts

How do I transmit a single hexadecimal value serial data in PuTTY using an Alt code?

I am trying to sent a specific hexadecimal value across a serial COM port using PuTTY. Specifically, I want to send the hex codes 9C, B6, FC, and 8B. I have looked up the Alt codes for these and they are 156, 182, 252, and 139 respectively. However, whenever I input the Alt codes, a preceding hex value of C2 is sent before 9C, B6, and 8B so the values that are sent are C2 9C, C2 B6, and C2 8B. The value for FC is changed to C3 FC. Why are these values being placed before the hex value and why is FC being changed altogether? To me, it seems like there is a problem internally converting the Alt code to hex. Is there a way to directly input hex values without using Alt codes in PuTTY? Answer What you're seeing is just ordinary text character set conversion. As far as PuTTY is concerned, you are typing (and reading) text , not raw binary data, therefore it has to convert the text to bytes in whatever configured character set before sending it over the wire. In other words, when y...

linux - Extract/save a mail attachment using bash

Using normal bash tools (ie, built-ins or commonly-available command-line tools), is it possible, and how to extract/save attachments on emails? For example, say I have a nightly report which arrives via email but is a zip archive of several log files. I want to save all those zips into a backup directory. How would I accomplish that? Answer If you're aiming for portability, beware that there are several different versions of mail(1) and mailx(1) . There's a POSIX mailx command, but with very few requirements. And none of the implementations I have seem to parse attachments anyway. You might have the mpack package . Its munpack command saves all parts of a MIME message into separate files, then all you have to do is save the interesting parts and clean up the rest. There's also metamail . An equivalent of munpack is metamail -wy .

ubuntu - Why does my USB hdd returns SG_IO: bad/missing sense data?

I am able to boot and run commands from external USB hdd; the message in question appears for about 45 seconds then booting continues. GRUB2 is installed on internal HDD. When choosing to boot directly to /dev/sdb the message doesn't appear, however boot time is about the same as booting to internal HDD. /dev/sdb: Timing cached reads: 1018 MB in 2.00 seconds = 508.97 MB/sec Timing buffered disk reads: 80 MB in 3.03 seconds = 26.37 MB/sec pfeiffep@de:~$ sudo hdparm -i /dev/sdb /dev/sdb: SG_IO: bad/missing sense data, sb[]: 70 00 05 00 00 00 00 10 00 00 00 00 20 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 HDIO_GET_IDENTITY failed: Invalid argument Gparted correctly identifies the drive as SAMSUNG MP0402H. Any ideas how to remedy the HDIO & SG_IO messages?

Desktop reboots itself on sleep or hibernate

I have been using an ASUS M2NPV-VM motherboard for main home desktop workstation, operating Windows Vista x64. This computer has right from day one not been able to enter hibernate or standby; after Windows performs its final actions and brings the machine down, it would automatically revive itself for a reboot. Updating to the second latest BIOS (1201)has not helped (the latest BIOS revision would induce video refresh problems rendering it unusable). I have been reading related discussions on incidents similar to mine to no avail of a true workable solution. They appear to be more speculative guesses rather than actual knowledge on the inner workings of motherboard hardware. Does anybody have any electronic engineering experience on PC energy-saving standards to provide a more informed opinion how to go about getting this to work? More stories: this motherboard could not even reboot properly the first thing i used it. It was due to refresh rate of the onboard GPU, which had no influe...