Lynx: Open Source Native Cross Platform framework used in TikTok | Hacker News Reader

Lynx: Open Source Native Cross Platform framework used in TikTok

lynxjs.org

249 pointstheanirudh93 comments

Are we intentionally reusing project names now? Lynx [1] is a fairly popular tui browser. I guess it doesn’t support ssl or tls, so is now obsolete? W3M [2] is a decent tui alternative that does. Or if you use kitty, awrit [3] is nice, with images and mouse support (it renders chromium in a terminal window)

[1]https://en.m.wikipedia.org/wiki/Lynx_(web_browser)

[2] https://w3m.sourceforge.net/

[3] https://github.com/chase/awrit

Lynx absolutely supports TLS (2.8 and later, possibly earlier).

> Lynx absolutely supports TLS

And has done (in various forms) since at least 2001!

Project name reuse has been a thing since the dawn of projects.

I can't tell if this is satire or a hidden advertisement for various tui browsers? A project that is a "popular tui browser" (for the literally dozens of people that use tui browsers?) does not have ownership claim to the name of a big cat genus which has 4k+ other results on GitHub with the same name.

Lynx is a well-known project which has been around for far longer than Github even existed - since 1992, in fact - which is in any case irrelevant, since it's not developed on GitHub: the commits for ThomasDickey/lynx-snapshots are snapshots of the code from the website proper.

Removed the star count as any sort of “evidence” to popularity - the point still stands though. It feels absurd to claim a name being “reused,” or implied stolen, when the name is a generic animal name.

Bikeshedding about names, article titles and tangential annoyances is HN’s favorite pastime, and an easy way to get karma. As old as the hills.

I believe it's spelled as "pastime" :)

^art

i am not affiliated with any of the projects, its just what i use.

It's not even the only TUI browser pronounced that way. See: links

https://en.wikipedia.org/wiki/Links_(web_browser)

Lynx pre-dates Mosaic... "As of 2025, it is the oldest web browser still being maintained"

https://en.wikipedia.org/wiki/Lynx_(web_browser)

It's not even a good name for a cross platform tiktok network.

I would have stolen a name like Transmission, or Bing instead.

The Lynx project was originally named independently without thinking this far ahead. Since so much code and so many users already rely on it, we decided to stick with the name rather than change it just for open-sourcing.

Open-sourcing is/was the perfect time to change it. Otherwise it might be even more painful if a name change is needed later.

And why exactly would a name change be needed later?

Names are just names, many people have the same names, and projects can too.

Generally agreed. I think unique names for projects are nice, but unless the projects have very similar goals, I think having the same name isn't really a big deal.

Plenty of people have the same name, but if you call your child Brad Pitt, it will be interpreted as a reference to the famous guy bearing that name, and nobody will believe it is a coincidence.

Lynx doesn't have a large user base (I think) but it is installed by default on many linux distros. Having to install two programs with the same name is a pain which is only resolved by renaming one of them (at the distro or the user level).

Bit misleading. The analogy is closer to naming your child "Genevieve" when another Genevieve exists in the school. Lynx is a fairly common and well understood word.

> Lynx doesn't have a large user base (I think) but it is installed by default on many linux distros. Having to install two programs with the same name is a pain which is only resolved by renaming one of them (at the distro or the user level).

This is a fault of the distros. At some point keeping niche software will cause issues and conflicts.

> This is a fault of the distros. At some point keeping niche software will cause issues and conflicts.

I am not sure I agree with this argument. This gives a vibe of "make place for me, away with the old guard!"

What if someone called their program "vi" with the argument that noone uses vi anymore?

Besides, who decides what is niche and what isn't? Is a program like lynx which offers better accessibility features than mainstream browsers not worth distributing because it's niche?

Blaming the distros for already having software named like what you decide to call yours isn't terribly cooperative.

Figurative "you" of course, not meaning OP here.

You took my argument and basically butchered the main point.

Everyone and their brother knows what "vi" is. Many (presumably millions) use it.

How many people would know what Lynx is. My guess is very few.

> Is a program like lynx which offers better accessibility features than mainstream browsers not worth distributing because it's niche?

Its trivial to understand why this is a bad argument (appeal to emotion)

> How many people would know what Lynx is. My guess is very few.

And mine is that it's more than you think. Especially when compared to the number of people who know what vi is. Neither of us have figures to prove our points. My indirect argument was that the fact that lynx is included by default hints that I am not entirely wrong. Your response to that is essentially that distros packagers don't know what they are doing. I won't get in a debate on the competency of people and accept this opinion as yours.

> You took my argument and basically butchered the main point.

This was absolutely not my intention. If I misunderstood your point, please correct me and tell me how I should have read it.

> Its trivial to understand why this is a bad argument (appeal to emotion)

No. The argument for including software that have specific accessibility features is not to appeal to your emotions. The reason for having accessible software is that, as niche as it may be it is useful. No one should care how non-disabled users feel about this, and certainly no one should care whether you or I think this is too niche.

Every Go developer knows how awful a bad name can be. Just try to google anything Go related, you need to type "golang" instead.

A go developer doesn’t Google go, they go to pkg.go.dev to get go’ing.

SEO.

Try running NextJS on your NeXT

$ apt-cache dump | grep '^Package' | awk '{print $2}' | wc -l && apt-cache dump | grep '^Package' | awk '{print $2}' | uniq | wc -l

152007

You can call it lynx all day long, but it won't be lynx in the Ubuntu repositories as that name is taken, and as you can see above, there are no duplicates.

Since it is a library, it won't be named lynx anyway even if there is no name collision. D3 is packaged as libjs-d3, for example.

Yeah that's true.

Oh no!

Anyway...

>Are we intentionally reusing project names now? Lynx [1] is a fairly popular tui browser.

The kids who built this are probably younger than the Lynx project and likely don't know it exists.

Indeed, they should rename it to Jaguar.

Just tried to browse the web on my Atari Lynx.

>Please don't complain about tangential annoyances—e.g. article or website formats, name collisions, or back-button breakage. They're too common to be interesting.

Name collisions are so common that the guidelines tell us not to talk about them. Why would you assume that this is intentional? And then the rest of your post is just about different tui browsers? How is this the top comment?

oops, sorry!

Likewise sorry for writing what was on re-reading a rude comment.

I remember listening to a Syntax FM[0] with Zack Jackson from ByteDance, which mentioned this framework briefly. Some quotes and notes from that part of the podcast:

- All ByteDance products, even native apps, are web-based

- They have an in-house framework called "Lynx" which is essentially their version of React Native[1]

- "All apps are Lynx apps. Everything is a Lynx app. It's all backed off the same stack."

- This approach allows them to maintain a unified architecture while having specialized teams focus on different aspects (algorithms, compiler, kernel, etc.)

[0]: https://syntax.fm/show/860/module-federation-microfrontends-...

[1]: The one being released, in the podcast they confirmed they would be open-sourcing it this year

I work at ByteDance (though not for long). Most of the "sexy" part of the app is native, or C++ crossplatform for things that would normally be in C++. I'm not sure I'm allowed to share what Lynx is used on, but it's a small subset of the app.

I went to watch the podcast – he's heavily overplaying how much cross platform web dev is used on the native platforms. He doesn't seem to be in the biggest channels for both iOS and Android developers (the largest 'IOS'/'Android' release/global channels).

Thanks. What do you mean with "native" (vs. C++)?

What I assumed from the podcast is that there's a lot of internal reusable tools in C++, and the web-based stuff is mostly about the UI layer.

Is there a lot of WASM to use both at the same time?

Native as in Swift/Objective C (iOS) or Kotlin/Java (Android). Probably 99% of the UI (and lower layers, outside of specialty cases!) is based in that. C++ is used for typical high performance cases; e.g. Video rendering (though there's more than that).

There's a lot of internal tools. I don't know what they're written in. I would not be surprised if it's Lynx, but I'm a mobile developer on the TikTok app, and haven't ever looked into it.

I haven't come across any WASM in my time here, but I'm not in a position to declare that it doesn't exist.

Makes sense, the announcement blog said it was only specific screens, such as search.

Apple and Google need to talk. They need to collaborate on something truly native to put an end to these bloated frameworks. My goodness the amount of engineering work being put into terrible mobile experiences is worse than how much Nasa is spending to look for water in space.

Any mobile app designers out there think TikTok has good UX? I mean scrolling video is great, but everything else?

> Any mobile app designers out there think TikTok has good UX? I mean scrolling video is great, but everything else?

You wouldn't ask a doctor if she thinks crack is great, right?

Silly nitpick but wouldn't this be more like asking a crack dealer if this particular crack is great? No matter what their answer, the fact there is a crack epidemic tells you the real answer. People aren't always wanting top shelf crack, some times the hood crack is the good crack.

Google already did the talking for both of them. It’s called Flutter.

Does Google use Flutter for building Gmail, Google or other major mobile apps?

Yes they do. Google Pay or Google Earth are using Flutter. They also contribute to Kotlin Multiplatform and use it for Google Workspace (gmail, drive, etc)

Flutter and Kotlin Multiplatform are two different frameworks. Flutter is mainly for cross platform UI. KM is mainly for common-core (or common business logic). My point still stands that these are all bloated and ugly.

Google products are an absolute mess. They create android, KM, and Flutter for mobile development. They also created MLKit, MediaPipe, and TensorFlow for AI. They're all over the place with their solutions

Look for the headline "Flutter at Google"

https://flutter.dev/showcase

Ah, yes, the framework from the dominant browser vendor, that happens to eschew everything good about the browser, to achieve marginally-improved typography, by rendering the whole site in a canvas tag. That should certainly be our #1 model for cross-platform development.

I don’t advocate for using web target, Flutter is much nicer than whatever clusterfuck Apple and Google call native toolkits.

"We should repeat the same code in multiple languages" is a pretty weird thing for professional pride. Try to be a better engineer instead of promoting gate keepers' platforms and give up "quality" as their responsibility.

TikTok is impressively multi-device. It works well on Android, iPhone, Android TV, and web.

Compared to Instagram, where the web version has always been behind the mobile version, TikTok really seems to make each device version the best possible.

I'm pretty sure for Instagram it's by design, not because of their technical stack.

Instagram web is so bad that it’s the only logical explanation.

I have worked in the Instagram ios source. It is FUCKING AWFUL.

They’re impressive in their business success given their code quality.

There isn't even a real iPad app (╯°□°)╯︵ ┻━┻

Worth considering that this is an explicit business decision (not enough iPad users to justify the development effort), versus a technical one. There were various builds with proper iPad layout floating around internally, but no one ever got the management buy-in to polish it enough to ship.

I don’t think I’ve ever worked anywhere that had decent code quality

They want you to use the app because it's harder to adblock in apps than on the web.

Got curious enough to wanting to give it a go, so on the quickstart page it says:

> Lynx Explorer is a sandbox for trying out Lynx quickly.

And then it asks me to use either the iOS Simulator or the Android Simulator, which based on experience, neither are made for anything resembling "quickly".

Anyone know if there are any "pure-web" instructions around? Skimmed around the docs, website and repository but didn't find anything that looked like it was made for just web setup.

> iOS Simulator or the Android Simulator, which based on experience, neither are made for anything resembling "quickly".

It's not adb / avd or some device emulator. It's an app that you install on your device, and then it can load your app from your development device using a link. I was able to run it in 5 minutes without having android studio or any other android development kits on my laptop.

Thats neat. It makes me think of expo [0] that does it in the same way, pretty useful to try stuff quickly, but forces to go though their servers.

[0] https://expo.dev/

I recently looked closely at cross platform mobile mobile development.

Despite distrusting Google and despite knowing react I chose flutter.

I want something fast with close to total cross platform compatibility.

Net Maui is not well spoken of. React native seems slow.

The only real choice fir my needs is flutter.

Let’s see how long it holds up.

Try ReactLynx first.

I see there's Lynx for Web[0]. Does that also adhere to the two-threads model (one for UI, one for background)? Also, is this something that React/Native may adopt itself one day? Curious to know if this is a technique that may be adopted by the React ecoystem itself, or if it's just too far outside the bounds of what React is supposed to handle.

[0] https://github.com/lynx-family/lynx-stack/tree/main/packages...

Thanks for finding that, was out after the web portion myself (https://news.ycombinator.com/item?id=43269966)

As far as I can tell, it has the same model, at least judging by the directory names (web-worker-runtime, web-worker-rpc, web-mainthread-apis, etc).

I'm also a little confused about how to get started. According to the docs you have to embed it into an existing app[0], but the quickstart suggests otherwise[1].

[0] https://lynxjs.org/guide/start/integrate-with-existing-apps....

[1] https://lynxjs.org/guide/start/quick-start.html#quick-start

Seems interesting. Coming from the react native world, competition is good! This seems to have a from scratch layout engine, which is pretty exciting. It looks like it supports some of the more recent/advanced CSS features. It also has a javascript engine based on QuickJS.

I can't wait for a more technical deep dive into how this works and compares to react native.

I see very little native on that.

Native now means mobile only? That’s a depressing state of affairs. Cross platform and then not even mentioning you’re limited to 2 out of the 5 more popular OSs. Just say you rewrote React Native if that’s all.

Also supports the web, and it looks like they might intend to support more platforms.

`readonly platform: 'Android' | 'iOS' | 'macOS' | 'pc' | 'headless';`

> Not only is the core engine of Lynx framework-agnostic, but it's also agnostic to host platforms and rendering backends ... expand to even more platforms, such as Desktop, TV, or IoT devices.

> However, Lynx isn't limited to React

I think this could be huge! I'd love to see a SolidJS version

the code examples etc though is 100% react at this point. for me I would like to see a vue.js version.

Quickly tested it out, and gotta say that I enjoy being able to write just normal CSS to style things. The getting started process is really speedy.

https://lynxjs.org/ doesn't scroll smoothly.

What browser/platform?

Flutter apps don't scroll at 30/60fps on my aging Mate 20 Pro, but that website is fine for me in Chrome.

Chrome on a 2019 Intel Mac.

To be completely honest there is no reason for anything 2D on the web to run badly on any machine over 200 bucks released in the last 15 years.

Interesting if Lynx uses host OS widgets like RN does or renders stuff itself like Flutter.

Support for CSS transitions (and animations) on a native UI thread would be huge. React Native Renanimated is cumbersome and painful to use everywhere.

Sadly I finally ditched my MacBook for a Linux PC since Expo/EAS has liberated me from needing Xcode. I won’t be able to try this out for iOS development.

So it is basically a rebranded React Native? Or completely different?

Or could someone ELI5 / TL;DR? The whole blog post is basically saying how good Lynx is and what problem it solves without telling me much technical details.

No. No, it isn't. Lynx is a browser I use in the terminal. This is something else.

Not to be confused with Lynx Browser: The Land That Time Revived https://news.ycombinator.com/item?id=43119238

Would be interested in using this if not for the lack of LLM support.

Can you also explain what the advantages are over React Native?

Lynx doesn't support LLMs or there's not documentation?

I bet they mean LLMs don't know the syntax, and won't be helpful in writing code for it.

By this criteria, we can literally never release any new programming languages or frameworks ever again...

If LLMs are going to be as useful as they claim, they have to be smart/flexible enough to adapt to new information.

> Would be interested in using this if not for the lack of LLM support.

I'm not even in the anti-LLM crowd, but that sentence made me shake my head in disgust

How so?

Primarily because the sentence doesn't even make any sense? It seems like rage bait.

Reply on news.ycombinator.com