“Usually” eww or w3m.el. (usually meaning once a year when X breaks or when I browse documentation)
Tried to post with eww but despite looking fine it doesn't seem to get through.

I usually don't. The web is a multimedia platform, and a text only browser is fairly antithetical to that.

>>3
Well it is true that the web uses images, video and so forth, often times these are unnecessary in the situations that one would be using a CLI. Take the website called Medium for example, there's a popup overlay, useless pictures of random hipsters on a laptop in a coffee shop and inane video ads (if you don't already block ads that it). If you're just looking for technical assistence or a guide to piece of software, there's no point in wasting your time with those distractions.
Additionally one can view images/videos in a CLI. I'm not completely certain if one can embed them into a terminal or if there's a browser that has an option for that. if one wants to keep it text only you could try using the caca libraray for imagemagick: it converts images and videos into text.

>>4
I know links2 has an image mode.

https://web.archive.org/web/20150701000000*/http://annasagrera.com/on-ascii-youtube-and-letting-go/

>>1
I'm using w3m. It's a pretty neat way to beat down distraction, which is a big problem I have with normal internet.

>>6
That's a nice factor too. One can get a similar feeling when using an RSS reader to read articles or blogs on a particular subject. You have to make sure to set images to not be loaded, which most of them seem to have such an option. That way instead of having bloated attention grabbing pictures or ads, if they weren't blocked, you get just the raw content.

>>1
EWW is nice every once in a while, but I don't have much of a use for browsing the web with a CLI. If it's something that requires JS then I'll have to use Firefox, because there I can at least control which scripts do and do not load. If there's no downside to using Tor for it then Tor Browser it is. If it's a simple task that I want to automate then Curl can probably do it.
I've also been experimenting with Next, but it crashes often when used through torsocks, and has poor features to control/block JS. If I did want to script something interactive though then that might be the choice because common lisp is nice.

elinks as well

>>6,8
Semi-off topic, but I used to use w3m.el, but it often took a long time to render the pages. And if you resized the buffer, it would freeze up for a second, so I always bounce back to eww, even though shr can't render as many sites. With w3m you can write a simple function to fetch the url and start a graphical browser

(defun w3m-external-open-graphical-browser ()
(interactive)
(browse-url-graphical-browser w3m-current-url))[/code]
In eww, that can be done with shr-external-browser. I find myself opening sites mostly in eww and then switching to qutebrowser when I need it. Conversely, if you use exwm, you can make a macro to fetch the url of the current site on qutebrowser and send that to browse-url. Took that little trick out of exwm-surf.

You can also direct sites to different browsers according to regexp with browse-url-browser-function--little-known fact.
[code](browse-url-browser-function
'(("\\(.+\.mp[34]$\\|.+\.jpe?g\\)" . browse-url-image-viewer)
(".+&loop=1$" . qrthi/browse-url-image-viewer)("duckduckgo\.com\/\\?&.+=images.+$" . browse-url-graphical-browser)
("." . w3m-browse-url) ; where w3m-browse-url is a catch-all for everything that doesn't match the above regexp's
))

That approach is a good alternative to that part of that enjoys obsessively tweaking browser, in my opinion--which is bad because it detracts from the time spent obsessively tweaking Emacs--by offloading all the configurations and bindings to Emacs. Don't even get me started on webjump.

I use lynx because its the only one that I know which also works with gopher. Having said that it's really clunky and not that great, so I should probably just move to something else.

When I'm in emacs I use eww, and it switches to elpher when I go to a gopher site, but currently I'm in an acme phase so I don't use emacs.

>>12
Just use raw netcat commands, the protocol is REALLY REALLY simple.
As far as I remember, it only has a command set of two:
"
" (enter): get a directory listing
"/path/to/fileorresource": cat this file

>>13 How does one pipe those from netcat? e.g. to a file for later reading, through a typesetting filter to display,,?

>>14 https://stackoverflow.com/a/13591423