I now yearn for the old plain HTML pages with "up / next / previous" links [1]. I used to hate how basic they were, but with "modern" web sites loading so much JavaScript to display static content, I'm starting to see the appeal of basic websites with no styling whatsoever.<p>[1] <a href="https://www.freesoft.org/CIE/Topics/index.htm" rel="nofollow">https://www.freesoft.org/CIE/Topics/index.htm</a> - I learned from this twenty years ago. Still relevant.
I've been thinking of migrating my blog back to text only. When I was a student I kept a "blog" as a list of HTML files that I manually updated. I've been using Wordpress for the past ~15 years and am growing weary of having to keep up with plugin, theme or CMS updates. My blog is not worth this much effort to maintain, I just want a place to share essays and occasional photos. I need time to make the switch, but the simplicity of text only format is attractive.
This is great. My only feedback is that, yes you're using text only but please don't shy away from minor CSS changes that can make the site look more pleasant. I have been working my homepage (<a href="https://usmanity.com" rel="nofollow">https://usmanity.com</a>) for years to make it minimal, organized, and still pleasant to look at.
Nice! This gives me more confidence on my decision to keep my websites text only.<p>Please excuse the plug(feedback is welcomed)<p>A problem validation platform - <a href="https://needgap.com" rel="nofollow">https://needgap.com</a><p>Curated list of startup tools - <a href="https://startuptoolchain.com" rel="nofollow">https://startuptoolchain.com</a><p>Startup/Entrepreneurial blog(Has occasional comics, but with transcription) - <a href="https://hitstartup.com" rel="nofollow">https://hitstartup.com</a>
The EU version of USA Today is not text-only, but it is damn fast, with no trackers, no js, and a really fast CDN: <a href="https://eu.usatoday.com/" rel="nofollow">https://eu.usatoday.com/</a>
Chiming in with Piero Scaruffi's website[1]. I used to read it a decade back and it really hasn't changed that much; the information is plentiful and interesting, the site is fast, and it seems like he nailed down the "static content is all that matters in the end" bit not only way before it became a common reaction to the overencumbered web of today, but also way before the web got overencumbered at all.<p>[1] <a href="https://www.scaruffi.com/service/aboutw.html" rel="nofollow">https://www.scaruffi.com/service/aboutw.html</a>
I'm a big fan of simplicity. HTML/CSS index generated by tree v1.8.0. All content US-ASCII text files:<p><a href="https://every.sdf.org/" rel="nofollow">https://every.sdf.org/</a>
Being pedantic, the list strikes me as poorly named. Websites listed do have images and other media, the unique trait they have in common is very minimal CSS (for better or for worse).<p>They tend to use little to no decorative images outside of main content though, which is nice. Not to imply a list like this doesn't have the right to exist.
Not quite text only but I just launched a website with a collection of easily copy-and-paste-able lists in plain text.<p><a href="https://copypastelist.com/" rel="nofollow">https://copypastelist.com/</a>
The Wiby search website is so cool! There is something really comforting and nostalgic about text-only websites. In one of the searches I came across Project Gutenberg Australia. <a href="http://gutenberg.net.au/" rel="nofollow">http://gutenberg.net.au/</a>
Thank you for sharing this.
I should probably email Sijmen but I've been using <a href="https://engineeringblogs.xyz/" rel="nofollow">https://engineeringblogs.xyz/</a> to help me keep track of hundreds of product/software development blogs and am keeping it as plain as possible.
I like small simple websites like these. I especially appreciate knowing about thin.npr.com<p>I don't think completely text websites have to look bad. E.g., my personal website looks pretty good with relatively little CSS:<p><a href="https://goodlifeodyssey.com" rel="nofollow">https://goodlifeodyssey.com</a><p>Some of the formatting is a bit sophisticated too, e.g. for block quotes of poetry:<p><a href="https://goodlifeodyssey.com/hesiod" rel="nofollow">https://goodlifeodyssey.com/hesiod</a>
I'm not usually a pedant, but something about this drives my brain totally insane. None of these websites are text only. Alternatively, all websites are text only. Several of them even feature that unholy technology which must be struck from the Earth according to HN: javascript. What is the line being drawn here? You might as well call it "Websites that load very fast and have a very plain aesthetic that I enjoy".<p>If I wanted a clear definition of a text only website, I'd say a text-only website is a website which is served with a Content-type: text/plain header. This is far from that.
I had a dream of a text only "underground" web the other day. Then executed it in a very poor way. If you want to check it <a href="https://tilde.pt/~fimdomeio/index2.html" rel="nofollow">https://tilde.pt/~fimdomeio/index2.html</a><p>Actually A text only web as a few interesting advantages. Almost instantaneous load even in a poor connection. Very fast development time since you throw away so many things away.
I am thinking about a little side project to launch this summer, just for fun, and I'm very tempted to build a simple card game, like kards.com, but text-only. I think it would be super fun, and I think that if well designed, you don't need fancy graphics to appreciate it.<p>Probably going to use Javascript and angular.js, but open to suggestions. (I used to be good at JS, never used angular so far).
I'm working on static site generator for using plain html...<p><a href="https://findingmyhtmlgoddess.com/" rel="nofollow">https://findingmyhtmlgoddess.com/</a><p><a href="https://github.com/jonascript/htmlgoddess" rel="nofollow">https://github.com/jonascript/htmlgoddess</a>
I really like Gemini but insisting on TLS and closing connections totally defeats the purpose for me.<p>I also came up with this concept <a href="https://github.com/runvnc/noscriptweb" rel="nofollow">https://github.com/runvnc/noscriptweb</a> which I may never actually code because of time constraints.
How about <a href="https://wordsandbuttons.online" rel="nofollow">https://wordsandbuttons.online</a> ? Sure, it uses JS for interactive visuals but all the pages are still under 64K by design. There are very few pictures, too. Even the logo at the bottom of pages is the reused icon file.
I try to do text only myself. Unfortunately Fossil requires CSS (although it does still work reasonably well with Lynx), but other stuff I write is usually just plain text, or if HTML is needed, it rarely has CSS. I also have a gopher server too.
My personal site is text only [1]. There is a tiny amount of JavaScript from Plausible, but there are no web fonts and no images.<p>[1] <a href="https://benovermyer.com" rel="nofollow">https://benovermyer.com</a>
The rate.sx site is pretty nifty. It provides a simple REST endpoint that server cryptocurrency exchange rate info. The nifty part is that it's built to be used on the command line!<p><pre><code> $ curl rate.sx # outputs a nifty table of coins and their market rates
$ curl rate.sx/:help # to see how to use the api
</code></pre>
I whipped up a shell function wrapper that makes this feel more like a standard cli command. In the off chance anyone is interested, I'm sharing it below (unfortunately, HN turns my tabs into spaces, so indentation is now a single space.)<p>Note that I was careful to not let it clutter up your environment, so feel free to include it in your .profile with impugnity. Also, this sort of exhibits my current experiments with terse shell style, so don't hate on me for that, but otherwise the function should provide all the typical ergonomics you expect of a standard cli utility.<p><pre><code> ratesx() (
set -o errexit -o nounset -o noclobber
RATESX_DOMAIN=${RATESX_DOMAIN:-rate.sx}
RATESX_CURRENCY=${RATESX_CURRENCY:-USD}
RATESX_NROWS=${RATESX_NROWS:-10}
RATESX_FETCH_CMD=${RATESX_FETCH_CMD:-curl --silent}
usage=$(cat <<-USAGE
usage: ratesx [<options>] [<command> <arguments>]
Show information about cryptocurrencies. Command defaults to 'table'.
Commands
graph <coin> [<interval>]
Display exchange rate graph for <coin> over <interval> (defaults to
\$RATESX_CURRENCY hours)
value <amount> <coin> [<amount> <coin>] ..]
Convert sum of <amount>*<coin> to <currency> (defaults to
\$RATESX_NROWS).
table Display table of top coins by market capitalization
coins Show list of supported cryptocurrency coins
currencies Show list of supported currencies
info Show upstream API info
Options
-q Quiet mode. Hide header and footer
-F Hide "Follow" line
-T Disable ANSI color
-n <nrows> Number of currencies (rows) to display in output table
-c <currency> Base currency for exchange rate calculations
-d <domain> Use API endpoint located at <domain>
Variables
RATESX_DOMAIN=$RATESX_DOMAIN
Default domain name of API endpoint
RATESX_NROWS=$RATESX_NROWS
Default count of rows to display in output table
RATESX_CURRENCY=$RATESX_CURRENCY
Default base currency for unit conversions
RATESX_FETCH_CMD=$RATESX_FETCH_CMD
Default command with which to submit API requests
USAGE
)
fetch() { $RATESX_FETCH_CMD "$@"; }
table() { fetch "$1/${2:+?$2}"; }
graph() { fetch "$1/$4${5:+@$5}${2:+?$2}"; }
value() { u=$1; o=$2; c=$3; shift 3; p=0
for v in "$@"; do
p=$((p^1)); [ $p = 1 ] && r="${r:+$r+}"; r="${r:-}$v"
done
printf '%s %s\n' "$(fetch "$u/$r${o:+?$o}")" "$c"
}
info() { fetch "$1/:help"; }
currencies() { fetch "$1/:currencies"; }
coins() { fetch "$1/:coins"; }
while getopts ':qFTn:c:d:' opt "$@"; do
case "$opt" in
q) quiet=q;; F) follow=F;; T) ansi=T;;
n) nrows=$OPTARG;; c) currency=$OPTARG;; d) url=$OPTARG;;
:) error "Need argument: -$OPTARG" "$usage";;
*) error "Unknown option: -$OPTARG" "$usage";;
esac
done; shift $((OPTIND - 1))
nrows=${nrows:-$RATESX_NROWS}; currency=${currency:-$RATESX_CURRENCY};
url=${url:-$RATESX_DOMAIN}; cmd=${1:-table}; shift || :;
url="https://$currency${currency:+.}$url"
opts="${quiet:-}${follow:-}${ansi:-}"
opts="${opts:+$opts&}${nrows:+n=$nrows}"
"$cmd" "$url" "$opts" "$currency" "$@"
)</code></pre>
I wanted to try out making gemini pages. I got some of the example servers, but no where in the official gemini spec, both via http and via gemini, could I find examples of the actual freakin gmi/gml formats!<p>I started up Wireshark and then remembered that the traffic is encrypted by default ... so yea, that's not going to help.<p>Are there any gemini tutorial on how to make a gemini site?