Learn Enough Command Line to Be Dangerous (Draft)

kill -15 $pid || kill -2 $pid || kill -1 $pid || kill -9 $pid

I could be wrong but I think beginners learn about -9 because beginners are more likely to start processes that they change their mind about and then cannot stop. Nothing wrong with this in the spirit of experimentation.

For more advanced users more familiar with the processes they start, I would guess -15 should be sufficient most of the time.

Could you expand on this? In my experience, kill frequently fails without -9, so I basically always include it (and it's never bitten me). I'm open to suggestions, though, so if you have a recommendation for a better way I'd appreciate the feedback. (Feel free to shoot me an email at the address in my profile.)

Thanks for the feedback. The reason for the verbiage is that this is the first of eight planned prerequisite tutorials [1], so Learn Enough™ Command Line to Be Dangerous is setting the stage for all the others as well. The introduction to each future tutorial will be focused exclusively on the content of that one document. I'll definitely look at putting some of the intro into a sidebar, though; that might work well.

Regarding Command-space, I've added a clarification regarding ways to start Spotlight. Thanks again for the helpful suggestions.

[1]: http://learnenough.com/

My cmd + space is actually somehow bound to both, so to search I do:

cmd + space ; esc ; cmd + space

Every once in a while I debate uncolliding them, but for some reason I get a mild kick out of occasionally unexpectedly seeing cyrillic dump out on the command line.

There are times i wish rm had a dry run kind of switch, where it would list out all the files that would be affected before asking for approval.

Indeed, I hope that the three-tutorial sequence covering the command line, text editors, and Git will be a useful package for the works-with-developers crowd.

Maybe you could create a cheat sheet of the combination of summaries?

Cool, that's a great idea. It should be easy to extract all the relevant tables and put them into a single document. Thanks!

Heh, I've been using 1G for so long I didn't realize g works, too. I'll think about mentioning it. Of course, one advantage of 1G is that it generalizes to values of n other than 1.

Feel free to abstract away the entire shell. The shell is merely a command-driven REPL that interacts with the kernel API, rewrite the shell however you want with an embedded interpreter like how Emacs eshell works (learn one shell, works across every OS Emacs is ported to).

Can also just script shell commands and create levels of abstraction to suit whatever simple layer of abstraction suits you. This is what things like Nautilus File Manager do to manipulate files with a mouse.

I enjoy Unix and the Unix philosophy, but I'm legitimately interested: what would a better system look like to you?

I'm just riffing here but maybe the reason we're stuck on *nix is because it came from a time before The Cloud, subscription software, sharecropper software, etc. It was made to make you more productive; Not to make you #<whatever> of #<whatever> user of Super.os.

OSes weren't worth a whole lot; Money was made on the hardware and the applications. How do you make people productive? Give them an OS that enables productivity.

Unix succeeded because AT&T didn't recognize the value of what they had.

demonstrate your improvement

The Learn Enough tutorials use the Softcover system (http://softcover.io/), which stores the HTML on S3 and uses JavaScript to pull it in dynamically. The flexibility and convenience of Softcover come at the price of a little extra JavaScript—a price that, from my perspective, is well worth paying.

What has the world come to where the command line is something foreign to a software developer?

What in the world are you talking about? The whole point of this tutorial is to teach command-line basics as a first step toward becoming a software developer. From the tutorial:

Many programming tutorials either gloss over the command line or assume you already know it. But understanding the basics of the command line is absolutely essential to becoming a skilled developer. Indeed, if you look at the desktop of an experienced computer programmer, even on a system with a polished graphical user interface like Macintosh OS X, you are likely to find a large number of “terminal windows”, each containing a series of commands at a command line (Figure 1). Proficiency at the command line is also useful for anyone who needs to work with developers, such as product managers, project managers, and designers. Making this valuable skill accessible to as broad an audience as possible is the goal of Learn Enough™ Command Line to Be Dangerous.

A different type of world?

Who claimed this? You totally missed the point of these tutorials. It's "Learn Enough Command Line to Be Dangerous" NOT "Learn Everything About the Command Line to Be a Guru".

You should just be able to scroll down. If that doesn't work, send an email to the address in my profile with details on exactly what's confusing you. For future iterations of the UI, I'm looking at adding access to the table of contents as well.

Update: You probably just need to enable JavaScript, as the Learn Enough tutorials use the Softcover system (http://softcover.io/), which stores the HTML on S3 and uses JavaScript to pull it in dynamically.

By the way, the other replies to this comment refer to Mark Bates's Conquering the Command Line [1], which is a completely different product.

[1]: http://conqueringthecommandline.com/book

Maybe whip up some curl commands in a command line, post-processed with sed and awk.

I enabled JS from a couple of domains, then gave up and backbuttoned out.

Here is the table of contents to make things a bit easier so you won't have to go through all the URLs: https://softcover.s3-us-west-2.amazonaws.com/8/unix_commands...

Use the command line to get the URLs for each chapter?

  Go to conqueringthecommandline.com/book
  Save the page source
  Insert the filename into the script below
  Alternatively, omit filename and pipe the source of the page to the script 

  Works with BSD sed as well as GNU sed; first lesson: portability

   a=$(printf '\004')
   sed 's/s3_url/'"$a"'&/' \
   |tr '\004' '\012' \
   |exec sed '
   s/\\u0026/\&/g;
   s/.*s3_url.:.//g;
   s/\".*//g;
   /https:.*Amz/!d;
   ' filename