Hacker News new | past | comments | ask | show | jobs | submit login
A collection of Unix terminal/console/curses tools (kkovacs.eu)
149 points by tsudot on Feb 28, 2014 | hide | past | favorite | 49 comments



I know lists like this will almost always contain at least one command that I, or any other sys admin on here, would use fairly regularly. But it would be nice if just once an article about "obscure" CLI commands didn't include stuff that's the most basic of essential every day staples for any *nix administrator.

Including stuff like vi and emacs is just ridiculous, quite frankly. If you haven't heard of them already then you really have no place in the command line, let alone reading a document like this.

Most of these types of articles should be rebanded as "beginners cheat sheet for handy command line tools" or just not written to begin with (because, frankly, they're just repeating all the same recommendations as every other article like this that pops up each week and few of them have any descriptions that are accurate either).

Maybe my job this weekend will be to write my own version of these documents but with a decent selection of programs, descriptions and example usages?


> Maybe my job this weekend will be to write my own version of these documents but with a decent selection of programs, descriptions and example usages?

Maybe your job this weekend should be to get a little exercise before writing on the internet?

It's a nice little summary of cool tools. "Some are little-known, some are just too useful to miss, some are pure obscure".


> Maybe your job this weekend should be to get a little exercise before writing on the internet?

Maybe you should learn some manners and stop posting wild assumptions about physique of your peers.

> It's a nice little summary of cool tools. "Some are little-known, some are just too useful to miss, some are pure obscure".

It's barely a summary. A summary requires some details about the points being raised; albeit a briefly phrased. That site was more akin to a bullet-pointed list.

The descriptions are lacklustre, the screenshots are useless for all bar the ncurses programs, and including "just too useful to miss" programs like vi and curl but with zero detail nor examples about how to use them is absolutely worthless. Anyone who's every used a Linux / UNIX command line -even briefly- would already know that vi and curl exists - what they might not know is how to use it. And that's where articles like this should focus if they're wanting to incorporate commonly known commands along with obscure ones. Otherwise you're trying to appeal to everyone yet failing to please anyone.


Hey man, the guy put together a nice little list of tools, nothing more nothing less.


> little list of tools, nothing more nothing less.

Which was my point. :P


Yes


I'm a big fan of Joey Hess' moreutils - https://joeyh.name/code/moreutils/

Examples:

- sponge: read in all of stdin, then write to the given file. Great for pipelines: sed "s/root/toor/" /etc/passwd | grep -v joey | sponge /etc/passwd

- vipe: easily drop a $EDITOR instance in the middle of a pipe chain

- ts: timestamp all standard input; great for long-running output


I hope my own collection of related tools is also useful?

https://github.com/skx/sysadmin-util

More focused on those a sysadmin might prefer than general-use, but contributions or updates are welcome.


Although there's some real gems in there (I can't believe I never ran across vipe before!), its Debian package conflicts with the GNU parallel package: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=718816 Quite annoying.


How is sponge different from > filename?


For one thing, it allows your pipeline to both read from and write to the same file, as it defers the output until there is no more data to read from "upstream". For instance, this won't work (it truncates the file for writing before it is ever read):

    grep foo <some_file >some_file
But this will have the intended effect:

    grep foo <some_file | sponge some_file


It can be used where ">" does not work, e.g. when using sudo. So you can do

  echo data | sudo sponge /etc/a.file
`tee` can be abused in a similar way, but it also prints the data to stdout, which is ugly.


sed can edit in place though.


indeed!

sed -i -e 's/root/toor/' -e '/joey/d' /etc/passwd


The best not-so-well-known tool he mentiions is socat. If you use netcat, socat is like netcat on steroids. It's just a bit hard to remember all the options, but next time there's something crazy you want to do with netcat that it doesn't support, like tunnelling raw packets from a wireless interface on a machine in firewalled corporate network through an SSL tunnel to home machine and dumping those packets locally to a serial port, socat can pretty much do it all.


I hear nmap's ncat is great as well but I haven't tried it yet. I agree that socat is invaluable, especially when dealing with serial devices.


I don't really think it's fair to call anything that's in a POSIX standard obscure (like xargs).

Most of these are fairly common, and many of the descriptions are terrible, for example: `rsync` doesn't "[Keep] filesystems in sync over SSH"; it's a versatile `cp`-like for dealing with remote systems efficiently that happens to support transfers over SSH.


As I can no long edit my comment: when I made the parent comment, the title referred to them as _Obscure_ tools.


Yes, the title of the post on HN is now changed, but the url of the original post still refers to them as "obscure" tools.


For being "obscure", htop manages to find its way into a very large proportion of *NIX screenshots on the web.

Not to mention vim, rsync, curl, and ack, all of which I (and I had assumed everyone else) use on a regular basis.


I still don't use ack, mostly because it's not installed everywhere I go on a regular basis and so my fingers still go 'grep' when I think 'text search'.

(I also rarely use curl, because my fingers are trained for wget -qO. That one is a personal quirk, though, I didn't know about curl until I'd had that one imprinted on my brain. The cool thing there, though, is that wget -qSO will send headers to stderr and content to stdout, so you can get fancy.)


Installation is simple enough assuming you have perl:

    curl http://beyondgrep.com/ack-2.12-single-file > ~/bin/ack && chmod 0755 !#:3


I'd highly recommend checking out httpie as a replacement for curl. I find the interface much more intuitive.


I'd love to see a site just like this, but cataloging all the best Unix console tools.

It sounds ridiculous, but you can tell a lot about a console program from a screenshot.


I would like to see a graphical terminal emulator. When I mean by this is that the terminal emulator works the same as usual- it should display normal output, support ANSI escape sequences and have scroll-back. But it also should be able to display graphics. If your UNIX command line program happens to output a graph, it shows up on the terminal. The graph is embedded inline with the text and you can scroll back to it.

Interactive programs should also possible- think of a schematic editor or a figure editor running right on your terminal emulator.

The point of all this is to retain the context of whatever you're doing without having to switch windows. That and to advance the state of the art of console tools- certainly they have not changed in decades.


On the topic of ways to improve the terminal, I'd like to see output separated from input.

Put the prompt 1/4 of the way down the screen instead of at the bottom. Above the prompt, we have a backlog of the commands we input. As we press up, it scrolls through the backlog, as normal.

Below the prompt, we see the output of the command.

I'd also like to see tab completion (and maybe some form of hinting) on programs, not just file paths.


>On the topic of ways to improve the terminal, I'd like to see output separated from input.

>Put the prompt 1/4 of the way down the screen instead of at the bottom. Above the prompt, we have a backlog of the commands we input. As we press up, it scrolls through the backlog, as normal.

>Below the prompt, we see the output of the command.

That would be more fitting for a shell than for a terminal.

>I'd also like to see tab completion (and maybe some form of hinting) on programs, not just file paths.

bash already does this. zsh and at least one variant of ksh will do completion if you type the first letter of a command. I was surprised that tcsh doesn't.


Check out terminology [0]. The video is quite long but impressive.

There was a work by someone on github for building a new kind of terminal. It wasn't line-based, rather "context-based": the output of the command was written next to the input, but you could write commands in the same terminal next to each other. It was some amazing stuff, but I can't find it (I think the logo was an octopus or something like that, and the "backend" was emacs)

[0] https://www.gnu.org/philosophy/free-sw.html


I recently discovered netbrake. Super awesome tool to limit the bandwidth used by a process. http://www.hping.org/netbrake/

Also its written by antirez :)


Incorrect title for the post. Many of the tools mentioned are very commonly used. Also, except nice screenshots, this article is just a listing of tools with very little useful info to go with them.


The author mentions App::Ack, but I doubt you'll go back to using `ack` once you've tried The Silver Searcher (`ag`).


I was certain this had already been posted here: http://news.ycombinator.com/item?id=2567186 (from 2 years ago[0]).

I think the difference is in the trailing slash of the submitted link.

[0] Oh god, no.


Curl is obscure?


I used remind and wyrd a fair amount before joining Google + switching to google calendar. Remind is really powerful, and wyrd is nice.

Here [0] is a post detailing how I got calendar notifications working with remind/wyrd + screen + URGENT hint.

[0] http://lists.roaringpenguin.com/pipermail/remind-fans/2006/0...


ITT people who don't think those tools are obscure


Good list thank you. Another good one is "VnStat/VnStati" I use it generate traffic graphs.


> nethack & slash'em

> Still the most complex game on the planet.

Does anybody know if a HTML5 version of Nethack exists?


Very likely, but potentially even better is this WebTiles version of Dungeon Crawl Stone Soup: http://crawl.akrasiac.org:8080/#lobby


I think you might be underestimating the complexity of nethack


Here you go: http://gateone1.rs.liftoffsoftware.com/

That's a direct link to the Gate One demo. You can play Nethack (among other things).

Here's the source to Gate One: https://github.com/liftoff/GateOne


If you like command line Unix tools, the Inconsolation blog is pretty good. He's currently running through the tools alphabetically.

http://inconsolation.wordpress.com/


They missed out ls and cd.


My latest discovery: jq - command line json processing, formatting and colouring.


I was going to mention jq. It's great, and replaces the previous json tool I was using (can't remember the name now, but it wasn't as useful.) I hope it becomes more popular.


I'm curious what tools the author doesn't consider obscure.


smem is pretty rad and obscure (http://www.selenic.com/smem/).


the full topic is "cool, but obscure unix tools"

most of these aren't really obscure, but i still think it's a cool list. thanks for that.


vim and emacs are obscure?


s/Obscure/Handy/gi




Consider applying for YC's Spring batch! Applications are open till Feb 11.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: