← Back to context

Comment by treesknees

3 days ago

Several news sites offer text only versions.

https://lite.cnn.com/

https://text.npr.org/

https://wttr.in/

More listed at https://greycoder.com/a-list-of-text-only-new-sites

It’d be great if there was some standard that allowed these to be easily found, and supported on the local news sites.

That CNN website is great, except it still has a huge cookie banner. Looking at the cookies of the site, I think the only cookie it sets is that i clicked on the banner. Most of the size of the page is also related to the banner it seems.

  • You can’t put a price on some round-rim glasses wearing EU bureaucrat named Klaus-Dietrich von Regulieren sleeping soundly because of that banner.

    • If I’ve understood the grandparent post correctly, they don’t need the banner. They wouldn’t need it if the only cookie they set were a functional 1st-party cookie, and since that sole cookie is just to track cookie banner status, they especially don’t need it.

      6 replies →

I looked at a CNN "lite" article, and it includes 560KB of stuff (lots and lots of CSS declarations) in addition to the actual 11KB of article content.

  • While still wasteful, CSS is one of those things you can do astronomically wrong before it starts being noticeable. Case in point here: inlining 560 kb of CSS with the page and just sending it with the entire HTML file each time is only ~61 kb of actual network transfer to load the article (due to brotli encoding).

  • The point is that it would still work if you block JS, CSS and graphics.

    Check it out in lynx for example

In the Netherlands the public broadcaster still publishes news through Teletekst:

https://tweakers.net/reviews/11700/hoe-werkt-het-vernieuwde-...

  • Which is cool but it's not web, and few people have working TV reception that supports it at the moment. The web version of Teletekst (https://nos.nl/teletekst) is over 3 MB in size.

    • It's better than web! You don't usually view it with Internet. You view it with either an antenna or with broadcast cable. It's up even when the Internet is down.

    • Are there no APIs with smaller and faster responses? The Swedish equivalent has --even a fairly respectable terminal client.

Using the lite subdomain is a great way to read all the subscriber articles as well. Was reminded of the lite site during some annoyingly aggressive A/B testing CNN was doing a few months back.

In terms of a standard, it would be nice if "reader mode" were standardized to request a text-only minimal formatting version of the site.

  • Oooh... can you imagine if servers actually took the hint and sent only text if the client provided Accept: text/markdown, text/plain headers?

    • > Accept: text/markdown

      funnily enough, the rise in agentic coding has actually made this on the rise

Maybe there could be a service which translates any website into a trimmed-down text-only version.

  • Firefox's Reader mode is pretty great. Doesn't reduce network traffic but makes almost any page more readable.

Thanks for sharing, i almost was not sure if the last part was sarcasm. Html itself was the standard, then when it got bloated we got rss. This seems like it’s not a problem of a lack of standards. It’s the company choosing not to promote it.

  • It is more to do with the fact that vast majority of people aren't going to bother subscribing to an RSS feed. I am on a freelancer slack group that supposed to have a RSS feed for jobs. The feed is often broken for weeks because most people don't use it.

    Even when it isn't broken the display output is broken in Thunderbird because the dev isn't going to bother checking Thunderbird as many people don't use email clients like that anymore and instead use webmail.

    I never have used RSS that much as normally if I want to check for new things on a site, I will just go to the site and look myself.

  • I suppose I meant more of a best practice - if every news site could be found at the subdomain of lite.XYZ.com, or perhaps some way for the browser to request specifically no images or styles, it’d be easier for the end user to find.

    RSS is a good point that I didn’t consider. Although it tends to be a summary and hyperlink to the main site.

    • Ideally IMO this would be accept headers. You're asking for the same semantic content but a different format. I'm not sure if there's a nice way of specifying html but in a minimal sense (we do quality with images, perhaps linked), however these could mostly be text/plain or text/markdown (and it'd be nice if that was then formatted properly by the browser).

      This often makes a really nice API if you can do other formats too - the main page of cnn could respond to rss accept headers and give me a feed for example.

  • RSS is just a list of links to webpages, maybe with summaries. The readers generally fetch those webpages and filter out the text, but every browser has equivalent functionality now. You can do it with literally any HTML page, though some websites try to fight it (since depending on the reader, it neuters ads).