Another Alternative to Conky

[dkeg]

So I've been experimenting with LemonBar boy. Check it out here. Easy enough to get set up. I at first just piped a simple one-liner conky through it, but then thought, I could use like osd_cat. I have the variables set, just needs some fine tuning. So as DJ wrote up, another nice conky alternative.
*Disclaimer, I'm absolutely positive that there is a better way to do this, but it works, and does what I need ... for now.*

This is the script I'm using. Name it whatever.

Code: Select all

#! /bin/bash

temp() {
    tempstat=$(cat /sys/class/thermal/thermal_zone0/temp | awk '{print $1/1000}') 
    echo "$tempstat"
}
cpu() {
    cpustat=$(top -bn 2 -d 0.01 | grep '^%Cpu' | tail -n 1 | awk '{print $2+$4+$6}')
    echo "$cpustat"
}
mem() {
    memstat=$(free -m | grep buffers/cache | sed -e 's/[buffers/cache -+ :]//g' | sed 's/.\{4\}$//')
    echo "$memstat"
}
clock() {
    clockstat=$(date | sed -e 's/EST 2014//g')
    echo "$clockstat"
}
batt() {
    battstat=$(cat /sys/class/power_supply/BAT0/capacity)
    echo "$battstat"
}
mail() {
    mailstat=$($HOME/mail) 
    echo "$mailstat"
}
music() {
    musicstat=$($HOME/music) 
    echo "$musicstat"
}
vol() {
    volstat=$(amixer get Master | tail -1 | sed 's/.*\[\([0-9]*%\)\].*/\1/')
    echo "$volstat"
}

#OUTPUT=

while true; do
    echo  "     \f5m$(mail)  b$(batt)  r$(mem)  c$(cpu)  t$(temp)  v$(vol) \c\f9$(clock) \r\f9$(music)     "
    sleep 10s;
done

Place in .xinitrc

Code: Select all

/path/to/script | bar -p &

P.S. I did borrow your volume string Joe, thanks!

http://postimg.org/image/sixymv9tr/

[wuxmedia]

Nice.
I'm slowly getting fed up with i3bar...
No shame in promoting something as great as your site dkeg!

[DebianJoe]

Very cool. There are 1.5 billion ways to skin an osd_cat, but new ways are always a welcome sight.

[johnraff]

Thanks dkeg - alternatives to conky are always welcome! Is bar lighter/have more features?

Wonder if you could simplify those functions, like perhaps

Code: Select all

temp() {
    tempstat=$(cat /sys/class/thermal/thermal_zone0/temp | awk '{print $1/1000}')
    echo "$tempstat"
}

could be

Code: Select all

temp() {
    cat /sys/class/thermal/thermal_zone0/temp | awk '{print $1/1000}'
}

and still give the same result? I haven't yet installed bar to try this, so just guessing...

[machinebacon]

^ Why using cat anyway? :D

[dkeg]

check it out on github. Actually comes with a paletter script similar to Rho's to pull your xcolors to use within bar. it's very light for sure.

Just one of those things you find and want to get it working. As far as the script 'cleanliness'. I'm sure its bloated up a bit. I don't know what I'm doing. Tbh, once I got it working, I just let it be. The improvements and perfecting is for all of you guys!

and why use cat? I dunno. I literally just took what I was using for osd_cat, hence calling it osdbar, and tuned it to work here. Of course, that doesn't mean I need it in the osd_cat either.

[DebianJoe]

Code: Select all

temp() {
    awk '{print $1/1000}' /sys/class/thermal/thermal_zone0/temp
}

Should do exactly what MB is suggesting.

[machinebacon]

cats are bloat :D thanks joe!

[johnraff]

cat-killing is a time-homoured geek sport of course.

[DebianJoe]

Side note on cat and pipe killing:

My awk if pretty sketchy yet, but I like breaking these guys down. You can also totally avoid grep for searching with regexps using awk. For example:

Code: Select all

awk '/(^Mary)|(^Elizabeth)/{ print $3 }' /home/joe/blackbook

Assuming that I have a file called "blackbook", this one will find lines starting with either "Mary" or "Elizabeth" and give me the 3rd column of info on that line. (Which, for the sake of discussion and staying 'on-topic' we'll assume is going to be sent to by lemonboy bar. For even more fun in theory, let's assume that's where I keep the phone numbers of people in ~/blackbook and that I'm trying to simplify the process of making booty-calls while drunk by not even having to open a file.)

awk is nuts with what all you "can" do with it, such as in-line iterating, but I'll just leave it at that.

[machinebacon]

Side note: http://linuxbbq.org/grillers/PDF/sed+%2 ... Arnold.pdf

:)

[johnraff]

You can also totally avoid grep for searching with regexps using awk. For example:

Code: Select all

awk '/(^Mary)|(^Elizabeth)/{ print $3 }' /home/joe/blackbook

...awk is nuts with what all you "can" do with it, such as in-line iterating, but I'll just leave it at that.

Yes! Awk (and Dillo) fan here. You so often see grep 'something' | cut something to pull a bit of data out of a file when a single call to awk could have done it.

A couple of days ago, however, I ran into how you can use sed to search for a certain line then pull out a bit of it that fits a certain regex, all in one command. Use -n option to suppress normal output, /p switch at the end of a s/a/b/ substitution to print only that line, and so s/regex\(bit\)moreregex/\1/p to print only the bit inside the ( ).
The specific example was to extract as a 3-character string the audio format inside a container file (eg mp4) from the output of ffprobe:

Code: Select all

ffprobe "$file" 2>&1 | sed -n 's/.*Audio: \(...\), .*/\1/p'