Ask HN: Who wants to help promote RSS?
Regularly on HN we see threads about the apparent demise of the feed file as a technology and how sad it is. Despite the answers to those threads showing that RSS/Atom is not dead at all, what is clear is that it is not in a great spot. Less sites support it, percent-wise a smaller amount of internet users knows what it is and how it might be of use, browsers dropped their integrations.
I have a plan on how to counteract this. It includes getting people together who feel that this technology is elemental for the open web and should be preserved. It then vaguely continues like this (not necessarily sorted by feasibility):
1. Open PRs for browsers with provided-by-us code to integrate RSS again, with UI elements that make existing feeds visible and supports subscribing via a feed reader
2. Run a site that can be used for this, that takes a "I want to subscribe to a feed"-request and presents RSS readers to do so.
3. Implement default feeds for CMS that are missing them.
4. Implement easy to use libraries for programming languages that might miss them.
5. Create a website that explains why this is important, that hosts or links to the specifications and that contains additional documentation.
It's nothing I can do on my own, but I really think that if some people come together, that we could have a real influence on this development. Even if the browser integration step fails (which might be likely) there are alternative ways forward. And I'm sure there will be many more ideas on what could be done.
If this is of interest to you, consider joining https://gitter.im/FosterRSS/community.
169 comments
[ 3.2 ms ] story [ 244 ms ] threadThe code is all open source, at https://github.com/symkat/BlogDB/ and I wrote an article on my own blog about initially writing the software from design to deployment: https://modfoss.com/building-blogdb.html
One minor thing: The preview images take up a lot of space on mobile, I have to scroll forever just to look through all the options. Perhaps the preview could be text only at least on smaller screens?
https://en.wikipedia.org/wiki/Atom_(Web_standard)#Atom_compa...
(a) That RSS doesn’t support relative URLs, whereas Atom supports xml:base. I think some clients support xml:base on RSS, but certainly not all, so all links should be converted to absolute in RSS for compatibility.
(b) That RSS titles are of an unspecified format and bodies are only mostly HTML, leading to things like angle brackets in titles being mangled in different and incompatible ways by different clients, and plain text bodies being unreliable (indicated by the <content:encoded> mess which is I think theoretically identical in semantics to <description>, but they felt the need to duplicate it because clients had made such a mess of <description> because RSS developed “organically” in the worst way). By contrast, Atom gets it all right, allowing you to specify in each case whether you’re providing text or HTML. (And HTML is useful in titles: things like <em> and <code> are completely reasonable, and it’s a crying shame that almost no content management systems support inline formatting in titles. But if you are converting to plain text and know you’re dealing with HTML, you can strip tags. I do wish it supported both, e.g. “You don’t need <code>cat</code>” and “You don’t need `cat`” so you get backticks if monospace formatting is denied you.)
RSS is a mess that should have been completely retired about fifteen years ago. If you’re doing podcasts you have to use RSS because Apple started doing podcasts a few months before Atom was finalised, and they’ve chosen to just keep layering hacks upon hacks, namespaces upon namespaces, duplication upon duplication, for the next 16 years, rather than ever switching to Atom which would have resolved a lot of that (and maybe worked to standardise other parts, rather than duplicating fields half a dozen times for RSS, Apple, Spotify, &c. &c.). But if you’re not doing podcasts, please let RSS die. As a term, too, in favour of just “feeds” or “web feeds”, because it’s harmful, leading to people only hearing of RSS and so implement RSS because they didn’t realise they should be implementing Atom.
how many github pages exist? currently, that number is apparently 2,866,300 [0]. imagine a tool that generates an rss from all the html files collected. before pushing back with git, you run a script that generates the rss. any new html files are added to the rss. and it doesn't have to be just github pages, it can be any html/blog website. i can build such a tool but am currently working on something else.
0: https://github.com/search?q=.github.io
To be clear, I also think the web has gone wrong and miss coherent and useful web search, exploring interesting user created content (people have mentioned the return of web rings) and curating my own news feed via RSS.
But it's strange to me that when I started everything was about what we were going to do in the future, and increasingly the conversation is about going back to the past.
I also like mechanical calculators, horse drawn plows, and fine tinsmithing all for the skills and craftsmanship that go into each. And yet I wouldn’t think to start campaigning to bring back animal powered plowing. The new generation of technologists and programmers entering the workforce doesn’t give a damn about your ugly RSS feeds or your plain text productivity workflow or your finely tuned EMacs config and I think that mindset shift is going to be irksome to a lot of established old heads.
Besides, I could say the same thing about electric cars. Your neighbor's new Tesla? Why is he going back 100 years in time? Nostalgia? Or new potential?
However, I see the death of RSS as the symptom of a larger problem: when platforms get big enough, they restrict access to their data. RSS feeds disappear, but so do other machine-readable endpoints. If it wasn't for GDPR, there would be no way to export that data. GDPR gave us clunky one-time exports, but even those are often incomplete. Making RSS popular again won't fix that.
The industry has a strong incentive to kill RSS, since the readers can strip the valuable bits (content or data) from the business bits (analytics, monetisation). RSS users are hard to count or monetise. This might not suit every business model.
This is a battle worth fighting, but it's not one you should expect to win.
You say that the industry has incentive to kill RSS, and yet Google has just added a Follow button to Chrome for RSS feeds. Go figure.
Chrome is not the industry. It's a web browser built by a content aggregator, not a website.
If we're talking marketshare, Chrome won: https://gs.statcounter.com/browser-market-share/desktop/worl...
Disclaimer: not a Chrome user.
I get why they do this as a new listener will want to start off with a full list of episodes and may want to pick up a show from the start.
It would be useful if an RSS could specify a separate URL to query that will only list new episodes, returning an empty RSS (or a standardized HTTP status code) until the new episode is available.
The original RSS with the full list would still be the primary URL for the podcast, the one I get when I right-click and select "Copy RSS URL".
I don't use my GitHub account for logging in to other sites without a decent amount of consideration (I didn't sign up to GitHub to use it as an SSO system and generally don't like using it for that purpose). And I have neither of the other two accounts.
So I would agree with the parent that removing some of the friction and offering people an email form to drop their email into might get you a higher uptake?
I am however interested in these efforts. For reference, my email is in my profile if that helps.
I created a form: https://docs.google.com/forms/d/e/1FAIpQLSc9lWrieLAn9azMI_88... . Does it work for you?
But the problem lies elsewhere: the blogs I regularly visit still have RSS feeds, but most content now isn’t in blogs anymore. For-profit sites have crippled or completely dropped their feeds to make users see ads. Many people that would have been bloggers a decade ago just publish threads on twitter now. And discoverability is a whole topic for itself…
So even if rss was in-your-face in every browser, how would you get content creators to support it?
My hope is that the one would follow from the other. If consumption is easy and awareness better, creators would want their content available via a feed. Think Youtube: With ads integrated into the videos every feed subscriber is a money source, both for the platform and for the creator.
I know a podcast I like but never listen to it, because it's behind a strange podcast platform without a feed. If they had that they had at least one listener more, and more listeners would mean more potential for advertising. Simple awareness could help here.
Also, I think it's somewhat of a misconception that ads are impossible to add to feeds. The article itself can contain them, the feed can even have special ads, there are no rules against that. Sure: We are talking static ads like banner ads, not Javascript-tracker-driven dynamic adspaces, but that's a direction the web should go to anyway. There has been technology and industry in that direction in the past, why shouldn't it exist again?
Enjoy your new RSS Feed about Quantum Mechanics, featured by BIGCOMPANY, we sell the shitty products you like.
[1] https://feedburner.google.com/
I also want to use the opportunity to propose a new name for RSS, which is 'Really Social Sites' or 'Real Social Sites'. RSS had name changes in the past, so it's only natural it gets a new name when it will revive in the coming years. And since all this web3 sh1t (names without meaning) I really felt the need to put energy in RSS and also it's name. To make it more accessible to everybody, not only devs and nerds, it needs a better name. Nothing else has to change, RSS is rock solid as a technology. Can't wait to use my own software to read the personal blogs/websites of all you guys and gals. Social Media without Big Tech.
I began to expect my lists and read status to be synchronized across my phone, tablet, and laptop. This meant using a service, and I really didn’t want another paid or fermium service.
This all changed last year when I finally set up my own instance of freshrss which I love. But I don’t know what you’d do for “regular users”
Even in the heyday of RSS how many of us just had an overgrown 10000+ unread RSS reader that we simply stopped using?
Promoting RSS seems backwards to me because that ship has sailed. Like promoting IRC when people have chosen solutions like Discord and then refusing to acknowledge why people prefer Discord. RSS (and IRC) didn’t just fall out of favor because people refuse to see how amazing it is nor is it just because of ads.
The answer to the post-RSS world is something more like a better, polished Fraidycat that can track content on any website. We can do better than RSS: a solution that doesn’t rely on everyone implementing it.
You’re right about RSS. I just see a better solution for this idea of a content-delivery system being one implemented as a smart client that can handle all websites rather than a dumb client that hopes every website in 2022 implements RSS. That way there’s no “you’re just using it wrong.” ;)
I don't care what you use it for. You can use SMS messages primarily to make a phone stuffed in your underwear keep vibrating, but if you want to complain that it's an inconvenient teledildonics protocol and client, that's on you.
Though, if you think you can code up a "smart client" that doesn't use puny RSS, but can "handle all websites" and will work better at tracking and delivering content to you than RSS does on zillions of websites (including new ones), you open up your editor and get at it, then. Have fun.
I find invoking the NYT especially funny because they have MANY RSS feeds, letting you filter by sections.
The Sunday NYT is a bit of a challenge, mind you.
I agree that the way how RSS clients were designed 10 years ago wouldn't work well today, but that is an a problem with architecture and UX design of rss client. I don't think that it has anything to do with the RSS or ATOM XML data structure standard.
Having machine readable data in open format is useful so that anyone can use it, and create a it's own solution without spending too much time on interpreting the data. While with unstructured, javascript heavy sites one would need to build it's own little search engine to be able to process it and desing/train it's solution for reasonable information extraction, which increases barriers to entry.
As many websites are built by clever people and/or based on somewhat ready-made frameworks, when you DON'T see the RSS/Atom icons appear -> take a look at the source of the website and search (ctrl-f) for the word "rss" or "atom" or "xml" -> you might end up finding a surviving RSS link in there.
https://www.rssboard.org/rss-autodiscovery
Imagine a service that offers a social media experience comparable with Twitter/Insta/Tumblr with rich clients, but under the hood it's all RSS, so not only can you subscribe to third-party feeds out of the box but your "profile" is also an RSS feed that other people outside of the platform can also subscribe to.
I think a way to do this and still allow private or restricted feeds is to take advantage of the fact that URLs are encrypted under HTTPS. Have the RSS served via FastCGI or a dedicated handler that verifies read permissions for a URL token against a permissions database.[1]
I think you'd want the permissions database to affiliate auth tokens with email addresses or mobile numbers.
When you request to follow someone, you input your email address or mobile number. If they approve your request, you'll receive a link to generate a feed URL with your auth token included. You can then easily input that into your feed reader of choice.[2]
If you lose the feed URL/auth token, you can request a new one using the same email or mobile number. When and only when the resulting link is activated will the old auth token mapping be replaced.
--
[1] This design is inspired by Slashdot's old auth system, which allowed you to log in via a saved URL. While that was probably horribly insecure over HTTP, you could restrict it to an IP/subnet. In any case, now that HTTPS is widespread the security concerns are substantially mitigated.
[2] An issue I haven't figured out is how to prevent feedly, inoreeder, etc. from caching private feeds.
I do think there's a gap to fill in RSS based social media, because users of current social media would like to send a private message (or at least not public) from time to time. I basically cover this with a contact form and email, at the moment. Sounds crude and not very techy, but it works like a charm.
I don't use social media except a single, oft-neglected Mastodon account and get all my info from RSS. According to Newsblur I have ~500 feeds subscribed and ~5000 unread items (can't keep up) every week.
I don't doubt that RSS is getting less popular, but perhaps due to moving in circles where RSS is popular I really don't feel like it's dying.
Apart from that I'm in a similar bubble as you, RSS never stopped being relevant to me.
I am certainly interested in trying to improve this. There has to be a less privacy invading and engagement promoting system that can work and compete with the mess that is the big silicon valley sites.
You can follow aggregators, you can follow curators, and you can filter what you follow. Discovery comes from what you get from who you follow. Humans going, "This is interesting.", you agreeing, and you following another blog or outlet.
Anything else definitionally is the "Silicon Valley algorithms" people say they don't want.
You connect to RSS feeds when you "thumb up" or directly submit a link that this feed posted. The more of the content you like from the same feed - the more other content from that feed will be prioritized. So feeds that post too frequently naturally go down in the ranking. And feeds that have a high signal-to-noise ratio are at the top.
On LinkLonk I am subscribed to 202 RSS feeds (you can see your subscriptions here https://linklonk.com/feeds), and I don't get overwhelmed by too much content.
You also connect to people who like the same content as you did. In that case each user is similar to an RSS feed (a user has "channels"). The more in common you have with a user - the more content you see from them.
No obnoxious adverts, no trackers and it's free. Would love to get any feedback (email in profile).
From posts like this, it seems like RSS still has a hardcore group of fans but it is not something everyday users ever found that useful.
2 was different. I was thinking about what happens when a user clicks on a feed - it's not very user friendly, especially now that browsers just show the XML file content. If we had a mechanism that wraps that request and gives the user a way to subscribe directly via one of the many available feed reader, that would help a lot.
The idea is not completely novel, https://github.com/superfeedr/subtome goes into the direction but I think it is broken. A website carried by an independent group could be a solution. Maybe even something that could help in case browsers are open for some changes here, but don't want to create the UI part. Or when working on extending a CMS and instead of implementing that UI again, we use that mechanism to wrap the feed. As a web developer making it a website was probably an idea near to me, but there might be better solutions.
I agree. But I'm also willing to think that a simple button to copy the feed URL would suffice. Since feed URLs can be read by all kinds of reader software (and you don't know upfront which ones, definitely way too many possibilities) there's no way as a subscriber to go from a website with a feed link to your reader app without just copying that link somehow. It's too much baked in to the protocol to do it otherwise, I guess.
But sure, the button to copy the feed on by default would already be a great improvement.
https://www.rssboard.org/rss-autodiscovery
Back in the day, I think Firefox even put an icon in the address bar when a feed was available.
1) Do work and support RSS or
2) Do nothing and we do it for you for free
Which do you think they'll choose? And then you're stuck running a big clearinghouse for RSS feeds, with fund-raising problems, users griping at you because some feed isn't perfect, etc.
Myself, I've had decent success just asking if someone could please add an RSS feed to their site the few times it's come up. It's apparently easy to do in many CMS's that don't do it by default.
https://indieweb.org/ActivityStreams
https://indieweb.org/ActivityPub
This is their podcast where they talk about what they are doing; https://podcastindex.org/podcast/920666 That podcast RSS feed; http://mp3s.nashownotes.com/pc20rss.xml
They also have their own mastadon where all the developers working on Podcasting 2.0 hang out. https://podcastindex.social/web/home