I’ve got to confess, I have for years been guilty if not reading the documentation. I simply go with the flow and hope it works…
But not anymore! And why the change you may ask? We’ll, I’m reading the f…ing documentation on Rocky linux and I’m just blown away from the amount of great information!
If you’ve been guilty of not reading the documentation, let me me know what changed it for you
If you’re not reading the documentation, this is your time to confess!
I have found the docs the best place to start with anything, but have found that some don’t know how to write good documentation.
Also man pages and the tools own help -? Or -h
If you run something that has pants docs, you could always see if there is a way to help update it
I had a boring manufacturing job with long gaps between batches of work, so I read every help file in Windows NT4.1. While reading them, I found a way around our IT limitations on which apps we could run, and learned how to write scripts. So I wrote a password protected launcher tool using a macro feature in a terminal emulator I had access to on my workstation, and then started reading the man pages in Unix sys-V.
i stopped reading most docs after like 95 unless they are rfc or reference and i had a memory that was stellar
now, i read all of them over and over and over because i got a tbi from electroshock “therapy” and i am working with shitty autobiographical memories and cant get to the details. so i read, keep reading, and make sure all the mans are at hand along with my references. now i get frustrated and wanna die but i still get it done but im always like yeah uh no
What the … LOL
Gotta start somewhere 😅
I find that the docs usually consist of a quick start guide covering some ultra tight scenario that doesn’t apply to most people, and reference material that’s just some total brain dump of every possible command without any kind of context.
If documentation is written in a readable and confluent way, RTFM isn’t such a big deal. The issue comes with overly draconian and non-confluent documentation.
Confluent?
Flowing/coming together.
I think what they are referring to are docs where pieces are explained individually, but not in a consistent or cohesive way, obfuscating use.
In my experience, all the Linux documentation I have read has been written for peers of Linux developers, who are familiar with technical terminology and several concepts and steps are left out and implied rather than explained.
It’s a way for developers to ensure that Linux never receives adoption past other developers. Literary equivalent of pulling the ladder up.
who are familiar with technical terminology and several concepts and steps are left out and implied rather than explained.
Said it before and I’ll say it again: had to manually install some software to make Steam tinker launch work, and the instructions for installing it were to download and prepare the GitHub folder, then “do the usual and move the completed file to …”
Ive used git in the past and it still took me multiple minutes to figure out they meant the “make && build” command. Why was that so hard to fucking write??
Highly specialized people live in bubbles and assume that everyone else lives in their same bubble and so if someone else doesn’t understand, they aren’t worth communicating with.
XKCD 2501, basically.
Thank you for this. It’s beautiful.
I mean, it’s technical documentation. There’s a limit to how exciting it can be. lol.
You’re absolutely right
I thought you wrote confluence and wanted to grab my pitchfork.
rivers
Looking at you, Nix documentation
Day 564: I have become lost in the forums amidst flake debate threads. Do not search for me, it is already too late.
There is a way with chmod in bash to change files and folders with files getting no execute bit and folder do (rwX instead of rwx). It’s in the man pages but good luck finding it via Google. Stackoverflow just suggests using find over and over again.
That did it for me.
Nothing teaches you what the documentation says like plowing ahead without reading it, fucking something up badly, having to crawl back to the documentation hat in hand and actually read it.
Oh it happens to the best of us. I was working on a simple cron the other day with the cron string that would insert the cron into my cron config something like ‘echo’ and the normal string you’d recognize, and ended with a ‘-’. I wasn’t paying attention and issued the command which did insert itself into the cron config, but in a manner in which I didn’t want. It replaced the whole cron file with that one string. #$@^$$ Luckily I have a cron to back up the crontabs.
I worked next to a technical writer for Unix; the Unix. One of the things we were known for, actually, was the amazing documentation. This guy and both teams of writers (that many) maintained the doc as their entire job. It was written well, it was spell-checked, it was accurate, it was accessible. If you installed the machine, it was on http://localhost/doc or so.
Almost all tech writers were turfed after Y2K. They cost money and didn’t earn directly.
If you notice a lack of good docco like you notice a lack of mentoring in code dev (I see you, Systemd), then we know how we got to this stage.
If you become CEO, just keep that in mind.
As a technical writer, I always get a bit giddy when someone shows appreciation for good docs haha Thanks for sharing!
I also don’t RTFM
I would say that I RTFM about 75% of the times (give or take). Though I only do it to see if I can find something other than what I intended to use the software or hardware for.
I mostly try to read the docs, but sadly good documentation is pretty rare.
While investing money to create good documentation is the preferred way, I cannot trust it to be accurate in this day and age of cutting corners. It takes a bit longer but I’ll always look at the code itself to get me closest to the truth of what is going on under the hood.
Lol reading the source has trained me to try reading the documentation.
If it’s good, it’ll save hours or crawling through code.
RTFM. …
… The last thing you try.It’s weird that Linux certification requires rote-memorization of commands. The only people who make any effort to memorize commands are newbies and people studding for exams. You will always have access to bash history, man, and --help, even from an offline machine.
Every command I’ve memorized is simply the natural process of repetition. Is that your experience?
Yes. But also, despite having done it literally thousands of times, I still can’t tell you which way round to put the target and the link name for a softlink on the first go.
My first guess is always
ln -s $NAME $TARGETNo amount of repetition will fix this.
I used to have that problem with ln until I realized it’s essentially the same ordering as cp: source, then destination. The source being the existing file that you’re linking to, and the destination being the link that you’re creating
My trick to remember:
You can link to a target without giving a name to the link. ln will use the basename of the target file then. You can’t create a link without a target, so target has to go first since it’s not optional. Did it for me
People are worried about losing skills to AI while all the skills have already been lost to Google and stack exchange 😅
Are you trying to say I’m not a newbie with over 20 years of experience?











