But you know, if I’m being honest, the experience was not entirely unpleasant. Sure, I missed certain niceties from the graphical side of things, but there were some distinct benefits to living in a shell. My computers, even the low-powered ones, felt faster (command-line software tends to be a whole lot lighter and leaner than those with a graphical user interface). Plus, I was able to focus and get more work done without all the distractions of a graphical desktop, which wasn’t bad.
What follows are the applications I found myself relying upon the most during those fateful ten days, separated into categories. In some cases, these are applications I currently use over (or in addition to) their graphical equivalents.
Obviously, among OSAlert readers, the terminal is a prized tool many rely on – but I wonder how many of you truly live entirely within the terminal, never touching the comforts of a graphical user interface.
Thom Holwerda,
I use the terminal so often that I’ve bound it a a keyboard shortcut.
For many of us the terminal is the primary method of developing software and interacting with servers, which is quite a bit of my work. My linux distro doesn’t even have a GUI. Obviously all terminal work can be done within a GUI shell, so if you live within xterm, does that count?
Personally I cannot do my work without a web browser, so it wouldn’t be practical to never touch a GUI, but beyond that the terminal covers most of my needs. One feature that I miss sometimes when not inside of a GUI is copy & paste using the mouse.
If that’s all you miss, you can achieve it with GNU screen + gpm.
What I don’t like about a full terminal set up is the lack of graphical support. The article mentions browsing with lynx or links, but the web really isn’t all that browsable that way. Something lightweight like netsurf is light-years better.
Personally, if I’m in a situation where I’m doing a lot of CLI work, I use dwm. Once you learn some of the shortcuts (and even added a few of your own) it is incredibly useful.
teco.sb,
I do put ‘screen’ to good use. One thing I absolutely despise now days are terminal applications that intercept the mouse with an inferior handler that prevents copy/paste from working. In the past year or two debian/ubuntu released new versions of VIM which do just that.. VIM is my go-to editor for things like config files, but it’s absolutely infuriating not being able to copy/paste things like ip addresses, this anti-feature is such a huge productivity regression. Same deal with “aptitiude”, I go to copy/paste only to discover the mouse has been hijacked. When I have root access I’m able to modify VIM’s configuration to disable it, but overseeing so many different systems that come and go it’s been quite a nuisance!
I may end up building a tool that removes these kinds of bad features automatically.
Agreed.
Yea, there are so many different options. I don’t need much at all so I gravitate to lightweight environments, but sometimes it’s disappointing that they don’t have the usual keyboard bindings mapped out of the box. Like XFCE leaving the ‘windows’ button unbound rather than mapping it to the applications menu as one would expect. Even though I don’t need a comprehensive desktop like Gnome or KDE, I will say I think the defaults are better out of the box.
I used to, about 30 years ago.
I still do, but I used to, too. (Sorry couldn’t resist)
I’ve tried it off and on in the past; I’ve had the best experience in distros that are considered “minimalist” by today’s standards like Slackware, Alpine, and Void. The latter two are especially easy and quick to get up and running with modern tools; they both have dependency-resolving package managers that are ready to go out the door. Slackware is great too though, if you don’t mind fiddling with package management manually.
I’m tempted to do this again on my 2006 Mac mini. It runs amazingly well with a lightweight distro like BunsenLabs or Void, it can only get better with CLI exclusively.
…so did I, and then I discovered Norton Commander and became just so much faster at everything file- and discovery related.
Nowadays I still love using the commandline, scripting and batching but there are also hundreds of things that I prefer to do with a GUI (and Total Commander + Everything is still a great toolkit)
Quickly see if everything is working fine on a machine: Just have a glance at device manager. Have an exclamation mark? Right-click and fix the driver through the GUI. But I also love bulk-saving/restoring drivers through dism.exe on the commandline when running deployment-scripts. Have a look at the (GUI) reliability-monitor first before using the commandline to churne through logfiles and event viewer events
Browsers, games, communication tools and even IDE’s are all better with a GUI unless you are living in that one app. GUI is an addition, not a replacement. The same goes for mouse/touch by the way
I grew up on terminals, so I think I can safely say that living in a terminal is crazy. Using one frequently? I still do. Using it almost exclusively? Cray-cray. Lock that dood up!
I would LOVE to be able to operate my phone entirely within a terminal …. a voice interface terminal. I think it’s an advertising motivated conspiracy scenario that phones don’t have a useful built in voice mode for driving. Siri literally tells you she can’t do certain things by voice when driving so you have to either stop or work the screen. In dash screens are only marginally safer. CarPlay is not a full solution, Voice is the new terminal and it needs work! Instead of trying and failing to go straight to natural language, I think it would be OK to have a specified and user configurable voice command set, just like a command line.
Classic Sci-Fi had really useful voice AIs that are nearly withing reach via today’s tech for a narrow but highly capable domain. I bet if someone analysed Star Trek’s OS (Original Series) computer it was only asked to do about 20 things.
Alexa is the most useful of the voice bots at present, by small margin. Amazon should build an Alexa voice skin for smart phones with the ability to learn user specified commands.
Post Jony Ive, Apple should go back to true Industrial Design and look at what people need their phones to do, beyond the pure physical form factor. Ops (the new ID host at Apple) might even do that because maybe they don’t know better and have to be fed up of building fragile over sized sheets of glass for under sized pockets.
Torvalds should leapfrog the antiquated screen and keyboard game and build a voice first OS for the future!
…. or Stallman, He already has carpal tunnel and would have far fewer closed source drivers to stress over in a voice world. Imagine a voice emacs :-). No quad key presses, just quad compound words, like German.
Or Microsoft. They could do it in their sleep and finally be first at something.
One of the things I miss the most about Windows Phone was the unmatched voice control capability. You could control just about everything on the device via Cortana. It would automatically ask if you wanted it to read an incoming message to you, and would give you several options to handle the message after. Android still doesn’t come close, and iOS is laughable in this arena. It’s a shame Microsoft wasn’t willing to sell their mobile division to a company that cared enough to run with it, though I guess since it’s so woven around desktop Windows 10 they couldn’t really break it off.
Bixby (when I first played around with it at least, I hate talking to my phone/devices, even though I occasionally talk to myself) had pretty decent control of various functions of the phone. I haven’t tried out the latest versions to see if it’s improved at all.
You mean you haven’t heard of MS-DOS Mobile?
https://time.com/3766895/microsoft-ms-dos-smartphone/
I love terminal but i have to be honest the internet will not work on a text only browser so i use a tiling window manager and use the “suckless” programs https://suckless.org/