The Bluesky firehose viewed in the style of a Windows XP screensaver | Hacker News Reader

The Bluesky firehose viewed in the style of a Windows XP screensaver

firehose3d.theo.io

602 pointsconsumer451149 comments

A lot of commenters here are having their minds blown by this. And while I also love this I get the sense that many others here are maybe too young to remember that this kind of open access to data used to exist for lots websites. It inspired companion sites and loads of creativity. I find it tragic really, what the internet has become. I hope federated, and even more-so p2p, protocols take significant foothold on the internet and help revive this spirit of the web. The corpo-web is so fucking boring.

It’s worth noting that twitter itself owes a lot of its popularity to its openness in the early days. In the early days there were third-party clients, RSS feeds, XMPP support, etc. You could post from a curl command in a cron job, leading to all kinds of interesting automated feeds. Then they walked it all back in the early 2010s.

I like that Bluesky’s federation model makes it harder for them to do an “open platform” bait-and-switch like Twitter did.

> I like that Bluesky’s federation model makes it harder for them to do an “open platform” bait-and-switch like Twitter did.

Why would it? They can still lock everything down and few Bluesky users will even notice. This is similar to what Twitter did, or what Google Chat did, etc. Compare this to other federation platforms where a server that locks itself down loses access to a huge chunk of the network, once the other servers reciprocate.

> Why would it?

Since migrating your personal data was a thing they thought about since day one, migrating to another network than the current one would be way easier than any centralized service and also easier than ActivityPub.

Seems there is one piece of the puzzle missing yet ("AppViews") in ATProto to be able to run completely independent, but seems they're currently working on getting that in place now.

You could still migrate all _your_ data to another service in Twitter quite easily, and most definitely you could in Google Chat. This did not change things.

> You could still migrate all _your_ data to another service in Twitter quite easily

Yeah? I don't remember being able to migrate from/to Twitter and taking followers/following etc with you without having to ask/request others to do something too.

But I'm guessing that you'll also have to request your followers to use a different AppView if Bluesky did a Twitter.

You can host your personal data, but as long as Bluesky Social, PBC is running the main Relay and AppView, and there's no easy way for Relays to talk to each other (like e.g. it works in Mastodon), they are in reality a centralized service.

Oh yes, and they are the only ones owning and developing AT Protocol, which makes it much more a currently-open-sourced protocol rather than a standard that is jointly developed by the industry.

The big central place is still the PLC directory that effectively means all accounts are centralized at BlueSky, even if your posts are not. They haven't planned to make it any decentralized in the future.

Nothing technical is preventing any federated platform to stop sharing the content with its peer. Only thing that prevents it is if there are multiple big peer and they can't afford a network partition, which is not the case with bluesky. Eventually VC's money will run dry and they don't have any solution for this.

They are a public benefit corporation that use that as a selling point but then don't disclose their charter. That seems really shady to me, but less than what twitter has become.

>In the early days there were third-party clients, RSS feeds, XMPP support, etc.

Right. This is something I keep pointing out in threads about RSS. Some people will say RSS never left. Well, it left Twitter for one. Google News and Craigslist for others.

I almost wonder, to GP's point, if people have just completely forgotten all of this, which is why they think nothing was lost.

The 2000s Internet felt way more innovative than the one we have today, despite all of the WASM, WebGPU, JIT optimizations, and other technologies that have been developed in more recent years.

We had torrents, open data, open protocols, and people were sharing data and remixing it freely. Mountains of stuff like this Bluesky demo was released every single day. We had link aggregators to point to the cool things that were happening, and we even had tools that let you pipe data sources between various APIs to enrich and recombine things easily.

Platforms stopped this. Facebook, Google, and even Apple put an end to the wildly evolutionary behavior by delivering a canned experience to the masses.

We need a return to P2P where single platform silos and their army of product managers don't shape how we interact with technology and the bulk flow of information.

I see 2 way to do this. A company (and PM) sees demand for the feature and they include it, ot it is forced by regulation.

A lot of these companies that originally had open standards formed with huge amounts of VC money and they prioritized growth over everything else. Then when they reached a certain scale, investors valued profitability and they slowly squeezed and monetized users until all of those open standards features were gone.

> The 2000s Internet felt way more innovative than the one we have today

Because it seems like this stuff is taught in Management 101 in all of the business schools: once you establish yourself with all this talk about "openness", etc. then the only way to succeed is by creating a walled garden, either through abuse of your monopoly position or by regulatory capture.

Cases in point: OpenAI _and_ Anthropic both pushing for regulation of AI, now that they have a dominant position.

I swear, the moment MBAs get involved, they try the same crap everywhere.

It's a common trope to blame MBAs for all the ills in the world.

But the reality is that having a moat and how to defend it is a fundamental strategy that every CEO is expected to know. Because it will be one of the first things you get asked from YC, investors etc.

And using regulation to lock out competitors definitely did not start with OpenAI and Anthropic.

> And using regulation to lock out competitors definitely did not start with OpenAI and Anthropic.

That's what I'm saying: it started with the MBAs.

Indeed.

Question is whether this all that was in the 90s can again be relevant for the young who grew with TikTok, insta and fb.

I don’t have the answer, only doubts…

Now with GPTs more than ever people can regain their web presence. But do they do it? I guess not as much as one would expect.

You may say we need to go p2p again and perhaps Tim Berns Lee actually meant p2p with the HTML, but are people aware of this daring need?

This is just revisionist history.

We still have all the tools you talk about today. But with the benefit of much simpler languages and SDKs and tools like LLMs to help generate code. I've seen children learn programming far faster with Swift Playground on their iPad than I had to with C++ books.

And these sort of canned experience are what helped bring technology to everyone. Which was always supposed to be the main goal.

It's a little bit of both in this case.

Kids are all over Discord, Roblox, Minecraft, and VRchat. They're writing scripts and mods, and that's great. They're probably having a blast and a good number of them are learning a lot.

But they're doing all of these things in someone else's walled garden, on buttoned up platforms that typically constrain what they can accomplish. There are fewer degrees of freedom and a lot less ownership in the work they're doing.

These platforms also cost kids money. They use toxic gotcha mechanics and peer pressure to monetize. This part is strictly worse.

It's a lot different carving out your own clubhouse and culture when you're renting. Especially when you're made to speak a certain language and abide by a rigid set of rules.

irssi had a plugin to read and post to Twitter from IRC, it was neat.

I’m struggling to recall some of those early fun accounts. Do you remember any interesting automated feeds from Twitter’s early days?

I used to run a cron job that would scrape my university's daily bulletin and post every day right after it was updated. It may have been the first “presence” the university had on twitter. I remember following things like weather stations that had automated accounts, as well.

https://techcrunch.com/2009/12/12/newlywed-sex-tweets/

Best Man Rigs Newlyweds' Bed To Tweet During Sex.

They’re on the job! #1 – Action commenced at 12.21GMT. Weight: 84KG.

They’re off the job! #1 – Action concluded at 12.24GMT. Duration: 3 m.15 s. Frenzy Index: 8 (scary). Judge’s Comment: “Is that it?”

BS is a cynical data-grab and will lock down and back-stab its users as soon as it can do so and be competitive.

I'm thinking that it's a cynical way to put all the data on an open network rather than a gated and paywalled one. As far as I can tell, Bluesky is still entirely funded by Twitter and that will eventually be cut off. They're betting on the Google/Mozilla type relationship to be maintained for now. They need to make the indexing service easier to replicate before that happens or the network will collapse when funds dry up.

They have nothing to do with Twitter anymore. Last round was funded by Blockchain Capital

https://bsky.social/about/blog/10-24-2024-series-a

Even if true, which I'm skeptical of, social media sites don't have to be permanent. It's no different from any other service being enshittified, just use it until it sucks and then stop.

As opposed to X, which told users it would be using their content to train Grok.

No matter which way you spin it, Bluesky is significantly better than X.

except on X you can actually say what you think.

https://x.com/search?q=bluesky%20banned&src=typed_query&f=to...

I can't even view that link without logging in

https://x.com/elonjet

…weird

not really. its stalking. it poses a direct threat to someones safety. and people arent even supposed to have that information under the current rules... thats why its only possible to get his jets location by crowd-sourcing it. anonymity for high profile passengers is a legitimate concern, enough for the FAA to bake it into how these flights are tracked. soon this loophole will be closed and then what will you point to? how could you seriously compare this to being banned for saying something that is politically incorrect?

you can't even say "cis" on twitter without the tweet getting hidden

Elon just posted a Tweet that if you post pictures of food you can get banned. Food. It's run at the caprice of a thin-skinned Napoleon at this point.

Tweets seem a truly terrible way to train Ai.

In my opinion : yes. But when you think about who Elon became, I think … Yeah a truly terrible way to train an AI.

A long long time ago, LiveJournal used to display all new user-uploaded images in a firehose type page.

It was a fascinating glimpse into the shared lives of people all over the world. It was definitely from a simpler time; there's no way something like that would be available now due to the violence/abuse material that gets uploaded.

I'm not too up to date on Twitter clones. What's the difference between this and Mastadon?

Mastodon has about 1 million active accounts, bluesky has 12 and you can also host your own data. Threads (by meta) has 100 million but you can't federate it yet so bluesky is the way to go currently for selfhosters and privacy advocates.

Mastodon is also confusing for a lot of people and when you point it out techies will appear to yell at you and tell you that "no in fact it isn't."

Whereas Bluesky has a familiar consumer app feeling to it.

Each of these sites also has a distinct vibe. Threads, for example, feels a bit like when Time Square turned into a kind of Disney property. It's clean and safe! But it also lacks a bit of soul.

Threads has about ~300m MAUs so far higher than 100m accounts.

And you can federate it with ActivityPub just not fully supporting it yet.

https://github.com/theosanderson/firehose

I just love that the open nature of Bsky is allowing people to hack things like this.

Now that their growth is crazy, let's hope that the work they did on the protocol and corp structure keeps it this way.

Yeah I was looking into the firehose as a potential way to source to discover new domains for my search engine. Even though it didn't pan out, I really appreciate how accessible the data is.

Why didn't it work out?

I decided agains it because it had an incredibly bad signal to noise ratio. Almost all links I saw were either to big websites like newspapers, patreon, onlyfans; or behind url shorteners.

Dunno, I may explore it further down the line, but for now the juice didn't seem worth the squeeze.

I created a website like 10 years ago called birdmine that indexed every link you or one of your followers shared on Twitter, in a Solr search engine so you could search stuff that had been curated to an extent. It was pretty cool, I think I’m the only person that ever used it though.

BSky could use a better search, or a better curation of feeds. I'm certain there is more room in the discovery play.

I might have asked this before, but did you look at adding this type of ingestion for ActivityPub?

Briefly, but I've come to learn there's a contingent of aggressively search-engine hostile people that has made a home on the fediverse. The federated nature of it makes it somewhat tricky to untangle the search engine friendly people from the hostile.

I don't need the inevitable DDOS:es and death threats you get when upsetting a clique of mentally ill people online.

You're right, I forgot about the backlash that this type of efforts got from people that don't understand the technology and just make assumptions about element visibility in the network, though maybe calling them mentally ill is a bit much.

The sorts threatening physical violence almost certainly are. But it's a minority to be sure.

A long time ago in a galaxy far, far away ... there was a twitter firehose and people loved how the open nature of twitter is allowing people to hack things ...

And then they decided to not-so-openly collaborate with government regimes who sought multiple times to implement a ministry of truth and censor everyone they didn't like. So now we have to settle for poor decision making.

Crashes like a Windows ME screensaver. Jokes aside, it's very fun to see open firehose access like this. I seem to recall that Dorsey had said that twitter limiting their api access was a mistake, hope we can keep this going.

For people experiencing crashes, Theo mentioned that this will likely work: https://firehose3d.theo.io/?discardFrac=0.6

I was hoping it was some RegEdit hack to fix driver settings.

If I understand correctly, the whole point of At-Proto being decentralized is that if bluesky were to shut down we’d still have access to this data. Someone else could create a client for posting and we’d be off to the races. But maybe not?

I absolutely love these firehose projects. So much fun. Makes me hope Bluesky gets 10x bigger soon :D.

If you add the following line just prior to the return in createTextTexture() the blurriness goes away:

  texture.anisotropy = renderer.capabilities.getMaxAnisotropy();

The perf could probably be largely solved with reusing texture objects as a pool instead of creating then destroying them as needed. I'm too lazy for that though :p.

Thanks - implemented some texture pooling (via LLM). I think the blurriness may [accidentally] help with the retro aesthetic, but I'll try to get that the anisotropy in as an optional parameter.

Yeah, managing your own memory is a good performance optimization a surprising amount of time when working with large data sets in Javascript. I've seen it used in 3D code, graph problems, etc, and so long as you keep it isolated, it's not too much of a hassle.

Bigger is not always better if it just fills up with the same type of people who fled Twitter in the first place. No one wants/needs a second Reddit.

This is cool! A long time ago I wanted to make something a little like this for my 20% project at Google/YouTube - a page that rained thumbnails of uploaded videos as soon as they became available.

Unfortunately, the idea was nixed since it had a pretty high chance of exposing ugly stuff that would otherwise have been lost in obscurity and never seen.

Do you think your idea has any more risk than sites like these:

- http://astronaut.io/

- https://obscuretube.com/

Absolutely 100% yes. Difference is these projects themselves are obscure. Opposed to an official Google branded service that will see significant publicity.

Reminds me of this old website I used to stare at for long periods, found it very relaxing: "Listen to Wikipedia" http://listen.hatnote.com/

Wikipedia Recent Changes Map - https://news.ycombinator.com/item?id=32649091 - Aug 2022 (36 comments)

Listen to Wikipedia - https://news.ycombinator.com/item?id=25560953 - Dec 2020 (34 comments)

Show HN: A Billboard-like chart for Wikipedia articles - https://news.ycombinator.com/item?id=10730695 - Dec 2015 (7 comments)

Listen to Wikipedia - https://news.ycombinator.com/item?id=9972781 - July 2015 (63 comments)

Listen to a melody made by Wikipedia article changes - https://news.ycombinator.com/item?id=8588576 - Nov 2014 (10 comments)

Listen to Wikipedia - https://news.ycombinator.com/item?id=6182576 - Aug 2013 (1 comment)

Live map of recent changes to Wikipedia articles - https://news.ycombinator.com/item?id=5693189 - May 2013 (13 comments)

Wikipedia Recent Changes (Live) Map - https://news.ycombinator.com/item?id=5687722 - May 2013 (1 comment)

Rcmap: real-time visualization of Wikipedia edits around the world - https://news.ycombinator.com/item?id=5645256 - May 2013 (3 comments)

Wow all the way back in 2013, I have visited it many times over the last 11 years :)

I'm saddened that Wikipedia has still not implemented an (AI) text to speech solution to read articles

I noticed that for messages facing the camera, ones further away from the screen occlude ones closer to the screen. I assume there's an alpha layering/rendering order error going on (assuming no order-independent transparency)?

Here's a version with that fixed https://firehose3d.theo.io/babylonjs.html (will probably replace the current one soon)

Seeing some depth-sorting issues with the text on Safari (macOS). Some distant head-on text (not on the sides of the "tunnel") is being drawn over nearer head-on text. Also, sometimes top of text is being clipped a bit.

Very cool though.

(Here's a version without that problem: https://firehose3d.theo.io/babylonjs.html)

This is SO cool I love it. Feels like I'm reading snippets from a million people's diaries one after another. Humbling!

This crashes my Safari browser on iOS.

Yes, best in Chrome I'm afraid

Firefox ran it, but very broken.

Weird, for me it worked great on mobile Firefox

"Works" on Firefox if you can stomach 300-400ms pauses every 2 seconds.

Edit: I just profiled it and it spends 42% of exclusive time in texImage2D. It would be better to allocate a set of textures up front and then use glTexSubImage2D to update their contents. glTexImage2D allocates a new texture every time.

Thanks - should be better now hopefully on Firefox

Still very stuttery before crashing in Safari on iOS on my iPad :/

Smooth and stable on iOS DB3 on iPad…

This did get rid of that bottleneck and the stutters are slightly better, but it still has some more work to go.

https://ibb.co/LgtJHMs

You'll want to get rid of glTexImage2D completely except for application startup (allocate a pool of N images up front, then re-use them and update with glTexSubImage2D). And short of being able to optimize the text render, which seems to be awfully stupid, you'll want to render offscreen to those textures ahead of time before you need to render them on-screen.

To be fair, you're crazy CPU-bound. This workload is peanuts for a modern GPU and there's no excuse for it not running at 500+ fps. But that's just how JS goes. You'd probably have better luck with C/wasm for this kind of thing if the web is your target.

For reference, while it does work much better on my old laptop now, on iOS 18.0.1 iPhone 11 Pro Max, it also crashes until I add https://firehose3d.theo.io/?discardFrac=0.6

It’s seems to work fir me ok on 18.1

According to the title, perhaps try IE6 on XP ;)

Lasts 5 seconds on iPhone 16 pro max

This is like a visualization of my anxiety. Too much coming in at once to possibly process, so higher order cognitive function just goes welp

Love it! I seriously want to use it as a background for my server.

(any chance to publish a version with a configurable speed?)

(Creator here), sure I just added something so you can play with the URL: https://firehose3d.theo.io/?speed=0.9 (but if you slow the movement down too much there will just be way too much content because it's real time)

For people on slow machines you can also reduce the number of messages with e.g. https://firehose3d.theo.io/?discardFrac=0.7

Perfect. Thanks!

Also just saw this project, which is a little night sky scene where stars are posts. No real interactivity but fun to see the playing around! https://nightsky.hctr.dev/ https://bsky.app/profile/hctr.dev/post/3lb3ywnxac22n

Whoa, Johnny Mnemonic (1995) calling.

Also, these experiments are good fun, anytime there's a plethora of data available to play with it's a good time.... but anyone else get the weird sense of having been here before? Early Twitter days lots of this kind of thing was going on too with all the tweet data. Until they weren't. When everyone at Twitter woke up and realized it wasn't sustainable financially and technically to keep open firehoses out there. And then the API limits started creeping in and never really stopped. Just saying, we've been here and it's hard to see it playing out a different way even with ATProto's sorta decentralized whatever future.

There could be some pessimism or learned hesitancy, but on the other hand perhaps we can just enjoy it while it is here? I thought the same thing about people building businesses on top of ChatGPT, yet they managed to have exits before any rug-pulls.

This is really cool, and within the first 5 seconds gave me a phrase that will be stuck in my head for a long time: arrogance is not a substitute for intelligence, Sebastian. IDK who Sebastian is, but I will be sure to let him know if ever I meet him.

https://bsky.app/profile/karabergkara.bsky.social/post/3lb5t...

I was like.. "Why do so many of the non-wall posts seem to have fire emojis in them?"

Turns out:

if (text.includes('fire-emoji') || wall > 3) { wall = -1; }

It's funny how palpable this snippet is while looking at the firehose.

https://github.com/theosanderson/firehose/blob/77225acb28985...

I like the possibilities given by the openness of bluesky

You can increase the texture resolution if you "zoom out" of your browser (ctrl+scrolldown or ctrl+- in Firefox)

Would be interesting to know how the app is deployed. I see there are some k8s yaml files: does the deployment happen manually (e.g., run kubectl commands inside the cluster)? Is there some sort of pipeline perhaps? (I don’t see any in the repo)

If you are interested if something usefull can build with atproto firehose, here you go https://www.graphtracks.com

Needs more Ellie Goulding: https://helloenjoy.com/lights/

This is trippy as hell on my 40" ultrawide. Love it.

43" 16:9... on M2Pro... it's like trying to listen to a group of your closest million friends, shroomed. It's also the first GPU-intensive "struggle" this machine has experienced (i.e. animation is occassionally choppy).

Which windows screen saver was this exactly? Is it maze?

I think they just mean feelings of early OpenGL (was that what it was?)

It appears to have elements of '3D Maze' and 'Flying Windows' (Windows 3.1).

Once upon a time, browsers had a stop button to pause animations. Now im hunting menus for that button so i can pause the firehose.

Really neat. Feels kinda surrealistic. A bit heavy on my broswer though.

Is there a way to run stuff like this as an actual screensaver (on Debian)?

The firehouse will allow the scolds and hall monitors to more easily hunt down the blasphemers and wrong thinkers.

weird take

That seems to the only purpose twitter ever served even before it was "ruined" was social witch hunts and mass mob forming.

Cool, but this nearly hard crashed my phone. I'm on a OP12 which is one of the fastest phones you can buy.

The creator is here reading the feedback, and committing code as we speak. I wonder how much HN feedback will help. In any case, this is all a fun experiment!

update: between when I posted OP and now, the site went from utter jank in FF to 90% smooth on my 7 year old ThinkPad Carbon X1 (5th gen, Intel HD 620)

Nice! This is one of the coolest comment->commit experiences that I've ever had!

For a comparison, I'm on a Pixel 8a mid-level device, and with Opera I get fairly smooth frame rates with some stuttering here and there. Maybe something else is running in the background?

$100 Helio G85 with 6G RAM runs fine. Fennex browser.

Check your OS or browser for problems.

It's a nice visualization, but it reminds me of why I avoid social media. Endless, worthless garbage spewed out into the void.

This is also social media.

It feels like Clippy is about to pop out and kick my butt. The authentic XP experience.

I would like to know how far through the firehose I have traveled

Use the force to sense how many pixels you've traveled.

Then, when you get to the exhaust port, fire the proton torpedoes.

I have no clue what "Bluesky firehose" even means.

Google to the rescue: https://docs.bsky.app/docs/advanced-guides/firehose

Bluesky == A particular Twitter alternative

Firehose == The raw live feed of all new posts from all users

Bluesky is a microblogging social network, like Twitter, or Threads.

However, Bluesky is the only one with open access to the firehose, aka all the activity. Here is a different, less aesthetically pleasing tool to see it:

https://firesky.tv

Mastodon is pretty open as well. Just won’t be the entire firehouse due to the decentralized nature of it.

I should have included Mastodon in my list, my apologies. Theoretically, it's the most copesthetic project.

There seems to be some projects that scrape(?) the most popular servers to turn it into a firehose, like this one: https://relay.fedi.buzz/

The atproto relays are crawling aggregations of independent servers. There’s no reason AP couldn’t accomplish the same if they worked on it

This pretty quickly crashed my tab, using Chrome on Android.

Works fine on Kiwi Browser on Android, although I'm using a fairly powerful ASUS phone.

I also tried it on Kiwi, using an s21 FE which is a few years old but not exactly a slouch and it crashed after about 10 seconds, same as Chrome.

Totally crashes chrome on iOS - repeatedly in strange ways

Really cool! Works smoother in Firefox than Safari

This is incredible. It's all I ever wanted.

Warning: May Cause Seizures

Not sure about your feeds but I got a lot of moping over Twitter/X.

crashed my firefox

Hack the Gibson!

Whoa!

This is fucking bonkers, bro.

Love it!

The internet, was supposed to be like this. I used to perceive it like this in its early days. And then it got enshittified.

Reply on news.ycombinator.com