RSS is undead
techcrunch.com
techcrunch.com
I started a side project that pulls in RSS feeds from various academic publishers and uses a simple regression analysis based on tiles and abstracts to recommend you papers. I've become busy with my PhD but if anyone is interested, it's on github:
By the way you can easily put that thing into a Lambda without or with Serverless. I don't know do you need any data to be stored in S3 right now but it shouldn't take long to do it. =)
Using machine learning would be for my own project only in the far future but still, data science yo. Single layer neural network with one activation function or what was the great description someone gave here for linear regression.
What might have worked is getting together with a few like-minded students and postdocs and doing it collectively; it would have lowered the individual workload, and created a social pressure to actually do it. It could be a sort of superficial version of a journal club. Machine learning is cool, but in life sciences, it's traditional to solve problems by throwing more postdocs at them.
You could perhaps do this with human effort, but mediated (and perhaps assisted) by machine. Imagine a tiny HN instance with a handful of users, where the new page is fed from RSS. Or, at a larger scale, Reddit, with subs for various disciplines, and some process for routing RSS items to the appropriate subs.
It didn't really preserve my sanity, though. I'm reading those papers in my spare time, and it got me to read many more papers than I ever did before.
Of course anything Google shuts down gets an immediate scream of pain from some people, but most everything else they've shut down has died out since then.
(The other one I still hear about is Wave. Though that one seems to have a lot of people lamenting what could have been, rather than what actually was. For Reader, the complaints are about what actually was.)
I always find these articles about RSS being dead to be pretty funny, considering I'm usually reading them via RSS.
https://tt-rss.org/ seems to be gone/down.
When Google Reader went away I looked at a few of the hosted alternatives but found tt-rss early on and never looked back.
There are clients that kind of did, but one of the cool things about Google Reader that a lot of people miss is the social feed. You could share articles you found interesting and your shares would show up on the feeds of anyone who followed you as just another RSS source. It's just an extra folder of articles from your GChat friends that you can read at your leisure.
And it's people selecting articles from their normal reading rotation, so it was refreshingly clear of clickbait. People usually shared stuff because they thought their friends would find it interesting rather than distributing content designed to "go viral."
Once the ecosystem fragmented nobody has the critical mass to replicate the social features. It was nicer than what Twitter or Facebook do because it wasn't in your face.
It happened back when "Social" was a big thing and Google was trying to push everyone onto its erstwhile Facebook competitor Google+ and it's erstwhile Twitter competitor Google Buzz.
First they killed the social feed and made the button auto-post to your Google Buzz account, but Buzz being a version of Twitter, had a lot more noise to signal so nobody used it that way.
They probably eventually killed it because they preferred having their development resources put towards more exotic social media efforts than maintaining a thing that didn't really make any money.
This long article has more of the sordid details if you're interested: https://www.buzzfeed.com/robf4/googles-lost-social-network?u...
All thanks to an ex-MS "cookie licker" (the term was used by Google people after he left the company).
Although nowadays they hide the feature in the interface, it is possible to subscribe to a feed by just pasting it in input text box when you are looking for new content.
As someone who relied on RSS for years, and did not use Google Reader, this equivalency between RSS and Google Reader annoys me greatly. Google didn't kill RSS. RSS is alive and well - almost all sites have a feed.
I used to have an RSS reader set up on my own server. For people who do not want to do that, just use an RSS software on your machine and update once a day.
Incidentally, at the time, the laws were such that if my house had been located 200 feet closer to a school than it was, I would most likely still be in prison. I got lucky there, the laws were "within 1000 feet of a school" and my house was somewhere between 1100-1200, and even though I absolutely NEVER sold anything out of my house it would not have mattered. The prosecutor went so far as to send surveyors to my property and dig up the blueprints of my house from City Hall and measure out the property lines and distances. I think I saw steam shoot out of that lady's ears when my lawyer read the surveyor's report in court. She was pissed.
Assaulted no. Did I have to fight? You bet. Inside, if you don't protect yourself and/or stand up for what's rightfully yours then you're not going to have a good time at all. At some point someone who is bigger, stronger, tougher, whatever than you _IS_ going to attempt to take all your shit. It's more of a test than anything else, to see what you will let people get away with. If you don't stand up for yourself, well, prison is full of predators especially if you're in a medium or higher security facility. This does not apply if your charges are for anything to do with hurting children or extreme violence against women -- in that case they're never going to stop coming... ever.
Word from the older guys who had been in there for a long time is that prison isn't what it once was in the 70's and 80's. Bad inmate on inmate assaults is almost always gang related, forcible rape doesn't really happen [at least in the facilities I was in and passed thru along the way] because they don't really have to do that -- there are plenty of willing participants (with exceptions for internal gang retaliation, but it would have to be something major). Another big change that reduced a lot of violence was they started giving people additional criminal charges, "outside cases" to use prison parlance, for crimes they commit inside such as stabbings, murders, serious contraband, etc. instead of just doing internal institutional discipline. This also cut way down on inmate vs. corrections officer violence -- stabbing a guard now became attempted murder of a law enforcement official -- an A felony, 20 more years to your sentence minimum.
That's not to say it's fun or peaceful or anything, it's still prison, still a city within walls segregated intensely by race and where you grew up, and there are some truly messed up people in there. It's 90% boredom 5% loneliness, and 5% small bouts of intense violence. But if you carry yourself with dignity, don't do anything stupid, and don't back down when challenged (particularly over something petty) you'll be allright.
Somebody did, The Old Reader (theoldreader.com natch) was setup to be as close to Google Reader as possible. I've been a paying user since it launched shortly after the Google Reader shut down, very happy with the service.
I've been using the oldreader (what a blessing) since Google Reader died (what a shame).
Love the google auth integration (as soon as I log in to my gmail, all my subscribtions are available, no matter what machine I am on ATM), but I never ever thought about paying for the service since it is more then complete for me as it is.
Maybe I should since I truly appreciate it and have found 0 alternatives over the years (and no, I am not going to roll my own, nor host anything, nor install apps across all my devices).
Shutting down Reader had zero effect on my RSS reading habits.
That's why I'm happy to rediscover the old reader. It is not only not a native app, it doesn't try to act like one.
Webviews are unwelcome.
I think that's what makes @dril so funny. He's incredibly skilled at walking the classification boundary of that mechanism.
The best thing about it is escaping the algorithmically curated feeds.
Every and service that I use has an RSS feed, except for Twitter. I use https://twitrss.me/ to follow users. If you don't find a feed, sometimes you just have to dig a little. You learn at which URIs the most commons CMSes presents their Atom/RSS feeds (hello /feed/).
Bingo. Plenty of my RSS feeds have ads, they just publish them as articles and say "This RSS feed is sponsored by. . ."
It's the same ad model as podcasts basically.
Other ones just give you an abstract or short blurb on your feed with a link to go back to the main site, which works just as well.
I've migrated from Google Reader to TTRSS when I was forced to make a change.
Then a few months ago TTRSS started to act up and started making bold claims like "your database doesn't respond" so I moved over to FreshRSS which was a much bigger improvement than I thought that it would be to begin with.
Runs fine using Postgres as a database.
Oh, exactly what happened to me. All kinds of strange errors, especially after upgrades. I eventually gave up and found FreshRSS. Been running (and updating) it over a year, without a single problem.
Honestly, I don't think algorithmically curated feeds are a problem. The main problem is the goal of the given algorithm which is often trying to maximize someone else's profit rather than minimize your time reading the feed...
That being said, I am always amazed at the ability of my brain to go through my feed and filter the interesting out quickly.
True that. Sadly, that is the business model for the larger part of the internet. My use of RSS is partly a way of trying to escape that.
Twitter is just a firehose, but my aggregated feeds are categorized and hand picked so I am far more likely to catch content I'm interested in on there and therefore be much more engaged with it.
(I am curious though, I may just try using feeds myself.)
This article really overstates RSS's problems, and underestimates "users," who it seems to view as epsilon semi-morons attached to credit cards. On the reader side, it's simply not that hard to control the rate of things you receive, and to quickly scan them for things you want to read. On the publisher side, it's trivial to keep track of the number of people subscribed to your feed(s). And if you want to know how long each of them spent on each article, and where their mouse cursor and eyeballs pointed that whole time, then too bad -- you are part of what's wrong with today's media ecosystem.
If you want a read count, then add pingback urls and make the protocol support it so various feed aggregators can report true numbers.
Curation should happen at the client, not the protocol. Tagging various feeds works well for me.
Maybe the opposite of why users are pushing for RSS, but without analytics content producers have reasonable justification for being hostile instead.
I'd say it absolutely is harder--on the one hand, you have an opportunity for more money. On the other hand, you have the mass of users that have proven that they don't care enough to react against irresponsible use in any meaningful way.
Simple decision in a business context.
Not all reaction is direct. It's a bit like other polluting industries; you can keep doing it right up to the point where public opinion turns on the industry as a whole, and onerous regulations get imposed. I think GDPR is a taste of this.
But yeah, more GDPR-like moves would be fantastic.
I just try to avoid screens and spend time with books in my free time. I still succumb to the infinite scroll thing with the reddit mobile app -- until just this weekend I didn't actually have the adequate lighting fixtures in every place in my home where I lay back to relax.
(Books. Not "ebooks".)
Facebook and RSS will both die before books do.
I assume that 80-90% of blogs already have RSS, just due to default configurations. Seems to me the context here is about pushing adoption and design significantly beyond that, both in terms of consumers and producers.
I can't say for certain that a pingback URL system would be acceptable, but I strongly suspect a dramatic comeback requires drastic measures.
I'm speaking from my own experience running a growing & profitable startup, Feedity - https://feedity.com, that helps with custom feeds for unstructured sources like webpages.
Our upcoming API and content explorer (a different kind of 'feed reader') have received thumbs-up from some customers who opted for early-testing.
No filtering. No comments. No proposal. No central way to find them.
It's not the absence of curation.
It's my curation.
It's easy: I like a site, I want news, I get the RSS. I delete it when it's not to my taste anymore.
The basic, deterministic, manual aspect of it is what's right with it. Not what's wrong.
Another weird idea is that RSS should be for everybody. No. It speaks to a certain kind of people, and that's fine.
Oh noes! Markdown is dead!
Oh noes! SOAP XML is dead!
Oh noes! RESTful JSON is dead!
How the fuck would a non-coding journalist know the difference?
And there's nothing factually wrong about RSS providing weaker data to ad-tech survellance. The only reason it's being pushed back onto the front burner, is because the Facebook election botch has got everyone spooked about analytics trackers, and lots of richer sources are going dark, putting RSS back on an even footing with the usual web tracker data sources, because everyone is blocking thirdparty scripts and resources.
Meanwhile, pretty sure the guy is operating as a tool for a paycheck sourced from the industry the article serves, and having qualifications only helps to support such theories, since credibility makes the opinion of a stooge all the more valuable. Maybe I should lump you in with him.
Plus they remove tracking and ads on the web page.
Should be win/win since we all hate tracking, ads and never click on them anyway and still actually want to reward sources we read, right?
Shameless(-ish) plug for my project, siftrss: https://siftrss.com/
You add a source feed, include/exclude items by keyword or regex match, and get a new filtered feed to subscribe to. Because it produces a normal RSS feed like any other, it works with all RSS readers.
It's totally free. No ads, no account, no email capture, nothing but totally disposable RSS feed filters.
Hmm, I wonder if there's a way to use it to turn a Medium feed into something useful. (By default, a medium feed is not, as it contains all comments by the author interspersed with actual stories/posts, with no obvious way for software to distinguish).
As for Medium feeds, it's all about whether or not you can find a pattern in the data to exploit. For instance, maybe links for comments contain "/c/" whereas links for posts contain "/p/" or something similar. There is, of course, no guarantee that there is anything to exploit or that what is exploitable won't change—that's the joy of dealing with third-parties! :)
I haven't looked at their implementation much, but I imagine it only works on Feedly itself whereas siftrss is meant to be totally agnostic to what reader you use. The tool also supports regex, so you can do some slightly richer matching with siftrss.
That is a neat idea, though!
Well, that's fine as long as that group of people is big enough for the sites I care about to consider it worthwhile enough to actually add an RSS feed :)
But let's not pretend it's always easy.
You gotta take in consideration the filters, the configuration, the cache, if you support rss and atom, the routing, and how you integrate it in your design.
There is no such thing as a free lunch.
One thing that could also be interesting is standardizing subscription/donation information as part of the protocol itself. So that users can easily do those things across all their sources at once (via their client), rather than going to separate websites and walking through their unique subscription forms.
All the articles published on the network are actually Atom elements (https://en.wikipedia.org/wiki/Atom_(Web_standard)), it's then super easy to import and export Atom/RSS feeds in XMPP (for example https://nl.movim.eu/?feed/pubsub.movim.eu/Movim).
On top of that Pubsub brings you a fully real-time solution (no need to pull hundreds of feeds each 15min, the articles comes to you) with roles, subscriptions, comments (also relying on Atom, see https://xmpp.org/extensions/xep-0277.html) and many other things :)
You can still access Movim using any other browsers (Chrome, Firefox) on your phone and bookmark it on your launcher with a similar result :) The Android app is basically a wrapper around a WebView.
But Movim looks really great, I should give it a go, thanks for the links!
I needed RSS enough to self-host Fever (on A Small Orange which is a POS that I need to migrate from, if anyone can recommend me a simple and user-friendly alternative).
Similarly, I've been waiting for Pinry to release version 2, but this seems to have eluded the team for a few years unfortunately.
It physically hurts when I see people purge their tweet history, because I can't imagine giving up all that history, especially when you can at least export your stuff first - which doesn't include your likes and DMs.
I don't think the latest Instagram/Facebook/Twitter brouhaha has been a vindication for RSS per se, but definitely one for self-hosting, especially as services die or get acquired left and right.
You might be cute and argue that for RSS being such a big deal, there sure aren't any overnight successes 15+ years in the making to migrate to at this point. I know Maciej Ceglowski of Pinboard fame was working on an RSS reader, but that seems to have stalled. And Digg just shut down theirs, which was really good. Fever is also shut down, but that doesn't matter much when it's self-hosted - except for the crazy fees ASO is charging people like me.
I've used SelfOSS which was fine, but didn't like some of the feeds I tried to import. It's pretty lightweight though and based on your like of Fever this seems like it might be a little too light.
I'm sure I'm going over stuff you've head of with Tiny Tiny RSS, so I won't go into that too much.
A really small shared hosting reseller with decent support I've used for years is Hole In The Wall Hosting. It isn't a big operation but it just works and does what you need it to do, SSH is extra but the plans are relatively inexpensive.
Maybe it's a failure of my imagination, but I'm having trouble coming up with anything at the protocol level that wouldn't make RSS worse as a general protocol, though it might improve at specific cases.
Leave the protocol simple, let innovation happen at the client level, and then if necessary enhance the protocol to support generally effective solutions.
"4. Adding value to the Internet lowers its value.
5. All the Internet's value grows on its edges."
Inoreader, a fantastic Google Reader clone that supports advanced filtering and following Twitter feed. https://www.inoreader.com/
HNRSS, for seeing all comments with some words. https://hnrss.org/
Nice paid service for visually scraping pages into RSS feeds https://feedity.com/
Convert summary feeds into full text feeds http://fivefilters.org/content-only/
> ...It’s an exhausting experience wading through articles from the style and food sections just to run into the latest update on troop movements in the Middle East.
> Some sites try to get around this by offering an array of RSS feeds built around keywords. Yet, stories are almost always assigned more than one keyword, and keyword selection can vary tremendously in quality across sites. Now, I see duplicate stories and still manage to miss other stories I wanted to see.
None of this is a problem with the protocol. The protocol just publishes a list of stories.
If people want a feed of the "most important" stories, you can build that. You can let people build their own personal feeds based on tags, popularity, or any set of rules imaginable, and you can de-dupe the items in that feed.
I used to subscribe to a podcast feed on The Conversations Network (no longer operating) which they let me personalize to include a mix of episodes from a variety of their podcasts. That was RSS.
RSS has nothing to do with how you determine which stories to publish on which feed.
More importantly, this is already solved by RSS readers. All the online RSS services I've tried have algorithmic feed ordering options for those who want it. This means that users who want algorithmic sorting/filtering can not only get it but also have their choice of providers.
More importantly, people like me who prefer to manually prune their sources and only subscribe to things they're actually interested in seeing can still do so. You can get a quick overview of what's trending and I can do a deep dive into what I like. Everyone wins.
Dropping the garbage styles is one of my favorite things about RSS. Using my TT-RSS via phone app is a lot speedier than the "enterprise value" provided by the native site.
This article loads 3.97MB with an adblocker. In TT-RSS, I can load the homepage and a full-page RSS article from Ars Technica for only 600kb.
I find the "pull" model of RSS much nicer to deal with than the "push" model of twitter and facebook. You simply cannot keep up with your twitter feed if you add 20 or 30 publishers that tweet every day. But those same 20 or 30 publishes in an RSS reader are easy to deal with.
Of course, the publishers dislike RSS because it messes with their metrics, discourages stickiness, and makes advertising difficult. 3 things that aren't my problem.
I currently use it for personal bookmarking and posting stories to HN.
It's based on tags, eg. http://handlr.sapico.me/Home/ByTag?Name=artificial-intellige...
Ps. Performance is slow on the main page for sorting calculated columns ( eg. The Algorithm ), that's why I linked to the newest page
If anyone knows how to performantly sort on non deterministic computed columns ( eg. the ranking algorithm for votes of posts + comments) in MS SQL server -> https://dba.stackexchange.com/questions/203331/slow-calculat... i'd be very happy :p, it's killing my "trending topic" performance...
Edit: added non deterministic
Even then?
Check out how reddit does it: https://medium.com/hacking-and-gonzo/how-reddit-ranking-algo...
The concept is that the time-based part of the score is immutable, fixed at post creation time.
That way you only have to recalculate when a score-modifying event occurs, not every time the page is refreshed.
No problem. Most had just moved URL slightly or changed format so finding them on the http site was easy enough. And while finding the new RSS URLs I often checked out the entire blogroll/webring/etc too and found new feeds to subscribe to.
For me RSS is wasn't ever dead. But it may have been left to itself in the nursing home for a few years.
However there are still a lot of issues that I see.
* discovery is still an issue - more and more people are removing link rel alternate from their content so it's becoming harder to find the feeds.
* ATOM vs RSS is still a thing even though all the feeds that I parse are actually mostly ATOM namespaced RSS feeds.
* syndication and push - Pubsubhubbub is still hard and not widely supported.
* Cross-Origin and HTTPS - for my needs I'd love to be able to fetch content directly from a web client and not have to proxy the content, however the web's sandbox model means that I need the pages to be on https and have the correct CORS headers - there's no clear incentive though for publishers to support this though.
RSS is also a prolific engine for curation, given how flexible and extendable the protocol is. The emergence of expert-run email digest frenzy should be built on foundation of RSS, where individuals can parse hundreds of sources and quickly construct a digest for easy circulation.
We finally are getting to standard formats in the digital age. JSON obviously is the leader in storing object data. RSS would be great for making multi-platform content.
As a programmer and content creator, writing something 1 time and it working everywhere is the future.
I still don't quite understand the fascination with video content: It's completely synchronous; a 10 minute video takes up 10 min of my time, with limited ability to consume more efficiently. Textual information puts the control back in my hands, with the ability to aggregate content and quickly consume. RSS has been the best way to do so while avoiding the dross that every social network has devolved into.
"We have to ruin RSS in order to save it."
No thank you. The "features" he wants to add to RSS are the very features that users are trying to get away from by switching to RSS.
FB, twitter, and the "more of the same old shit" algorithms that they try to pass as machine learning are currently failing in the market. FB is losing users because it has nothing interesting to show to them and because they have been taken hostage by paying customers that depend on idiots clicking on click bait. Twitter is desperately trying to get rid of real time because they believe it is more "interesting" to keep week old tweets on top that are somehow "more relevant".
The branding thing is a bit disingenuous; this is really about ads and tracking. Good luck with that, blocking all of it like most serious users these days and also GDPR is not going to help your cause.
RSS as a publishing mechanism is fine. The firehose needs work though. It needs reliable filtering, curation, and aggregation. That's the gap filled by hacker news and other aggregators. And having access to the firehose is nice sometimes.
A agree that it never really died, it's just hiding.
All news sources I can think of implement prioritisation.
Do users really want personalisation? Do you remember your Dad saying pre-web "oh, I just want a newspaper personalised to me, so I don't read outside of my core interests" or did he just sometimes read, sometimes skip less immediately-relevant articles? I'm not sure anyone really wants to live in a bubble, just some tech companies decided they should.
I prefer local news from my neighborhood over news from a neighborhood in another city in another state. I prefer technically detailed articles about software development over technically detailed articles about dog grooming. I choose feeds and sources that filter in my favor.
Your dad always had the option of subscribing to the Wall Street Journal or People if their local paper didn’t cover business or celebrity gossip extensively enough.
Maybe not bubbles all the way down, but it’s close.
I'm definitely in favor of personally choosing what to subscribe to, but algorithms are very far from where I would trust them to make those choices for me.
* A feed for all updates to a news story for a given news source
* A feed for all updates and developments for a news story across all news sources.
* A feed for all updates for a given area for a given news source. More specific than 'news category', more general than one specific thread of events.
* Feed for all updates for a given area (as above) across all news sources.
* [Client] Functionality to subscribe to related or more specific feeds from a given feed.
* Functionality to see related commentary for a given item from given forums (for each story I'm interested in I will normally look at HN or reddit comments).
I'm sure all these exist, but I haven't dug deep enough to find them yet.
RSS is the javascript of text.
Unfortunately it became unmaintained and is now abandoned by its author. It's been on my todo list for a while to write similar so I can have the same experience again.
Algorithm-generated feeds are mostly critisized for buidling filter bubbles and leading to dependency, but either can be avoided by simply switching to RSS. It is you who choose what to subscribe to in RSS, and it’s hard to imagine many people would offense themselves by subscribing to sources contrasting their own ideologies. And any experienced RSS user will understand the stress of out-of-control unread counts, which easily leads to the crumbling of your RSS-based reading system.
You may be like, “well, I’m nimble at selecting non-biased sources and spend a health time on my RSS feed.” Congrats, but one as rational and restrained as you is also unlikely to be solicited or tricked by News Feed. On the other way, you can’t expect one inclined to be biased and addicted to crappy stories to become less so just by using RSS instead.
And that’s without mentioning the high technical barrier of RSS that filters ordinary users away. For many popular content providers, there’s no full text output, even no conspicuous feed address. Shall I suggest my mom switching to RSS and subscribing to her favorite WeChat official accounts by setting up an Linux server running web crawler? I haven’t be nerdy enough to do that.
I do love the simplicity and transparency of RSS but don’t really think something invented in the 1990s can solve the problem of 2010s and decades to come. When the density of information available reaches a certain bar, you have to turn to something like algorithm. Being arbitrary and black-boxed is not inherently a problem. Haven’t we been happy with the works of DJs, curators, and food writers? What can be more arbitrary and black-boxed than we human? What needs be changed the most is not the tool people use to get information, but the mindset with which people live with the torrent of information in such a post-scarcity era of contents.
Do not follow newspapers through RSS. That's the golden advice. Most often they have daily or weekly digest newsletters (NY Times and Guardian has excellent ones), use them instead. More generically, use periodical digests for high-volume news sources, and RSS for following websites that post anything less frequent than a couple or so times a day.
All of the things you mentioned there, is basically a reason why I started Feedpresso. After GReader was shut down, and after I got pissed with unread counts from a local newspaper on Feedly, I've put some basic probabilistic machine learning together to "avoid a mess" (what a strange thing to do - to avoid a mess, you start an even bigger one :-D).
One thing led to another, and I have to say Feedpresso is a pretty decent reader that is easy to use (my dad uses it without knowing what's RSS) compared to the most of the RSS readers available.
My feeling is that this problem about content and updates boils down to a mix of a few things: - following local and international news - following writers that produce high-quality industry-specific content - how do I get that conveniently in one place? - how should I not lose sanity and hours of time doing that?
And you are right about the amount of the content. It extremely magnifies the mess.
Also, let's not forget the whole hell with ad-incentives that make publishers produce lots of low-quality content (with ad-bloat) just to get a few extra page views (impressions). I would even argue that "free content" and ad-incentives is what ruined the regular news, and it is what drives most of the stupid viral content on FB.
Anyway, give Feedpresso a go - maybe it will stick. PM me for a discount at tadas@feedpresso.com .
Well it's a protocol. Just as we have better e-mail clients than we used to, we can have better ways of dealing with feeds. Part of the push to re-establish RSS is the hope of more services exposing more and better RSS feeds to be consumed.
This is a feature, not a bug.
> Analytics increases revenues from advertising, and that means it is critical for companies to have those trackers in place if they want a chance to make it in the competitive media environment.
If you can't survive without tracking my behavior, then I guess you will die. Given the quality of writing I can expect from this "competitive media environment", it won't be a great loss.
> RSS also offers very few opportunities for branding content effectively.
This is a feature, not a bug.
> They need to actively guide users to find where the best content is...
Let me stop you here. An RSS directory is nothing challenging. You can make one by hosting an old timey "list of cool sites" .html on your own website. Brent Simmons included one when he developed and maintained NetNewsWire before selling to Black Pixel. Modern day Podcast clients often include a directory or access to a 3rd party directory. Anybody can create a directory, stuff it with all the criteria they need to sort by tags or algorithmically push users to the highest bidder and ship it in an RSS client.
RSS is a good technology. In terms familiar to a newspaper or press corporation, it is a wire service, except one that goes from newspaper to reader rather than from the Associated Press (or whoever) to newspapers to readers. It does one thing, and it does it well, and doesn't support anything less than it needs to do it.
RSS clients range the gamut from inboxes (NetNewsWire) to news rivers (River5) to machine-curated digital newspapers (Apple News), each of them solid form factors and useful for different types of feeds, and the only reason that variance of form factors is even possible is right there in the name: Really Simple Syndication.
Dave Winer was right to freeze the spec, it's not perfect, but it is good enough. Letting newer iterations filter through would lead us to a situation like with the web, where the www went from a really great networked document browser to an application container. That's why when I visit a site like say, techcrunch.com in a browser with scripting disabled, almost nothing remains of the actual website, although surprisingly the entire article was fully readable, and perhaps more so than if I had enabled JavaScript to let all that Branding™ shine through.
Let me be clear, the only parts of RSS that died are the parts that Google and a few social corporations within 50 miles of each other owned because they all had a stake in getting you into their own walled gardens and keeping you there. The Open Web is great, open standards are great, until you need to convert eyeballs into dollars and the exchange rate isn't looking good because you're too open and didn't introduce enough artificial scarcity into your business model.
RSS never died, by and large the feeds that you subscribed to a decade ago in Google Reader are still alive and kicking, ever present and always underappreciated even at the height of Google Reader's popularity before Google decided to rebrand it to death.
> This is a feature, not a bug.
I know where you're coming from, but .. I'm one of the rare people who, for a while, read a significant number of blogs without RSS because I actually liked having the different distinctive visual styles to remind me who I was reading. It's a pity we can't have a little bit of styling without going bananas.
Most RSS and read-it-later applications do a really poor job of playing embedded videos never mind displaying interactive infographics or other JavaScript applications, so going to the source is really the only alternative at the moment.
I hope apps like Pocket would get better at this. Pocket already plays Youtube videos just fine, it's the embedded stuff that doesn't work.
More often than not, when I read something, I care about who the other is, and secondary to that, I care about where it is hosted. For more than ten years I've used NetNewsWire, and now I also use Apple News and Overcast. NetNewsWire will show me the author name just fine so long as the feed includes it, and given the left to right style of the client, I can tell at a glance where I am. For most feeds though, it isn't really a problem as I have found that the NetNewsWire-style really only works for feeds that post relatively infrequently, not more than a few times a day if that, and even that would be overwhelming coming from one source if every source I had posted a few times a day to keep up the appearances of their Personal Brand™.
I think an optional CSS or rather a subset of CSS component would not be a bad thing. Just enough to style some text and formatting but not enough to screw with the layout too much. The client could completely dismiss the attached CSS and in my case I would almost certainly end up turning off the custom CSS, but I can appreciate good web design and will visit a site if I happen to really like their design. That said, I also fear that even that concession would open the dam to a lot more concessions than I would like, so I can live with the status quo.
A good feed reader is going to have a good reading experience already, and I care a lot more about having a good reading experience than I do about the opinion of a web designer and their corporate overlords. RSS, reader mode in Firefox and occasionally lynx -dump or when I can find them, print pages, are how I can even use the web to read anything on it anymore.
So HTTP+HTML5 is a mess, I would rather not see HTTP+RSS (or Atom or JSON Feed or whatever) turn into a similar mess so that websites can rebrand their feeds to death in the same way they rebranded their websites to death.
What a stupid thing to say... All this indicates is that RSS users didn't like those apps. I use elfeed, which has upwards of 500 stars on Github, and it's an Emacs feed reader. Many other feedreaders exist that you can run on your computer. Furthermore, Feedly and NewsBlur, along with other services, seem to be quite lively. My rather unknown blog gets some requests referred from the Flipboard app regularly.
But a fact is that lots of incompetent site makers do not add RSS feeds to the websites they build. My uni's RSS feeds are going offline one by one as they redesign the web pages, and I have to maintain a scraper script to follow the updates. Many institutions' feeds are badly curated, e.g. a museum I follow, I figured out today that I wasn't getting any news because their RSS feed had a <!doctype html> and some conditional comments before the actual feed. Some feeds include javascript and XML readers fail on them. And then there's the Wix or Squarespace or whatnot websites, and they are a total failure (especially because they can't do this RSS thing whereas even Tumblr does it easily).
I really believe we should somehow certificate programmers and some certain specialisations. That means ensuring that a "web developer" does know about the base specifications on which the web is built, and the most common best practices. If a web dev fails at making an RSS feed, that's a big problem.
As a non-developer, as an observer, if people want to call themselves “software engineers”, and I think it’s a fair title: you are designing and building systems people rely on, then you need to take the first steps toward self-regulation before it’s imposed upon you.
Yes, if you want to call yourself an engineer then you ought be able to prove under examination that you understand first principles. What those first principles are ought be determined by the industry itself in collaboration with a peak-body for oversight.
What we have not is practically the Wild Wild West that is actively doing harm in some areas.
Since the recent post saying we should to back to using RSS I've given NewsBlur a go (I'd tried Feedly when GReader died, didn't like the way it scrolls). Honestly though I looked at a few articles on the day I signed up but I've not checked back once yet. I'm like the person that pays for the gym then never goes when it comes to RSS now apparently.
I don't think I've ever had a problem with a site not having RSS feeds though, any site with regular articles that I want to follow has had one
[1] https://github.com/cadadr/configuration/blob/master/emacs.d/...
[2] https://github.com/cadadr/configuration/blob/master/emacs.d/...
I bet most of those older sites ended up getting RSS feeds by default as part of whatever Content Management System they had. And since it wasn't a feature actually requested at the time of deployment, the users of the CMS didn't miss it during a redesign. And since no one has been asking for it, and well-known tools like Google Reader are defunct, services like SquareSpace and Wix don't have it, and their customers are not web-savvy, and so aren't asking for it.
Hey, I wish just as much that more people were more competent at whatever it is they're doing, but saying that leaving out edge features that weren't requested is somehow the sign of a bad programmer or a bad engineer is plain wrong. Good engineers work to solve the problem posed, not their own fantasy work; they may try to convince that something is needed or a nice to have, or may refuse to work on a job they don't think is taking the right things into account. In the case of the former, stats most likely indicate that more traffic comes from AMP or twitter than it does from RSS feeds. And in the case of the latter it just means the customer gets someone else to do it.
RSS wasn't a specification on which the web was built. It's a useful addition, but the fact that the WWW existed before it, and continues to exist after RSS's heyday, means it's not considered critical.
RSS is as much of an edge feature as the silly animations and crappy one page app reditions of what essentially is a glorified blog is an edge feature. But the latter gets you paid more. On the flip side, e.g. at my uni which switched from wordpress sites to a home-grown CMS, departments have a hard time publishing updates now, and people are switching to Facebook or Whatsapp groups instead. I'm lucky that my department chose to use a google group instead, so that I don't have to convince someone to forward FB updates to me.
Just like RSS, JS was nothing else than a useful addition, and WWW existed long before it and its heyday. The current culture of web development is directly user-hostile in many ways, and builds its confidence upon the perceived approval of users, whereas the reality is that that users' approval is just people putting up with what they have at hand with the varying degrees of knowledge they have. I can't keep mailing every website maintainer and ask for a proper website, I just make do. Luckily I can write a script and put it on cron to fetch most important updates for me, but lots of others don't have anything else than word of mouth or facebook, which is just the digital version of the former. The rapport between the web dev and the user is that of submissive bondage, not really a feedback loop and joy.
Yep, I had no idea I was using it until I started a website years ago and saw RSS was everywhere.
Maybe every actor thinks RSS is a problem, but I as a consumer of content love it for all the reasons they find it problematic.
It seems like the author is conflating what's bad for their industry and their marketing dept. and what's bad for the consumer. Consumers don't want intrusive advertising, we don't want analytics, we don't want social interaction. We want content. Full stop.
There's a healthy and lively debate to be had about what needs to happen to allow news to work for both the people who make it and those who consume it, but currently, every discussion along those lines that I read like this are always written solely from the point of the publisher. No one is asking the important questions, like:
1) Did anyone ask the consumer why the "pivot to video" strategy failed? I can tell you in one sentence: No one on the consumer end wanted it.
2) Were consumers ever asked if we wanted our data shared with all manner of companies and possible bad actors? We answered anyway by ramping up adblock usage and noscript usage.
All of these articles read so impressively tone-deaf to the wants and desires of the people they allege to be serving, and then they wonder in thinkpieces why news as an industry is dying.
"Feeder"[1] looks like exactly what I want, but the requested permissions list is insane. Are there any that don't request to "Access [my] data on all websites"? I've tried a bunch and they all seem to - I guess so they can detect RSS links in web pages?
I know Firefox has the built-in Live Bookmarks thing, but it doesn't give a notification when there's a new post to read.
Now they can't handle RSS feeds natively so they need to access webpages to be able to register RSS links. Guess it shows how sacrificing usability for security doesn't necessarily lead to more security.
I'm currently using Brief, and have been using it for quite some time. Still haven't found an alternative I liked. And I guess with the live-bookmarks dead it's harder than ever to switch. I might give Feeder a try.
Not affiliated, just a user.
One of the solutions I saw on a local traffic alert site was that you could create a personalized feed URL which would only send you a notification when e.g. here is an accident on that specific road.
This might be useful to solve the prioritization problem if you could create a custom feed based on your interests. Secondly this also provides the newspaper with some sort of analytics to incorporate "relevant ads" in the feed (I know I'm playing the devil's advocate here)
Also, we've never stopped growing and are approaching profitability on one full time employee so don't believe the RSS-is-dead hype.
> I would gladly pay money for an Amazon Prime-like subscription where I can get unlimited text-only feeds from a bunch of a major news sources at a reasonable price. It would also allow me to get my privacy back to boot.
Facebook had a scandal over selling private data, so move to RSS for the privacy! RSS doesn't allow tracking (from Amazon), let's turn it into HTML.
RSS is not a business model. Selling things people actually want is a business model.
If you can't persuade someone to pay for your content (perhaps via willingly watching adverts), then it's not economically viable content. If you care enough to continue publishing it, great, but it's out of your own pocket.
"useless content". FTFY.
Who is using RSS? How are they using it? What depends on or is enabled by RSS? What are the feedback loops, and what are the trends?
There is difference between podcasts, news sites publishing a headline-only, twitter-like feed and a blog publishing full text essays. The protocol itself is a protocol. The interaction between consumer, publisher and protocol is what's interesting. That's a very wide field of possibilities.
Fwiw, in the context of a FB alternative, id be looking out for trends that blur the lines between consumer and publisher. Its certainly not sufficient, but I imagine that any open (or not) alternative to social networks will have this quality.
[0] https://feed.press/blog/2017/05/31/feedpress-adds-automatic-...
[1] https://blog.tryca.st/cast-update-experimental-json-feed-sup...
[2] https://feedbin.com/blog/2017/05/22/feedbin-supports-json-fe...
Three steps:
1. Make sure your browser has the feed detector enabled. It used to be standard, now you need to go into settings and configure it.
2. Surf normally.
3. When you see interesting content and the Feed icon goes active, click on it. Now you've added a feed.
Eventually you should categorize and prune your feeds, but this is a good start.
And one for iOS and Android?
I can only find frontends to online services like Feedly.
Unfortunately, as EFF points out in their recent series about encrypted texting, a one-size-fits-all solution is difficult.
I subscribe to hundreds of feeds, and none of them are news publishers. Still works for me.
InoReader makes a great assistant in this way. I'm not affiliated at all.
I hate some new media who provides App only. App can increase DAU and maybe better for their KPI, but will lose serious reader.
RSS hasn't gone or been anywhere.
Qutebrowser is a very interesting browser. It's easily scriptable using a config file in python, but not only. It has a nice Emacs Customize like UI, and then it has a system reminiscent of CGI for running external scripts against web pages or on events. A definite improvement for me over whatever is available. I maintain a handful of websites with js enabled (HN, reddit, youtube, a couple more), and for the web apps I'm forced to use, I just start up Chromium temporarily. It's been lovely so far.
* facebook.com * block
* fonts.googleapis.com * allow
* fonts.googleapis.com cookie block
* fonts.googleapis.com frame allow
* gstatic.com cookie block
* gstatic.com frame allow
* gstatic.com media allow
* gstatic.com other allow
* gstatic.com script allow
* gstatic.com xhr allow
* twitter.com * block