RSS is back as an underpinning to SlackOps
conoroneill.net
conoroneill.net
The only thing I'm not satisfied with reading news on RSS, is that news organizations push too many articles, to the point that reading the headlines alone takes quite some time. There's nearly 100 articles per day per source sometimes. Unlike a newspaper, which has a natural structure of priority and hierarchy, in an RSS reader, every head line has the same salience, and it's a pain to weed out what's important.
I kinda hope news organizations would make a separate "weekly digest feed", 30 or so articles per week.
Unfortunately training is per feed, there's a newer Premium Pro tier with global training coming out with it, but it is much more expensive [2].
1. https://www.newsblur.com/faq (under Intelligence section)
2. https://forum.newsblur.com/t/global-keyword-training/5419/6
I assume this is targeted at professionals that are trying to stay abreast of developments around certain topics/subjects. If you think about it through that lens it's reasonably priced, I suppose.
So you can have a different feed for Science and Technology, and one for Australia, and one for New York. By not using the front page firehose, you can keep the number of inbound articles to a more manageable level.
Though like many sites, I have no idea how you would reach the "feeds" page from the front page (I gave up and searched Google for it)
[1]: https://codemadness.org/sfeed-simple-feed-parser.html
(You can see my setup at https://acdw.casa/planet)
mike@klaus:~$ rss --list|grep ycomb
155. https://news.ycombinator.com/rss
mike@klaus:~$ rss --list-grep 155
74. title = (?i)\b((e-?|web)?mail|hardenize|irc|internet relay chat|grpc|hashicorp|rust|debian|c\+\+|perl|(bit|name)coin|tor|pgp|gpg|gnupg|openpgp|digitalocean|ovh|linode|grepular|email\s*privacy\s*tester|parsemail|ssl|https|backdoor|apache|exim|distribut|peer (to|2) peer|vpn|secur|anonym|webrtc|torrent|webtorrent|nextcloud|owncloud|graphql)(ity|ous|e?s|ing?|ed?)?\b
I'd be more than happy if you give it a go and share how it seems.
We need to get back to a world where we’d have daily or weekly issues of news that we all read so that we can’t point at the same thing and say, “that’s right” or “that’s wrong”.
> […] The company spent a good portion of its presentation specifically focused on podcasts, which it said had been “largely unchanged” for years before its entry into the market, due to the limitations of RSS.
> Spotify cited how unbundling podcasts from RSS technology has paved the way for Spotify to generate revenue through these popular audio programs — a sentiment that’s not universally beloved by those who support an open podcast ecosystem. Spotify has disrupted that market by bringing some podcasts in-house, where they can only be heard on its service, and competitors have followed. This has fractured the ecosystem and left consumers at a disadvantage as some shows are no longer broadly available.
> “We’ve been able to replace RSS for on-platform distribution, which means that podcasts created on our platform are no longer held back by this outdated technology,” Maya Prohovnik, Spotify’s head of Talk, told investors.
The plan was to embrace, extend, and extinguish the podcasting medium, and the hard part of that is done. "Podcasting" was a standards-based, open medium. Now it's any audio show, and we've lost the unique name for the open medium. As Spotify incentivizes more and more creators to not publish an RSS feed, standards-based podcasting will become a fond memory.
The whole article is pretty much just re-wording Spotify's press release.
There seem to be no sources in it except for Spotify's press release.
So, a very old form of organizational disfunction in journalism: The cheapest and quickest way to write an article is to use someone's press release as the only source, as it doesn't actually require, well, any journalism.
When I recently asked friends for podcast suggestions... several were spotify-only, i don't think the suggestors even realized it, cause they just listened on spotify anyway regardless.
So I think Spotify agrees with you and is working on it...
I occasionally venture onto Reddit or HN to see if I've missed anything big but rarely do.
Facebook. Twitter. Instagram. These would be incredibly useful things to have native RSS feeds for, but instead I have to dig around and either use some tool someone built for the purpose - a tool at the mercy of the walled garden whose wall they're peeking over - or build my own nightmare factory.
It used to; killed off ~2013:
* https://brodiesnotes.blogspot.com/2013/06/twitter-has-killed...
If Facebook had any kind of API I could probably build this, and stop using facebook otherwise. Which is probably why they don't?
Imagine that shortly after the turn of the century, Adobe created a proprietary platform for delivering internet content and called it the "web". Viewers must use Adobe Web Reader to use Adobe "web pages", which are DRM-protected PDF files delivered using proprietary protocols. Adobe has done deals with several hundred popular sites to turn off their standards-based websites and deliver Adobe websites exclusively.
This is what Spotify and others are doing, and nobody cares. Apple is effectively the last media giant left holding the flag for standards-based podcasting, but how long do we think that will last?
What I just love is when people use this excuse to justify "borrowing" a clearly defined technical term, using it incorrectly, and then insisting that their incorrect usage of that term is entirely valid because "language and meaning change over time."
(Not saying you're one of those people at all but that is one of the most frequent uses of that phrase that I tend to experience from other people.)
* https://en.wikipedia.org/wiki/Podcast#Etymology
There was a minor movement to label them "netcasts", but it didn't really take.
The history was that "podcasting" was an open medium, "open" being the point. A "podcast" was a show (e.g. "Smartless"). "Episodes" were audio or video files combined with metadata that lived in the podcast's RSS. Anybody could play in that ecosystem.
Now "podcasting" means anything/nothing. "Podcasts" are shows. "Podcasts" are episodes. "Podcasts" are things that increasingly need a proprietary app to play.
The distribution model was a client subscribes to the RSS feed on their computer, downloads new episodes (from the publisher's website), and then syncs them when the "iPod" is synced to the computer. Many PMPs but the iPod especially defaulted to syncing your computer library when plugged in (to charge). So the iPod just received no episodes of podcasts when people plugged them in to charge at the end of the day.
When iTunes added explicit podcast support it became even easier to subscribe and sync podcasts. This was and remains a good distribution model but Spotify et al have done their damnedest to co-opt the term to mean content exclusive to their platform.
https://en.wikipedia.org/wiki/Podcast#Trademark_applications
Those types have done a lot of the same damage to the word "wiki". It blows my mind that Sourcehut of all places is a willing participant in debasing the term.
> Imagine you're a mathematician, and fellow mathematicians start calling derivatives integrals instead; that's basically how badly the term[] is being misused: it's being used to describe systems that are almost the _exact opposite_ of the concept.
Many of these new-gen collaboration sites that bill themselves as having wikis result in a request for approval (pull request) instead of an actual edit when you attempt to make a change. (And in the Git-backed ones, the process is usually more egregious; see below.) This proposal/review/approval cycle is exactly what a wiki isn't. GitHub no longer imposes this workflow—there's now an option at least to turn your wiki into, you know, an actual wiki—but even today that option remains off by default so all pages are closed to changes unless the owner makes an effort to toggle the right setting[1]. What these are aren't wikis—they're anti-wikis; if you have a "wiki" that only you can edit, then stop calling it a wiki. (Notably, GitHub's setting doesn't control whether it will continue to advertise it as a wiki or not—instead of just, like, the "docs" tab or something.) Most of these are static sites with extra/fewer steps.
SourceHut is even worse than present-day GitHub, because for all its docs in the man.sr.ht namespace, this is the process you have to use to edit one of its anti-wikis: spot the typo, note the title of the page you're currently on, find the corresponding repo URL in the page footer, leave your browser to clone the repo to your machine, open that directory and locate the source file based on the page title you noted earlier, edit it, make a commit, and then (probably**) push your changes to the original repo. This flies in the face of the definition[2], etymology[3], and entire spirit of the word[4][5].
* Maybe you hit a wall because the specific page has been locked. Fine. Raise your concerns through linked resource for discussions. The fact that individual pages can be locked because e.g. they have a history of being the target of abuse does not negate the meaning of "wiki". If every page is locked down (including pages that don't even exist yet, so you can't go create them), then that makes a material difference—it makes it not a wiki.
** It may not actually be the case that trying to push your changes makes them immediately live on SourceHut—it very well may result in a request for approval by the project owner. No idea. I've never gone through the process on Sourcehut because of how much contempt is deserved by projects that have contributed to the campaign of debasing the term "wiki" to it-means-what-we-feel-like and who-are-you-to-say-that's-wrong levels of uselessness.
1. <https://docs.github.com/en/communities/documenting-your-proj...>
2. <https://en.wiktionary.org/wiki/wiki#English>
3. <https://en.wiktionary.org/wiki/wikiwiki#Hawaiian>
Does Google Takeout support Google Podcasts? That'd be the usual method of exporting everything from a Google product.
Nope. One of the few use cases where I actually looked towards Google Takeout as an escape hatch for my data, instead of a mere novelty ("a ZIP of all my YouTube interactions—that's neat, I guess"), but Google Podcasts is not there. I've filed several complaints about this and several other shortcomings in response to the Google Podcasts app's spammy notifications requesting feedback.
I ended up writing a short set of procedures to follow (an SOP—written like a software manual, really) that walks you through how to select page elements that are shown on screen while logged in, and then drag and drop them from the podcasts.google.com tab into the "export tool". The punchline is that the export tool is just the SOP document itself. It's meant to be read in your browser, and it takes advantage of the fact that the browser also happens to be a universal runtime...
The exporter will descramble the links and output your subscriptions in text/uri-list format.
It could really stand to have an associated 90-second video demonstration that walks through the steps because of how unorthodox the the-documentation-is-the-implementation paradigm that I was experimenting with is.
(I actually still listen to them on an Apple iPod nano....)
I could believe it being true of some of other platforms, trying to have "our-platform-only" content but I haven't seen it and am not finding confirmation. I don't think it's true of Google Podcasts, that they have any of their own content.
I migrated my podcasts to GNOME Podcasts app by selecting the RSS link manually from Google Podcasts app when I switched to a Linux phone.
It wasn't very fun constantly having to switch between my RSS app and my browser, so I just stopped using it.
Some publishers offer full RSS feeds as one of the benefits of their subscriptions. It might be something to explore for your favorite must-read sources.
1. Widely adopted protocol, clients in all programming languages.
2. Publicly readable without password
3. Get notification for new items in many tools like Slack
4. Easy to produce and consume
5. Many nice Web UI Reader available
<Add yours>
There's still this lurking mess of RSS 0.91 vs 1.0 vs 2.0 vs Atom. As one of the folks involved in creating Atom I'm frustrated with the outcome.
These days it'd probably be JSON, right?
It does have a few minor advantages: multiple attachment support, more "branding" images support (author avatars, post "banners"), easier to build JSON than XML in most API backends these days. The one major advantage is that JSONFeed is easier to work with it in tandem with WebSub and other pubsub systems that assume messages are natively JSON, you can use the same JSONFeed item generating code for both pull (HTTP GET) and push (WebSub) scenarios.
Certainly even the main advantage isn't a huge reason to switch to JSONFeed if you've already got RSS/Atom feeds, but I believe the hope for JSONFeed was always that it would spark some sites/backends that don't want to support XML, don't have good XML libraries, or don't want to use their template languages to drive RSS/Atom feeds to be able to build something simpler instead resulting in more feeds overall than if things were just left to the XML-ish status quo. I don't know how successful it has been on that front, though, but I appreciate it exists for trying.
I’ll repurpose an older comment[1]. It was in response to “I doubt very much [JSON feeds] will catch on”.
> There’s little incentive for websites to change to JSON feeds when their RSS feeds are already implemented, working well, and generated automatically. But several good RSS readers added support for JSON feeds when the format was introduced and that’s all you need for it to be viable; it isn’t important if it “catches on” after that when the flexibility is there.
> JSON feeds were great for me because I generate my own feeds for personal consumption and can now do so with simpler code. I understand I’m in a minority—most people consume feeds and never create their own—but the larger point is JSON feeds may have already done their job: they exist and are supported if you need them.
Are there any technical reasons to use RSS instead of Atom in new deployments?
(I can perhaps see leaving legacy RSS links created pre-Atom in place.)
There's no technical reason. Unfortunately many CMSes still emit barely usable RSS v2. There's also the branding issue of RSS the format essentially being the name of the technology. Someone saying "I want an RSS feed" ends up with the implementer making an RSS v2 feed rather than a better Atom feed and just labeling it RSS.
Maybe in theory, but as a heavy RSS user this has never been an issue in practice. For example, WordPress sites support both by default — RSS at /feed/, Atom at /feed/atom/, but apps like Feedly hide this implementation plumbing during normal use.
You could also just get the unique email address of a channel and let the email notifications go there, same result: https://slack.com/help/articles/206819278-Send-emails-to-Sla...
… and using Slack to consolidate RSS turns RSS from pull to push. (Slack probably has caching mechanisms across all workspaces to reduce pull bandwith).
This was created to avoid unnecessary polling of RSS/Atom feeds:
* with `HTTP 304: not modified` result, I'll come back in 60 minutes * if there is any update, I'll come back in 4 hours * if there is no cache control header, such as etag or last-modified, I will poll every 12 hours.
I actually prefer the ability to push things like the top stories of the day to me through either Slack on email rather than having the temptation to constantly refresh an RSS feed app. An added bonus is that I frequently add some custom logic to curate the feed to my liking (based on the feed).
A lot of these online feed-readers have terrible UIs for managing large numbers of feeds, but I guess they allow the same benefit of keeping track of state (read vs. unread) that email also offers.
I'm slowly extending support for other social networks, to my surprise, they are not that "social" and it's quite a pain to scrape anything from them.
It... wasn't ideal though, in hindsight we should've hacked together a UI so that the people sending the message could confirm the message and see how many people would be receiving it first. There have been a few Incidents of accidental test messages sent out. One was my fault, early on, because production and test were the same machine. The other was the operators' fault, but by then the app had two million installations and it caused a bit of a social media storm that day (#hajo).
It was funny to see that one play out, and how quickly there are print companies making and advertising merchandise about whatever is trending on twitter that day.
Prior HN discussions: Turn GitHub into an RSS Reader (https://news.ycombinator.com/item?id=27010144)
It wasn't long before people discovered that and began using our service to deliver relevant stories to their Slack teams and stakeholders. As the author says, RSS is the closest thing to an open MVP-style API that "just works," and hasn't been plagued by a company's closed garden rules.
I am wondering, what else could one use RSS for.
Long live RSS!
Google makes ad revenue, cuts out most of the sites themselves and increases google value, while reducing the value of the actual site producing the content.
I would advise against co-mingling an RSS feed with your team's main communication or coordination channels.
Isn't Feedburner part of Google nowadays? I remember reading somewhere that they inject ads in the feeds.
Not a lot of new features being developed, and some features getting turned off "to support the product's next chapter"[1].
There hasn't been a firm announcement about Feedburner being sunsetted, but given Google's history of killing things off (including Reader, apropos of this thread's topic), it does feel like the walking dead.
https://developers.google.com/search/blog/2021/04/changes-to...
I feel like Slack has already become a black hole equal to my Inbox in all respects and am hopeful that the Interoffice Mail envelope will stage a comeback for the things I really need to see.
[1]: https://github.com/maubot/maubot#readme
[2]: https://github.com/maubot/maubot/issues/110
[3]: https://github.com/maubot/maubot/pull/147#issuecomment-99722...