How to recursively exit nested shells? - bash

I use the Ranger file manager in my terminal to move around. Every time I use the S command to drop into a new directory, Ranger is actually launching a new shell. When I want to close a terminal window I need to run exit as many times as I have changed directories with Ranger. Is there a command that will run exit recursively for me until the window closes? Or better yet is there a different Ranger command to use?

Don't enter a sub-shell, just quit ranger and let the shell sync the directory back from ranger.
function ranger {
local IFS=$'\t\n'
local tempfile="$(mktemp -t tmp.XXXXXX)"
local ranger_cmd=(
command
ranger
--cmd="map Q chain shell echo %d > "$tempfile"; quitall"
)
${ranger_cmd[#]} "$#"
if [[ -f "$tempfile" ]] && [[ "$(cat -- "$tempfile")" != "$PWD" ]]; then
cd -- "$(cat "$tempfile")" || return
fi
command rm -f -- "$tempfile" 2>/dev/null
}
Press capital Q to quit ranger, after which the shell will sync directory automatically to the same one within ranger.
This is very flexible and you can use q to quit normally without syncing the dir back to the shell.
Update:
. ranger to open ranger is another solution, which is NOT recommended. Because compared with previous method, quitting from . ranger with q will always sync the dir back to the shell from ranger. You have no control on this behavior.
No real solution to this. It's oft requested but not as easy as people think.
You can source ranger, so start it like this . ranger. That'll cd to ranger's cwd when quitting.
References
Changing directories from ranger wiki
How to exit and cd into last dir you were on ranger? #1554

#Simba thanks for the great answer and the link to the docs.
In the end, the easiest answer was to just create an alias in my .bashrc like the docs recommend
alias ranger=". ranger"
Now when I use Q to quit, it automatically switches to the new directory, and only requires running exit one time to close the window.

Related

Hot reload on save

I'm currently using a terminal and vim on OSX as a development environment for Flutter. Things are going pretty well except that the app does not reload when I save any dart files. Is there a way to trigger that behavior?Currently I have to go to the terminal and hit "r" to see my changes.
Sorry for the plug, but I wrote a very simple plugin to handle this.
It makes use of Flutter's --pid-file command line flag to send it a SIGUSR1 signal.
You can achieve the same result as my two-line plugin by adding this to an autocmd
silent execute '!kill -SIGUSR1 "$(cat /tmp/flutter.pid)"'
And launching Flutter with the --pid-file flag.
I made a vim plugin hankchiutw/flutter-reload.vim based on killing with SIGUSR1.
You don't have to use --pid-file flag with this plugin. (Thanks to the pgrep :))
Simply execute flutter run, modify your *.dart file and see the reloading.
I did it with the excellent little tool called entr. On OS/X you can install it from brew: brew install entr. The home page of the tool is at http://eradman.com/entrproject/
Then you start flutter run with the pidfile as #nobody_nowhere suggests.
How do you run entr depends on the level of service. In the simplest case you just do find lib/ -name '*.dart' | entr -p kill -USR1 $(cat /tmp/flutter.pid)
But such invocation will not detect new files in the source tree (because find builds a list of files to watch only once, at the start). You can get away with slightly more complex one-liner:
while true
do
find lib/ -name '*.dart' | \
entr -d -p kill -USR1 $(cat /tmp/flutter.pid)
done
The -d option makes entr exit when it does detect a new file in one of the directories and the loop runs again.
I personally use even more complex approach. I use Redux and change to middleware or other state files does not work with hot reload, it doesn't pick up these changes. So you need to resort to hot restart.
I have a script hotrestarter.sh:
#!/bin/bash
set -euo pipefail
PIDFILE="/tmp/flutter.pid"
if [[ "${1-}" != "" && -e $PIDFILE ]]; then
if [[ "$1" =~ \/state\/ ]]; then
kill -USR2 $(cat $PIDFILE)
else
kill -USR1 $(cat $PIDFILE)
fi
fi
It checks if the modified file lives in /state subdirectory and if true does hot restart or else hot reload. I call the script like that:
while true
do
find lib/ -name '*.dart' | entr -d -p ./hotreloader.sh /_
done
The /_ parameter makes entr to pass the name of the file to the program being invoked.
You don't say what platform, but all platforms have a "watcher" app that can run a command when any file in a tree changes. You'll need to run one of those.
vscode has this feature. If you don't mind moving to vscode you can get it out of the box. You could also reach out to the author and see if they have any suggestions on how you could do it in vim or check the source directly. Most likely vim will have a mechanism to do so.

Leaving a Command Running on a Remote Server

Forgive me if this is something I've just completely missed, however, I have a remote server (a NAS) that I'd like to start a command running on, while I do some work locally. Now, I believe I could probably do this with a command like:
ssh foo#bar 'cp -Rl /foo/bar /bar/foo'
However, I need a return value in my main script from part of the command, so I need it to return but leave the cp command running. For example:
foo=$(ssh foo#bar <<- REMOTE_COMMANDS
cp -Rl /foo/bar /bar/foo &
echo "foobar"
REMOTE_COMMANDS)
However I don't believe this returns until the cp command has completed, but if I use exit I think the cp is interrupted?
Is there another way to leave cp running, or will I need to run two ssh commands (one for the cp, one to get the return value I need?)
You can use one the following choices :
tmux
nohup
screen
tmux & screen are some complete environments that can be attached and detached for 1 to N users.
If you need something straightforward, look nohup first.
You can use screen command.
Simply create a new screen using : screen -R screen_name.
Run your command or code and then exit that screen by pressing ctrl + a + d.
If you want to switch back to the screen, enter this command : screen -r screen_name.
Hope it helps.

use alias only once in bashrc

I wrote an alias in my .bashrc file that open a txt file every time I start bash shell.
The problem is that I would like to open such a file only once, that is the first time I open the shell.
Is there any way to do that?
The general solution to this problem is to have a session lock of some kind. You could create a file /tmp/secret with the pid and/or tty of the process which is editing the other file, and remove the lock file when done. Now, your other sessions should be set up to not create this file if it already exists.
Proper locking is a complex topic, but for the simple cases, this might be good enough. If not, google for "mutual exclusion". Do note that there may be security implications if you get it wrong.
Why are you using an alias for this? Sounds like the code should be directly in your .bashrc, not in an alias definition.
So if, say, what you have now in your .bashrc is something like
alias start_editing_my_project_work_hour_report='emacs ~/prj.txt &̈́'
start_editing_my_project_work_hour_report
unalias start_editing_my_project_work_hour_report
... then with the locking, and without the alias, you might end up with something like
# Obtain my UID on this host, and construct directory name and lock file name
uid=$(id -u)
dir=/tmp/prj-$uid
lock=$dir/pid.lock
# The loop will execute at most twice,
# but we don't know yet whether once is enough
while true; do
if mkdir -p "$dir"; then
# Yay, we have the lock!
( echo $$ >"$lock" ; emacs ~/prj.txt; rm -f "$lock" ) &
break
else
other=$(cat "$lock")
# If the process which created the UID is still live, do nothing
if kill -0 $other; then
break
else
echo "removing stale lock file dir (dead PID $other) and retrying" >&2
rm -rf "$dir"
continue
fi
fi
done

How can one specify a user's emacs init file to load with emacsclient?

Question
How can one specify a user's emacs init file to load with emacsclient?
emacsclient does not understand '-u user' and '-a ALTERNATE-EDITOR' does not allow me to quote it and provide '-u user'.
For example, running:
/usr/bin/emacsclient -n -a '/usr/bin/emacs -u <username>' ~<username>/notes.txt
returns
/usr/bin/emacsclient: error executing alternate editor "/usr/bin/emacs -u <username>"
Background
I'm using emacs version 23.1.1 and emacsclient version 23.1.
emacs itself supports '-u user' to load a specified user's init file.
I use the following bash function in my aliases file to invoke emacs
# a wrapper is needed to sandwich multiple command line arguments in bash
# 2>/dev/null hides
# "emacsclient: can't find socket; have you started the server?"
emacs_wrapper () {
if [ 0 -eq $# ]
then
/usr/bin/emacsclient -n -a /usr/bin/emacs ~<username>/notes.txt 2>/dev/null &
else
/usr/bin/emacsclient -n -a /usr/bin/emacs $* 2>/dev/null &
fi
}
alias x='emacs_wrapper'
Typing x followed by a list of files:
Connects to an existing emacs server if one is running, otherwise starts a new one
Executes as a background process
Opens the list of files or my notes file if no files are provided
This works great when I'm logged in as myself. However, many production boxes require me to log in as a production user. I've separated my aliases into a bash script, therefore I can get my aliases and bash functions by simply running.
. ~<username>/alias.sh
Unfortunately, this won't let me use my .emacs file (~<username>/.emacs) :(
This problem has been driving me crazy.
If you can't include command line arguments in your alternate editor specification, then simply write a shell script which does that for you, and supply that as the alternate editor argument instead.
#!/bin/sh
emacs -u (username) "$#"
The point of the server/client model is that you have an existing Emacs with its own set of configurations, and then you connect via one or more clients.
What should the client do if the server was already configured to show the menu bar (a global setting), and the client says not to show it? What if another client attaches saying to show it?
If you want to use different settings for Emacs, use different emacs sessions.

Root user/sudo equivalent in Cygwin?

I'm trying to run a bash script in Cygwin.
I get Must run as root, i.e. sudo ./scriptname errors.
chmod 777 scriptname does nothing to help.
I've looked for ways to imitate sudo on Cygwin, to add a root user, since calling "su" renders the error su: user root does not exist, anything useful, and have found nothing.
Anyone have any suggestions?
I answered this question on SuperUser but only after the OP disregarded the unhelpful answer that was at the time the only answer to the question.
Here is the proper way to elevate permissions in Cygwin, copied from my own answer on SuperUser:
I found the answer on the Cygwin mailing list. To run command with elevated privileges in Cygwin, precede the command with cygstart --action=runas like this:
$ cygstart --action=runas command
This will open a Windows dialogue box asking for the Admin password and run the command if the proper password is entered.
This is easily scripted, so long as ~/bin is in your path. Create a file ~/bin/sudo with the following content:
#!/usr/bin/bash
cygstart --action=runas "$#"
Now make the file executable:
$ chmod +x ~/bin/sudo
Now you can run commands with real elevated privileges:
$ sudo elevatedCommand
You may need to add ~/bin to your path. You can run the following command on the Cygwin CLI, or add it to ~/.bashrc:
$ PATH=$HOME/bin:$PATH
Tested on 64-bit Windows 8.
You could also instead of above steps add an alias for this command to ~/.bashrc:
# alias to simulate sudo
alias sudo='cygstart --action=runas'
You probably need to run the cygwin shell as Administrator. You can right click the shortcut and click run as administrator or go into the properties of the shortcut and check it in the compatability section. Just beware.... root permissions can be dangerous.
Building on dotancohen's answer I'm using an alias:
alias sudo="cygstart --action=runas"
Works as a charm:
sudo chown User:Group <file>
And if you have SysInternals installed you can even start a command shell as the system user very easily
sudo psexec -i -s -d cmd
I found sudo-for-cygwin, maybe this would work, it is a client/server application that uses a python script to spawn a child process in windows (pty) and bridges user's tty and the process I/O.
It requires python in windows and Python modules greenlet, and eventlet in Cygwin.
It seems that cygstart/runas does not properly handle "$#" and thus commands that have arguments containing spaces (and perhaps other shell meta-characters -- I didn't check) will not work correctly.
I decided to just write a small sudo script that works by writing a temporary script that does the parameters correctly.
#! /bin/bash
# If already admin, just run the command in-line.
# This works on my Win10 machine; dunno about others.
if id -G | grep -q ' 544 '; then
"$#"
exit $?
fi
# cygstart/runas doesn't handle arguments with spaces correctly so create
# a script that will do so properly.
tmpfile=$(mktemp /tmp/sudo.XXXXXX)
echo "#! /bin/bash" >>$tmpfile
echo "export PATH=\"$PATH\"" >>$tmpfile
echo "$1 \\" >>$tmpfile
shift
for arg in "$#"; do
qarg=`echo "$arg" | sed -e "s/'/'\\\\\''/g"`
echo " '$qarg' \\" >>$tmpfile
done
echo >>$tmpfile
# cygstart opens a new window which vanishes as soon as the command is complete.
# Give the user a chance to see the output.
echo "echo -ne '\n$0: press <enter> to close window... '" >>$tmpfile
echo "read enter" >>$tmpfile
# Clean up after ourselves.
echo "rm -f $tmpfile" >>$tmpfile
# Do it as Administrator.
cygstart --action=runas /bin/bash $tmpfile
Or install syswin package, which includes a port of su for cygwin: http://sourceforge.net/p/manufacture/wiki/syswin-su/
This answer is based off of another answer. First of all, make sure your account is in the Administrators group.
Next, create a generic "runas-admin.bat" file with the following content:
#if (1==1) #if(1==0) #ELSE
#echo off&SETLOCAL ENABLEEXTENSIONS
>nul 2>&1 "%SYSTEMROOT%\system32\cacls.exe" "%SYSTEMROOT%\system32\config\system"||(
cscript //E:JScript //nologo "%~f0" %*
#goto :EOF
)
FOR %%A IN (%*) DO (
"%%A"
)
#goto :EOF
#end #ELSE
args = WScript.Arguments;
newargs = "";
for (var i = 0; i < args.length; i++) {
newargs += "\"" + args(i) + "\" ";
}
ShA=new ActiveXObject("Shell.Application");
ShA.ShellExecute("cmd.exe","/c \""+WScript.ScriptFullName+" "+newargs+"\"","","runas",5);
#end
Then execute the batch file like this:
./runas-admin.bat "<command1> [parm1, parm2, ...]" "<command2> [parm1, parm2, ...]"
For exaxmple:
./runas-admin.bat "net localgroup newgroup1 /add" "net localgroup newgroup2 /add"
Just make sure to enclose each separate command in double quotes. You will only get the UAC prompt once using this method and this procedure has been generalized so you could use any kind of command.
A new proposal to enhance SUDO for CygWin from GitHub in this thread, named TOUACExt:
Automatically opens sudoserver.py.
Automatically closes sudoserver.py after timeout (15 minutes default).
Request UAC elevation prompt Yes/No style for admin users.
Request Admin user/password for non-admin users.
Works remotely (SSH) with admin accounts.
Creates log.
Still in Pre-Beta, but seems to be working.
I landed here through google, and I actually believe I've found a way to gain a fully functioning root promt in cygwin.
Here are my steps.
First you need to rename the Windows Administrator account to "root"
Do this by opening start manu and typing "gpedit.msc"
Edit the entry under
Local Computer Policy > Computer Configuration > Windows Settings > Security Settings > Local Policies > Security Options > Accounts: Rename administrator account
Then you'll have to enable the account if it isn't yet enabled.
Local Computer Policy > Computer Configuration > Windows Settings > Security Settings > Local Policies > Security Options > Accounts: Administrator account status
Now log out and log into the root account.
Now set an environment variable for cygwin. To do that the easy way:
Right Click My Computer > Properties
Click (on the left sidebar) "Advanced system settings"
Near the bottom click the "Enviroment Variables" button
Under "System Variables" click the "New..." button
For the name put "cygwin" without the quotes.
For the value, enter in your cygwin root directory. ( Mine was C:\cygwin )
Press OK and close all of that to get back to the desktop.
Open a Cygwin terminal (cygwin.bat)
Edit the file /etc/passwd
and change the line
Administrator:unused:500:503:U-MACHINE\Administrator,S-1-5-21-12345678-1234567890-1234567890-500:/home/Administrator:/bin/bash
To this (your numbers, and machine name will be different, just make sure you change the highlighted numbers to 0!)
root:unused:0:0:U-MACHINE\root,S-1-5-21-12345678-1234567890-1234567890-0:/root:/bin/bash
Now that all that is finished, this next bit will make the "su" command work. (Not perfectly, but it will function enough to use. I don't think scripts will function correctly, but hey, you got this far, maybe you can find the way. And please share)
Run this command in cygwin to finalize the deal.
mv /bin/su.exe /bin/_su.exe_backup
cat > /bin/su.bat << "EOF"
#ECHO OFF
RUNAS /savecred /user:root %cygwin%\cygwin.bat
EOF
ln -s /bin/su.bat /bin/su
echo ''
echo 'All finished'
Log out of the root account and back into your normal windows user account.
After all of that, run the new "su.bat" manually by double clicking it in explorer. Enter in your password and go ahead and close the window.
Now try running the su command from cygwin and see if everything worked out alright.
Being unhappy with the available solution, I adopted nu774's script to add security and make it easier to setup and use. The project is available on Github
To use it, just download cygwin-sudo.py and run it via python3 cygwin-sudo.py **yourcommand**.
You can set up an alias for convenience:
alias sudo="python3 /path-to-cygwin-sudo/cygwin-sudo.py"
Use this to get an admin window with either bash or cmd running, from any directories context menue. Just right click on a directory name, and select the entry or hit the highlited button.
This is based on the chere tool and the unfortunately not working answer (for me) from link_boy. It works fine for me using Windows 8,
A side effect is the different color in the admin cmd window. To use this on bash, you can change the .bashrc file of the admin user.
I coudln't get the "background" version (right click into an open directory) to run. Feel free to add it.
Windows Registry Editor Version 5.00
[HKEY_CLASSES_ROOT\Directory\shell\cygwin_bash]
#="&Bash Prompt Here"
"Icon"="C:\\cygwin\\Cygwin.ico"
[HKEY_CLASSES_ROOT\Directory\shell\cygwin_bash\command]
#="C:\\cygwin\\bin\\bash -c \"/bin/xhere /bin/bash.exe '%L'\""
[HKEY_CLASSES_ROOT\Directory\shell\cygwin_bash_root]
#="&Root Bash Prompt Here"
"Icon"="C:\\cygwin\\Cygwin.ico"
"HasLUAShield"=""
[HKEY_CLASSES_ROOT\Directory\shell\cygwin_bash_root\command]
#="runas /savecred /user:administrator \"C:\\cygwin\\bin\\bash -c \\\"/bin/xhere /bin/bash.exe '%L'\\\"\""
[HKEY_CLASSES_ROOT\Directory\shell\cygwin_cmd]
#="&Command Prompt Here"
[HKEY_CLASSES_ROOT\Directory\shell\cygwin_cmd\command]
#="cmd.exe /k cd %L"
"HasLUAShield"=""
[HKEY_CLASSES_ROOT\Directory\shell\cygwin_cmd_root]
#="Roo&t Command Prompt Here"
"HasLUAShield"=""
[HKEY_CLASSES_ROOT\Directory\shell\cygwin_cmd_root\command]
#="runas /savecred /user:administrator \"cmd.exe /t:1E /k cd %L\""
A very simple way to have a cygwin shell and corresponding subshells to operate with administrator privileges is to change the properties of the link which opens the initial shell.
The following is valid for Windows 7+ (perhaps for previous versions too, but I've not checked)
I usually start the cygwin shell from a cygwin-link in the start button (or desktop).
Then, I changed the properties of the cygwin-link in the tabs
/Compatibility/Privilege Level/
and checked the box,
"Run this program as an administrator"
This allows the cygwin shell to open with administrator privileges and the corresponding subshells too.
I met this discussion looking for some details on the sudo implementation in different operating systems. Reading it I found that the solution by #brian-white (https://stackoverflow.com/a/42956057/3627676) is useful but can be improved slightly. I avoided creating the temporary file and implemented to execute everything by the single script.
Also I investigated the next step of the improvement to output within the single window/console. Unfortunately, without any success. I tried to use named pipes to capture STDOUT/STDERR and print in the main window. But child process didn't write to named pipes. However writing to a regular file works well.
I dropped any attempts to find the root cause and left the current solution as is. Hope my post can be useful as well.
Improvements:
no temporary file
no parsing and reconstructing the command line options
wait the elevated command
use mintty or bash, if the first one not found
return the command exit code
#!/bin/bash
# Being Administrators, invoke the command directly
id -G | grep -qw 544 && {
"$#"
exit $?
}
# The CYG_SUDO variable is used to control the command invocation
[ -z "$CYG_SUDO" ] && {
mintty="$( which mintty 2>/dev/null )"
export CYG_SUDO="$$"
cygstart --wait --action=runas $mintty /bin/bash "$0" "$#"
exit $?
}
# Now we are able to:
# -- launch the command
# -- display the message
# -- return the exit code
"$#"
RETVAL=$?
echo "$0: Press to close window..."
read
exit $RETVAL
Based on #mat-khor's answer, I took the syswin su.exe, saved it as manufacture-syswin-su.exe, and wrote this wrapper script. It handles redirection of the command's stdout and stderr, so it can be used in a pipe, etc. Also, the script exits with the status of the given command.
Limitations:
The syswin-su options are currently hardcoded to use the current user. Prepending env USERNAME=... to the script invocation overrides it. If other options were needed, the script would have to distinguish between syswin-su and command arguments, e.g. splitting at the first --.
If the UAC prompt is cancelled or declined, the script hangs.
.
#!/bin/bash
set -e
# join command $# into a single string with quoting (required for syswin-su)
cmd=$( ( set -x; set -- "$#"; ) 2>&1 | perl -nle 'print $1 if /\bset -- (.*)/' )
tmpDir=$(mktemp -t -d -- "$(basename "$0")_$(date '+%Y%m%dT%H%M%S')_XXX")
mkfifo -- "$tmpDir/out"
mkfifo -- "$tmpDir/err"
cat >> "$tmpDir/script" <<-SCRIPT
#!/bin/env bash
$cmd > '$tmpDir/out' 2> '$tmpDir/err'
echo \$? > '$tmpDir/status'
SCRIPT
chmod 700 -- "$tmpDir/script"
manufacture-syswin-su -s bash -u "$USERNAME" -m -c "cygstart --showminimized bash -c '$tmpDir/script'" > /dev/null &
cat -- "$tmpDir/err" >&2 &
cat -- "$tmpDir/out"
wait $!
exit $(<"$tmpDir/status")
Can't fully test this myself, I don't have a suitable script to try it out on, and I'm no Linux expert, but you might be able to hack something close enough.
I've tried these steps out, and they 'seem' to work, but don't know if it will suffice for your needs.
To get round the lack of a 'root' user:
Create a user on the LOCAL windows machine called 'root', make it a member of the 'Administrators' group
Mark the bin/bash.exe as 'Run as administrator' for all users (obviously you will have to turn this on/off as and when you need it)
Hold down the left shift button in windows explorer while right clicking on the Cygwin.bat file
Select 'Run as a different user'
Enter .\root as the username and then your password.
This then runs you as a user called 'root' in cygwin, which coupled with the 'Run as administrator' on the bash.exe file might be enough.
However you still need a sudo.
I faked this (and someone else with more linux knowledge can probably fake it better) by creating a file called 'sudo' in /bin and using this command line to send the command to su instead:
su -c "$*"
The command line 'sudo vim' and others seem to work ok for me, so you might want to try it out.
Be interested to know if this works for your needs or not.
What I usually do is have a registry "Open Here" helper in order to open a cygwin shell with administrative privileges quite easy from anywhere in my computer.
Be aware you have to have the cygwin "chere" package installed, use "chere -i -m" from an elevated cygwin shell first.
Assuming your cygwin installation is in C:\cygwin...
Here's the registry code:
Windows Registry Editor Version 5.00
[-HKEY_CLASSES_ROOT\Directory\shell\cygwin_bash]
[HKEY_CLASSES_ROOT\Directory\shell\cygwin_bash]
#="Open Cygwin Here as Root"
"HasLUAShield"=""
[HKEY_CLASSES_ROOT\Directory\shell\cygwin_bash\command]
#="c:\\cygwin\\bin\\mintty.exe -i /Cygwin-Terminal.ico -e /bin/xhere /bin/bash.exe"
[-HKEY_CLASSES_ROOT\Directory\Background\shell\cygwin_bash]
[HKEY_CLASSES_ROOT\Directory\Background\shell\cygwin_bash]
#="Open Cygwin Here as Root"
"HasLUAShield"=""
[HKEY_CLASSES_ROOT\Directory\Background\shell\cygwin_bash\command]
#="c:\\cygwin\\bin\\mintty.exe -i /Cygwin-Terminal.ico -e /bin/xhere /bin/bash.exe"
[-HKEY_CLASSES_ROOT\Drive\shell\cygwin_bash]
[HKEY_CLASSES_ROOT\Drive\shell\cygwin_bash]
#="Open Cygwin Here as Root"
"HasLUAShield"=""
[HKEY_CLASSES_ROOT\Drive\shell\cygwin_bash\command]
#="c:\\cygwin\\bin\\mintty.exe -i /Cygwin-Terminal.ico -e /bin/xhere /bin/bash.exe"
Hope this helps. Let me know if it works for you. Thanks.
PS: You can grab this code, copy and paste it and save it in a name.reg file to run it... or you can manually add the values.
Just simplifying the accepted answer, copy past the below in a Cygwin terminal and you are done:
cat <<EOF >> /bin/sudo
#!/usr/bin/bash
cygstart --action=runas "\$#"
EOF
chmod +x /bin/sudo
Try:
chmod -R ug+rwx <dir>
where <dir> is the directory on which you
want to change permissions.

Resources