It's time for an RSS revival (2018) | Hacker News Reader

It's time for an RSS revival (2018)

wired.com

222 pointsalexzeitler92 comments

I do love me some RSS. I started using RSS seriously 2 years ago, and last year, I forked Bubo RSS[1] to have my own RSS reader[2]. It's just a static site that is built every hour using Github Actions (although you can easily turn this into a cron job on your Raspberry Pi or whatever).

I liked the result so much, I ended up doing something similar using RSS feeds to a build page with all my of Youtube subscriptions[3].

[1] https://github.com/georgemandis/bubo-rss

[2] https://kevinfiol.com/reader/#dailies

[3] https://kevinfiol.com/youtube/

EDIT: The Github repos if anyone is interested: https://github.com/kevinfiol/reader + https://github.com/kevinfiol/youtube

That's awesome! I'm tempted to try this approach on my website, but using Next.js' ISR to regenerate the feed pages periodically.

I’m becoming a real fan of using RSS for anything that fits it – websites, YouTube [0], email newsletters [1]; even books [2]! Honestly I’m tempted to start putting summaries (hacker news, stack exchange, reddit) in my feeds, just to keep me from endless scrolling.

[0] I think Feedly can handle this, but YouTube may have disabled this natively, which is a shame.

[1] I’m a fan of https://kill-the-newsletter.com

[2] Dracula Daily (https://draculadaily.substack.com/about) and Letters from Watson (https://lettersfromwatson.substack.com/about) are quite fun!

Feeds are definitely still enabled on YouTube at

https://www.youtube.com/feeds/videos.xml?channel_id=UC<xyz>

The only complicated part is finding the channel id, which is not always presented directly (sometimes you have a username, sometimes a channel "shortname", now you also have handles... even the YouTube data API itself can only resolve some of them).

If you give a channel URL to NetNewsWire, it will figure it out for you. I definitely recommend it, it's the golden standard of RSS readers.

To get the YouTube channel Id: on the channel home page (you may need to reload the page or open it in private mode to avoid caching), open the developer console and enter:

ytInitialData.metadata.channelMetadataRenderer.externalId

I am using the RSSPreview [1] Addon in Firefox, which works for me also on Youtube. The addon brings back the nice RSS Icon in the URL Bar. You can also change to color back to orange so you can spot it easily when a site has an RSS feed.

[1] https://addons.mozilla.org/en-US/firefox/addon/rsspreview/

My best idea was to subscribe to the RSS feed of my getpocket.com account. That way I can save any page anywhere to read it later in my RSS reader.

I created a bookmark to watch YouTube videos from Feedly, it basically creates a playlist from the YouTube videos you're following (20 at a time):

    javascript:!function(){const%20e=document.getElementsByClassName(%22entry%20entry--unread%22),t=[];Array.from(e).forEach(function(e){const%20n=e.getElementsByTagName%28%27a%27%29%5B0%5D,o=/https:\/\/www\.youtube\.com\/watch\%3Fv=(.+)$/.exec(n);o%26%26t.length%3C20%26%26(e.getElementsByClassName(%22EntryMarkAsReadButton%22)[0].click(),t.push(o[1]))}),window.open(%22https://www.youtube.com/watch_videos%3Fvideo_ids=%22+t.join(),%22_blank%22,%22noopener%22,%22noreferrer%22)}();void(0);

I still use it on a daily basis like many others.

But a revival isn't dependent on the technology or the type of apps that can consume RSS. It is the lack of outlets publishing on a regular basis. Social media's grip on attention has only grown stronger since RSS readers were more popular.

Blogs are mostly gone comparatively speaking to 10-15 years ago. Many authors do an email newsletter.

RSS is still around but it's a content supply problem and not a technology issue.

Many trade publications and media outlets have RSS feeds but not all of them.

Right. RSS still has some value today.

But it had far more value in a world with a lot of individual blogs, many of which were only updated occasionally, and without other reliable discovery mechanisms for those blogs--other than going to a site every few weeks to see if there was something new.

Have you considered an RSS reader that "reads" non-RSS publications?

That's what I've tried to do with Lenns.io. You can subscribe to any blog/website, regardless of their RSS support. Then, if they don't expose an RSS feed - the website's titles are followed. That works with about 90% success.

I'd be happy if you give it a go https://lenns.io. Bonus - you can set priorities to feeds and the number of items per source to be listed on your dashboard (so not a single subscription can overrun it)

> It's time to head back to RSS.

Some of us never left.

My sentiments exactly.

It's time for this, it's time for that...

Imagine how much further away humanity would be if these critics actually sat down and produced what they think is missing.

I have a family to support and I'm slowly and painstakingly crawling out of a huge hole both in physical and mental health. And my job is fairly demanding.

I don't have the bandwidth. Many others don't have it as well. So whoever feels so strongly about stuff -- by all means, please create the thing! You'll be celebrated.

Historically it's been pretty evident that "raising awareness" is a very frail strategy that doesn't achieve anything short-term.

I'm in an unique position that I have half a year now to produce...well whatever I choose to produce. What is the thing that you're missing most?

Distributed chain-of-trust RSS would be nice as a start.

Above everything else though, I am missing a decentralized, globally automatically replicated, blockchain-like storage network.

Think software releases. Package X releases version 1.3.11. We should have that in an IPFS-like storage network + several hashes (SHA-2, SHA-3 and what have you) so it can never be tampered with. We should also have that version linked to the previous one -- 1.3.10 (a la Bitcoin transactions I guess).

That's only one application. Another one would be: a network of friends and families all sharing e.g. huge photo travel albums. Or, if they are into it, torrented movies or music. Game hacks even. Whatever, it should not matter.

Wanna dive deep into a rabbit hole? Here:

- http://www.nncpgo.org/ (pay no attention to bad site, the idea of the project and what has been coded so far is solid)

- https://changelog.complete.org/archives/tag/nncp

- https://www.complete.org/nncp/

That's what we're missing. Everyone constantly acts like hosting 10GB somewhere is a huge deal and I am honestly sick of it.

Were I to retire financially tomorrow I'd work tirelessly on all this until my brain can no longer do programming.

We need a new internet. Badly. We need it 10 years ago.

BONUS POINTS: ability to automatically exchange keys in a neat UI e.g. I send a special message to a friend: "Join my inner circle of shared storage" or some such, they click it and the software does authentication and checking if both parties are who they claim to be, issue new keys (so them being compromised has a minimal impact). Think ZeroTier and/or Tailscale but even more self-hosted and even more reliable.

I don’t quite understand how your desired solution for software supply chains actually achieves much.

Content addressability doesn’t seem different from checksums or subresource integrity.

IPFS requires file pinning, so who is owning that and how is it different from a current state package distributer?

Further, someone/a build-machine still has to upload to the storage network… so Now we need smart contract to build (plus a coin)? … then we’ll need a coin for the compute providers?

And even if we have all of that, Charlie still employs his advanced supply chain attack upstream like he normally does… and we’re still left with malware plus a bubble of infra that didn’t quite solve much it was funded millions to solve.

I think there are more people who are just looking for the next Twitter than who are going to switch to RSS. That's not what I'd like to see happen, but it's the sense I get. Whatever people get from those platforms, it is not available in an equivalent dose through content syndication. Most likely they'll go back to Twitter, or one of these other platforms will win out, but the least likely outcome is that we return to a pre-social media world.

Interestingly pasting the article URL into my reader doesn't discover a feed for this site.

I can't even read the full article without subscribing so I wasn't surprised. But I found this list of feeds https://www.wired.com/about/rss-feeds/. I think there's some value in just getting the headlines but I'm sure HN already catches the worthwhile ones from Wired.

There's another RSS related post on the front page without a paywall https://news.ycombinator.com/item?id=34105572

Two RSS submissions today, from different sources. Interesting.

Haven't seen this in the comments for either: https://miniflux.app/

A very lightweight self-hosted web client for managing RSS feeds. I very much like it.

I love using RSS to subscribe to blogs. Many of my favorite personal developer blogs have RSS feeds, and I have them on my blog. I'm currently using Readwise to get all my feeds in one place. This filters out a lot of noise prevalent in social media like Twitter.

Feeds are NOT just a way to craft a personal selection of news in a personal LOCAL UI, with optional full-text-search, local storage for articles, tagging etc BUT also a simple form of asynchronous/lazy push-pull notifications.

For instance we can have feeds for:

- all software project we use/follow, to get info about a new release without having to go to a site/read an ML etc and perhaps finding the Changelog directly in the feed;

- be advised of a new invoice from some company, a simple way no enterprise I know offer, but on par with emails to give an invoice, perhaps with the attached pdf and an easy to auto-parse data fields to munge easily relevant infos in a personal system, like match the invoice with a banking transaction, adding it to a personal agenda etc ALL with few lines of code from ALL parties instead of utterly complicated crapware monsters;

- follow some books authors or special offers as a form of self-chose, not invasive ads

- automation as machine-to-machine, using the XML in automated (and eventually authenticated) fashion.

The real issues are:

- all modern IT is designed for CONSUMING contents in eterodirected ways, users MUST BE powerless dumb sheep;

- most users actually ARE CL from Simon's BOFH memories with systems force them in such sorry state.

Yes, some kind of asynchronous personal agent is a great idea. It is the ultimate anti-social media, your very own "super app" that handles your digital life for you.

I want it

The issue here is that you can only in part, the other part are the others who offer good feeds or not... Like journals who do not offer feeds, some that offer them but not with full contents making them of little use, projects who do not offer release feeds and so on...

Even most banks, at least in EU do not offer decent AUTOMATIC export of transactions, most offer some kind of limited and limiting manual export but nothing more.

That's the real damn issue... We can have a super-simpler IT infra for the humanity instead of the immense gazillion of crappy crap we have, but we miss it because someone for a personal corporate/political interest want a limited and crappy state of thing to have some short-term power...

Last week I tried to find all of the Odd Lots podcast episodes with Matt Levine as a guest. This turned out to be impossible in the Google Podcasts app that I normally use - you cannot search a podcast for keywords, and searching all podcasts gives you the wrong results. Incredible for a company that built their business on searching.

So instead I imported the Odd Lots feed into my tt-rss instance and immediately had the answer.

I am working on a smart RSS reader that I can make a demo of myself using.

I am ingesting articles w/ superfeedr, sending them to a webhook answered by a lambda function, putting the items into SQS, and then picking them up and putting them in an arangodb database.

It will show a single item at a time, Tik-Tok style, and you can thumb up or thumb down. 2004 vintage text classification was able to do this well with 1000 judgements which a moderately avid user could supply in 10 days looking at 100 articles a day.

It won't be easy to sell unless it can learn on many fewer judgements than that so it needs some semi-supervised classifier that does better with fewer judgements. A real product also needs some kind of clustering to deal with the reality that an event radiates stories gets echoed many times by different sources in the media and blog industry and whether you are interested or not interested in the event there is some limit (usually) of your tolerance of the event dominating the stream.

RSS is too lean, it's a format, the majority of people don't know what it is or how to use it. A lot of sites don't even support RSS anymore. Like IRC, its userbase is tech people. Sorry guys, you don't make up the rest of the world...

It's an outdated concept, we all know people don't surf the web anymore. And the larger it will get, the less likely people will. Have you tried using RSS, just measly collecting your favorite sites with the off chance that they support it? It's a mess and no surprise it died as everyone flocked to Facebook, Twitter, etc.

People need some sort of platform and that's why ActivityPub / Fediverse is a lot more promising. Social networks aren't the problem, just the one's we've been using.

Discussed at the time:

It's time to head back to RSS? - https://news.ycombinator.com/item?id=16721690 - March 2018 (491 comments)

Also, related ongoing threads right now:

Ask HN: What RSS Reader do you use in 2022? - https://news.ycombinator.com/item?id=34108413

This is the year of the RSS reader - https://news.ycombinator.com/item?id=34105572

RSS fails miserably on 2 things:

1. Discoverability 2. Monetization

Consider YouTube - it provides discoverability through the algorithmic feed. Giving creators a distribution platform they can't find anywhere else. It also provides monetization through ads, sponsorships and direct support to creators.

Consider Newsletters - no discoverability built in, but it does support monetization. Pay to subscribe, and only you will get the newsletter. RSS feed can't limit only to subscribers.

Authenticated or unique feed URLs are "a thing", but overall the difficulty in "monetization" is a feature to me.

I've been a heavy RSS user since 2005. It has served my purposes well and I'd just as soon it not be "disrupted", personally.

>RSS feed can't limit only to subscribers.

This practice was common when RSS was popular and is still used today by newspapers and other media outlets.

An individual feed provides only a snippet view of the article which encourages you to visit the site where a subscription was required.

You're not wrong. But I do wish more people were happy with those "failures", which basically describe the Internet of yore.

I'm a creator myself.

Why would I want to build an audience on a platform with no monetization possibility, ever?

Even sponsorships are not possible, since I can't reliably count reads (which I can even in email).

It's nice to think of the "classic" internet, but it's a naivety that will keep RSS on the edges. And I love RSS idea! I want it back! I hate subscribing to email newsletters in a way that makes it hard to manage and organize.

Plenty of creators create without the expectation of direct payment. Of course, many are not in a position to do so as well. But I expect that, for the most part, the platforms make a lot more money off the creators attempting to monetize than the creators themselves do. And, in so doing, they make the overall experience worse.

But, yes, RSS was always pretty fringe. And today there's enough hope being dangled that a newsletter or YouTube channel will make it big that it's even more fringe.

When counting email reads, how do you account for individuals who configure their mail client to not download images by default? I do this in Gmail to avoid tracking pixels from being loaded. I imagine there’s a sizable subset of more technical users who do this?

You can subscribe to newsletters via FeedBin, which costs $5 a month, and will act as a source to many feed readers (as well as being a web-based feed/newsletter/twitter feed reader itself).

I'm also a creator

I do it to hopefully help people out in life. I get enough traffic and thank you notes to make it worthwhile.

> Pay to subscribe, and only you will get the newsletter. RSS feed can't limit only to subscribers.

Patreon has this feature with their podcast RSS feeds.[1] When you subscribe, you get a URL with a personal, unique parameter. This feed has all the episodes that you pay for. Your version of the feed stops updating after you cancel your subscription.

[1] They probably also have it for normal RSS feeds, but I've only used their podcast RSS feeds

>RSS feed can't limit only to subscribers.

Authenticated rss feeds are a thing. I have subscribed to some premium podcasts which have a auth token in the feed url for authentication.

This allows paid content, and most importantly revocation of access at a later date (if e.g. the subscription is cancelled).

Some of us who never left RSS think those two things are features

>RSS feed can't limit only to subscribers

The feed doesn't give you the whole article. Only the first paragraph or two. If you've ever seen someone write "after the jump" they're referring to the content that you can only see when accessing the page directly.

I recently got a remarkable 2 tablet. If you install Koreader, it has a neat little function that generates ebooks from an rss feed!

Calibre has a similar feature I'm playing with. I'd like to just plonk it on my homelab and have it automatically scrape my rss feeds and throw the resulting ebooks at my tablet over the network. Doing it on-device is a bit slow due to the low power processor, but it's liveable

I think having a daily digest RSS reader truly helped me with my productivity as now I check it only one time per day instead of infinite scrolling.

Initially I used Mailbrew, but for the last 5 months I've been using my own open-sorce tool: https://github.com/piqoni/matcha

What rss readers for linux can you recommend?

Self-hosted Miniflux, dead-easy, just works.

https://miniflux.app/

Also my favorite, I used to host it by myself but went now for the hosted version. the only improvement I would like, would be better full site scrapping. anyway I mostly consume rss through an api (into reeder).

NewsFlash: https://gitlab.com/news-flash/news_flash_gtk

My own tool which creates a daily digest, easy to setup, just run it: https://github.com/piqoni/matcha

It will generate RSS in markdown which can be read by something like Obsidian. It supports also Terminal mode.

I looked into this. Obsidian uses a proprietary license and is not open source.

Newsboat is an excellent command line feed reader for Linux:

https://newsboat.org/ https://github.com/newsboat/newsboat

I tend to agree. I've said multiple times how much I miss google reader. HOWEVER... We need an RSS 3.0. Let's face it, parsing XML is an extremely tedious and slow task, even if you are using a fast compiled language. There are a million ways to do it now and optimize it by magnitudes. Speed and size-wise.

what rss reader can you recommend for linux?

QuiteRSS. For Android, Flym.

RSS, at the time of Google Reader, was, imho, really buffed up by the addition of Google Buzz for commenting contents with your social network. I can only speak for myself but it was really a very effective duo. (that Twitter or Facebook could barely reproduce, given their egotrip orientation).

It really bothered me when Google Reader went away. I always wondered why they did that??

I have been using RSS for the last 10+ years and have not been on social media for almost at least that long.

HN is social media.

HN has an RSS feed by the way: https://news.ycombinator.com/rss

If you want to call it that - it's more of a forum, than modern social media. I meant Facebook, Twitter, Snapchat, et. al.

Wow, that is my goal now. I'm curious, what RSS readers do use on mobile?

I still have `Press` - an old android rss reader app for android - installed and hooked to my self-hosted aggregator.

Feedly is the best thing I’ve found for iOS.

Reeder on iOS is pretty good. Been using it for years across multiple versions

I don't read RSS on mobile. Only at a desktop when I'm ready to read.. otherwise I just end up mindlessly scrolling.

NetNewsWire on iOS and Mac.

For Android, Flym. Free. No ads.

Truthfully I have never used RSS, beyond having to support it in a piece of software. I've never really understood the appeal and I've never been able to make it fit into my life.

Maybe I'm missing the point, I never read news or other such things, so perhaps there's just no use case for me.

You can use it to check for software updates in Github releases

edit: technically in Atom format but functionally it's the same, just add /releases.atom

In simplest terms, do you want to remember to visit 20 different sites each day or just remember to visit your RSS reader which aggregates all those sites?

There are other things you can use it for--including being the behind the schemes technology for podcasts. But, yeah, that was the primary not behind the scenes use case.

There's been a decline in low volume individual sites and a lot of people find sites like this one, Twitter, etc. provide sufficient discoverability for their needs in any case.

I use rss2email + opensmtpd to email me my rss feeds from a virtual machine every hour. It is setup so I can see it on my home screen but not notifications so I can read at my leisure.

Does anybody know why Randall Munroe (XKCD) has made a mockery of RSS on his site? I thought his joke was funny at first, but he has left it there and now the RSS feed from his site is worse than useless. Here's the content I get from his RSS feed on both my Android phone and my iPad:

"xkcd.com is best viewed with Netscape Navigator 4.0 or below on a Pentium 3±1 emulated in Javascript on an Apple IIGS at a screen resolution of 1024x1. Please enable your ad blockers, disable high-heat drying, and remove your device from Airplane Mode and set it to Boat Mode. For security reasons, please leave caps lock on while browsing."

This behavior began several months ago. Previously I could access his content on the same devices.

lmao

Is it still doing that? Whenever I tap the "RSS FEED" link, Safari asks to open my RSS feed. Even when I uninstall my RSS app, Safari asks to look up feed readers in the app store. I can't find anything like that quote when I view the feed's source on desktop.

Edit: Maybe that was a bug? I see that exact same quote in a tiny font at the bottom of every page.

I'm really not sure what mean. You can literally take a look at the feed and see the most recent 4 entries: https://xkcd.com/atom.xml

Works fine for me in NetNewsWire. I have never seen this behavior.

Yes, please.

Google took a number from Microsoft’s embrace, extend, extinguish playbook with its google reader.

NetNewsWire is my preferred rss since it’s rebirth a few years back.

Please, you recommend me RSS reader for Win desktop with customizable appearance? With UI that is not from early 00's.

QuiteRSS.

This is example how outdated UI looks like, which is not pleasant to use for me.

Feedbro is by far the best feed reader.

Which OSes does it run on?

Still haven't forgiven Google for killing reader. RIP.

Are Fediverse services compatible with RSS?

You can get an RSS feed of posts of any Mastodon user. Just add ".rss" to the profile url. For example, mstdn.social/@feditips -> mstdn.social/@feditips.rss

Details: https://mstdn.social/@feditips/108357998963885456

The conceptual difference between RSS and Fediverse is that RSS is read-only while federated protocols are read-write. For example, you can add comments through ActivityPub, but not through RSS.

i've been a customer of newsblur.com ever since google reader died.

i get the hacker news 100 feed there.

This should have (2018) in the title please :)

Added. Thanks!

#legend

Reply on news.ycombinator.com