Text-only news sites are slowly making a comeback
poynter.org
poynter.org
I see a lot of websites now which, in an attempt to reject all the nonsense, end up rejecting even basic CSS, and you are subjected to browser width text in Times New Roman with ugly blue underlined links (danluu.com is a good example. Such great content that I love to read but, to be honest, ugly). Surely there is a middle ground?
Something as simple as the following can make your website look neat and readable:
body {
font-family: Helvetica, Arial, sans-serif;
max-width: 800px;
margin: 0 auto;
}
a {
text-decoration: none;
}
My mum, my step-dad, my brother and his fiancee, my next door neighbours, my best friend and my worst enemy all have this in common: they go to a website and they want to see the content on the page. They don't care why or how, so they're never going to apply an optional CSS file, even in the unlikely event they understand what it is.
Sorry, that's not meant to come off as a brutal put-down: it's just the way most people are about the web.
Regular users don't optimize their work environment strongly, but that doesn't mean they don't care. And current situation is that full power over site layout is in hands of site designers. I believe this is a bad balance, and needs to be strongly shifted towards good and easily configurable defaults client-side. That is, I believe the situation will improve meaningfully for regular users if we could seriously dumb down HTML5/CSS - while responsible creatives will be saddened, the reality proves that the commercial web in general can not be trusted with so much power over layout.
That's not a reality, though, that's an opinion. If we were to "dumb down" HTML5/CSS and remove degrees of creative freedom from authors in favor of centralized, client-side solutions, we would do so for everyone, not just the "commercial web."
A web where everything looks and behaves exactly the same is a web which is only interesting to a minority of people. That web would probably have never escaped academia and the world would be worse off for it.
(Which makes the "web where everything looks and behaves exactly the same" better than the web where designers fuck with UX to get a fraction of a percent increase on an A/B test, because consistent interfaces are easier on the mind of the user.)
They do care. They just don't know better, and even for us power users content wins over the container. But everybody is complaining about slow websites that don't ever load, and this is leading more and more to the flocking away from the web to the apps. IDK if this is a good thing or not, but sometimes I think that if web became again a web of documents and applications again applications (albeit castrated tbh), that's at least better than what we have ATM.
Some of us consider it a feature that a website allows the reader to freely resize the browser window (and the text with-in) to the width they consider comfortable.
There are few reasons why a line of text should be longer than, say, 800 (css) pixels, but having the browser full screen might still be a good idea for more app-y sites.
But I agree that I commonly see people having their browser window maximized all the time (on huge monitors, to a great waste of on-screen space). I think the fact that they became accustomed to that is an unfortunate consequence of modern web design.
Of course, the Windows users whined and whined and whined about that until Apple decided to make everything go fullscreen super easily.
So anyway, I love seeing people use 34" ultrawides to browser websites with their browser window maximized. It's a ridiculous waste of space. XD
now whether these people are aware of that and whether they say that having 2800x1200 window is fine for them is a completely different thing...
> So if you want to use for example maps on one tab and read text on another, the best UX would be to have users resize their browser when changing tabs?
The best UX would probably for text and maps in this example to be in different windows, to be honest. But that causes other issues with the mental model people have about web sites. Still, having a website-provided width limit helps avoid a click or two required to split the site off and resize its window, at the cost of wasting huge amount of screen space.
Not sure what's the way towards a better UX, though. I myself like tiling window managers, in which I can efficiently position stuff with keyboard the way I like, but I understand this will be too complex for casual users. I also like grid window positioning, like http://windowgrid.net/ which I use on Windows, which lets me flexibly resize windows with the mouse while retaining a nice looking layout. I feel that a solution like this would be simple enough for regulars, too, offering them useful flexibility with little confusion.
Calling these things a framework seems like an exaggeration. I prefer to call them “a stylesheet”.
Sakura seems to be done by someone who's not too familiar with typographic conventions. The distance to a paragraph following a header looks to be the same as the paragraph preceding a header.
What's wrong with Times new Roman? I can read it just fine.
Times was designed for a different world, I would say that modern typefaces designed for the web are easier to read on the screen.
Apart from that, different content works better in different typefaces.
What you despise is not design, I’d say. What you despise is bad design.
Also, those who make websites, just please be nice people and keep the underlines below the links. Why confuse the reader? Why impede quickly skimming through a page for interesting links? If they are ugly, they are ugly to you, that's subjective. I like them, and think they are part of the visual culture of the internet.
And then there are the evil guys who style some random stuff like links, but they're plain spans with color: blue and an underline... Least surprise is best surprise.
Links should be blue and underlined.
Oh my yes. I even recently (as something of a joke) created a Gopher gateway to a web service.
I miss Gopher in a lot of ways. It was my first real experience with the Internet (via the public library). The signal to noise ratio was good, and I think the result of making some very strong assumptions / opinions on what a document client-server should do, and how it should do it.
I’d love to reimagine gopher for the 21st century—maybe with documents and document maps served up as S-expressions. But then I realize that making the document model more complex or flexible than it is could simply lead to a new webby mess.
Gopher is kind of a museum piece right now. It’s a shame, because I feel like it has value in forcing you to think through a non-shitty information architecture. The web offers such a higher ceiling, but all too often a much lower floor as well.
It would be great to have this assumed by web sites. They could offer several pre-made CSS versions, e.g. day and night, right via the user-side mechanism for choosing and editing styles.
Alas, this mechanism is missing by default.
I'll add some human ones:
- vanilla html webpages had way more actual content [1]
- they don't burden you with updates, social sharing
[1] nowadays very often you run into a site with header, menu, flashing area, related articles, footer, cookie warning, etc etc. In the middle of all this you can barely see the first paragraph. And when you scroll down you realize there's only two of them. html5/css3 improved the wrong factors, or at least gave people the wrong idea about where to spend time
Once a graphic webpage has traffic but isn't a front-runner, there's a temptation to keep adding more little graphic pieces of crap that slowly the site becomes completely unusable - see slashdot and yahoo.
Then as you attempt to exit the tab the "exit intent" crap kicks in "NOOOO don't go!".
Sometimes I miss the days when no one cared about the internet.
Though, apart from Ajax, HTML5 lacks a simple way to compose markup at the browser side, thereby avoiding full page reloads, like SGML had over 30 years ago ([2]). HTML imports are basically dead, and weren't great or widely used to begin with ([1]).
[1]: https://developer.mozilla.org/en-US/docs/Web/Web_Components/...
[2]: http://sgmljs.net/docs/templating.html (my project to bring back markup technologies to the Web)
I made my first website with it when I was 11 (still have it on a cd rom)...it was so simple to use. In fact, a lot of stuff in the late 90s to the early 2000s were very approachable as a curious noob. I started tinkering with QBASIC during 1997 on an old 286 with a monochrome monitor (still have it), and when I got a Windows computer and I tried to make a Warez site on Geocities lol....manually searching altavista for 50mb rar parts compressing Playstation ISO images on a 56k modem, hosted on myspace.com (when it used to be a file storage)...it took all summer to compile but again everything was easier because people didn't give a shit what the site looked like, they just cared about what they could get for free.
Now it seems impossible to get away without some trendy UI framework or experts reminding us how we are doing it wrong...which ironically increased the webpage filesize to 196kbps mp3 files that works on modern devices.
Reminds me I have to prototype some low bandwidth mesh network to see how fast it could go with simple html4 era websites.
ps: note that .. I was very eager to have full js websites when the web 2.0 thing started. failed experiment IMO
Please do.
We need a low-bandwidth, low-noise, no BS alternative to the festering mess that is the current state of the Internet. We need a meshnet, for information and learning, not for advertising and spam. The low bandwidth will help enforce simpler content, with less overhead.
You romanticize. We made confusing menus as well with JavaScript (called DynamicHTML back then), Flash, and Java plugins. We were proud of fire and water effects and other gimmicks.
And then Superfish when you wanted to be serious.
Ehm. What does HTML5 and CSS3 have to do with any of this madness?
Most sites exist to bring a profit. If they are paid / subscription-based, they can be nice as heck. If sell a product, they make examining and buying it as effortless as they can.
But the rest of for-profit sites are vehicles to sell eyeballs to advertisers. They afford to the advertisers as much comfort as they can without turning too much readers away. They exist not for the readers' convenience.
So many modern javascript heavy websites that are pain to build, pain to index, pain to maintain, pain to download and consume, painful to support older devices. We got here because it was trendy...and now we've come to a full circle.
I also like to archive many of the articles that I read, but I don't like keeping bookmarks. So I wrote a small utility to convert html to text by combining 'readability' (python) and lynx (text-base web browser), which does a nice job of formatting the content and storing the hyperlinks as footnotes. This archive is also nice and easy to index if you are so inclined.
On my Linux laptop I have no choice but to use a text browser. I currently prefer links2[1]. The speed along with the lack of distractions is something I enjoy when using a text browser. Text only sites like HN are great (except comments not indented). [1]http://pupnik.de/links2.html
By making the text-only version a separate destination, most of their traffic still flows to their multimedia site by default and by habit.
The light version is always advertised by them for specific use-cases only: quick access in emergencies and for usage in "emerging markets" (inaccurate marketing-speak for low-bandwidth connections on low-power devices).
The market forces behind text-only news sites are not unlike those behind AMP links, Facebook Instant Articles, and the Apple News app. News sites want their news to be read, but not at the cost of not getting paid. It's a difficult problem to solve, and these "secondary" text sites typically get the content in front of the most ad-averse eyeballs anyway, so they're not currently a threat.
There seems to be a trend in disconnecting content and interactive input elements, sourcing out handling of comment systems to providers like Disqus. It allows a smaller publishing interface without the need to invest in big backends. As soon as there is no more input handling needed on my infrastructure, migrating to static content is a logical step.
Secondly, I hold myself to the same standards I want to see from other website owners: no bloated adscripts, no tracking and no excessive execution of code inside my browser.
Pages that do not display content with disabled Javascript should rethink their priorities - it not only excludes people with no JS, screen readers or text-browsers; it also fails to make me recommend the link on Twitter or Facebook.
I use Jekyll to build my site, which is served up with nginx. 2+ years and 100+ articles later and I think it was a great move. You tend to strive to keep things fast and minimal, rather than bolt on a bunch of bloated plugins and other things that detract from the reading experience.
It's quick, easy to maintain, and has been completely mobile optimized since the flip-phone days.
It is built around showing you information, instead of information being an excuse to show you ads. Matt understands what he is providing and that insemination of information is more valuable than any extreme monetization strategies.
It doesn't have a confusing landscape of sections, icon fonts, ad videos and promotional content that prevent newcomers from understanding the flow. It takes but a few clicks before you are intimate with the interface.
It is the quintessential "if it ain't broke, don't fix it"
I just wish there were more publications of this style that weren't hyper-conservative and biased.
For another example of a beautifully designed site that puts information first, look at Cryptome.
https://www.alexkras.com/eznews/
If I am actually interested in the story I don’t mind waiting a bit for it to load.
Edit: Never mind, found it linked from your blog!
For anyone else looking: https://github.com/akras14/eznews/blob/master/index.php
Nice work!
Here is the blog post to go with it: Simple RSS Reader in 85 Lines of PHP - https://www.alexkras.com/simple-rss-reader-in-85-lines-of-ph...
http://www.magnusson.io/evil-feed-reader
It does very little, but exactly what I want.
we have seen our culture, the worlds culture, adopt computers and the internet. gaming, social media, all kinds of media distributed over the internet -- we have grown into the internet. and with this development, we are fast approaching a platou, a period of maturity where it is clear what exactly the internet is used for by most people in most cases, and also (critically) how those uses are implemented. when the internet was born, we made browsers with a turing complete language built in and we made all the tools very general, all because we didnt know what was going to happen. we had to be ready for anything. this is simply not true anymore.
off the cuff, the internet is used for reading forums, reading news, watching video and listening to music. perhaps a new kind of browser can be adopted that has those things baked in and nothing else. this kind of browser would meet most of peoples needs while being simple enough to make competition possible, thus aligning these new browsers with the interests of users. these browsers would be much, much safer as well.
for other kind of websites, sophisticated web apps that allow you to interface with a service, like email or time clocking or whatever, perhaps those things could be branched off into a new area where js, wasm and binary distribution live. perhaps a different class of browsers. with these kinds of browsers, there is more pressure to align with users interests because users can soft-boycott them because 90 percent of their needs are met by the simpler browsers mentioned earlier.
and perhaps a new standard for interfacing with documents can be baked into these new simpler browsers to further take necessity and power away from js and wasm browsers. for things like emails, remote document manipulation, assignment submissions, etc, a lot of that could be fit into a standard that is baked into the browser instead of being re-implemented in js for every instance.
a new class of simple browsers would create a new division of the internet that is clean, fast, simple and worry free much like these text only news sites.
if one could bake payment into the browser as well, creating the most friction-less payment experience on the internet ever seen, we could see a re-vitalization like that seen with patreon all over again.
If someone wanted to create a web-alternative for high singal to noise content in 2017, I wonder if public libraries wouldn’t be the ideal test case.
I've recently started using an RSS reader again and I'm finding that I can find some feeds, but what I don't know where to find is a good place to discuss the articles. More, it does feel like the feeds are a second class citizen, at best. Is amusing to see the page that lists some feeds, such as https://fivethirtyeight.com/features/fear-not-readers-we-hav... still list the demise of google reader. Which, sadly, reminds me that I haven't checked google+ is about as long.
The general trend was that news outlets generally offer some text accessibility settings, and big tech company apps did not offer any text settings at all.
My conjecture is that the minimalist design trend in Silicon Valley cuts against having many user-configurable options, which unfortunately affects accessibility.
1: https://medium.com/@BeeLineReader/the-importance-of-text-acc...
The phenomena we see as Gresham's Law, which does and always has exerted itself far beyond coinage -- the first use is found in Aristophanes' "The Frogs", in ancient Greece, referring to both coin and politicians, concerns a fitness or reward mechanism, within some environment, specifically to complexity or quality.
This can express itself as both a non-reward of higher quality, and of a temporary seeking of greater complexity, depending on a number of factors (and I'm still trying to sort out a description of how these may manifest).
In the instant case of Websites, chasing advertising dollars has rewarded more complex, and paradoxically, less informational sites, which are expensive in terms of the compute resources required to deliver them. There is a possible inflection point at which simpler, lighter-weight alternatives can present both more and better information, sufficient visual appeal (see my comments and links elsehere in this thread), and at far less complexity.
This also shows up in various technologies and progressions, particularly as a novel "worse-is-better" option emerges and overturns an established method or product.
Also if you want to convert any existing website to light weight you can use https://googleweblight.com/?lite_url=<News Article URL>
Then use UNIX filters to transform it into the format one prefers.
For example, NYT has a jsonp feed I have posted before.
# replace "world" with whatever NYT section you prefer
curl -o 1.jsonp https://static01.nyt.com/services/json/sectionfronts/world/index.jsonp
exec sed '/\"guid\" :/!d;s/\",//;s/.*\"//' 1.jsonp
As an experiment to enjoy text-only versions of popular news websites, convert the rss to html for a "text-only" version.
The author of radare2 wrote a filter called "rss2html" some years ago and it works well enough for this experiment, so there is no work involved here.[1] rss2html will transform a page of rss xml to html, txt, etc. Build it with libcurl and it can fetch pages too.
curl -o 1.xml http://www.bbc.co.uk/news/rss.xml
rss2html 1.xml > 1.htm
browser file://1.htm
http://www.bbc.co.uk/news/rss.xml
http://feeds.reuters.com/Reuters/worldNews
http://www.latimes.com/world/rss2.0.xml
http://www.wsj.com/xml/rss/3_7085.xml
http://rssfeeds.usatoday.com/UsatodaycomWorld-TopStories
http://feeds.washingtonpost.com/rss/world
Pitch: A "podcasts for textual content" optimized for low bandwidth.
An podcast like app that lets you subscribe to (textual content) creators, auto downloads for offline reading, allows reviews on individual posts or entire series, presents content in an ebook like aesthetic manner, and has the lowest bandwidth for transmission and storage because: 1. Transmission and storage is in markdown (not HTML) 2. Pretrained language models shall result in higher compression ratios than regular gzip of HTML common in browsers.
Feedback appreciated
I subscribe to the local paper, some financial blogs and some tech blogs. They are preloaded onto my device and presented to me in a common format every time I charge it.