← Back to context

Comment by SingleSourceAI

1 day ago

[flagged]

Both Samsung and Google already did it. My S26 Ultra supports Airdrop and I've tested it by sending and receiving photos with iPad

>It uses a time-sliced channel-hopping mechanism so the radio can serve both infrastructure WiFi and the direct peer link simultaneously.

This seems like such a basic solution that I'm surprised that it isn't required by any of the mainstream standards before WiFi Aware. I wonder if this was some sort of a patent issue or similar.

It is entirely possible to inject (unrelated) wifi frames while being associated to a BSS without violating the existing 802.11 standards. That’s why Apple is able to implement AWDL on standard compliant wifi hardware.

However the path towards this type of interoperability would likely go through additional standardization via IEEE 802.11* and the Wi-Fi alliance. At which point Apple will need to implement and support the new standards. There is no need to reverse engineer AWDL to meet the new European interoperability requirements. What is needed is for wifi chipset OEMs to implement such standardization. Something pretty routine of them.

It can be expected that Apple will also maintain the proprietary AWDL in order to support their legacy devices.

  • AFAIK, Wi-Fi Aware / Neighbourhood Aware Networking is basically the "standardised" version of AirDrop, and as of 2025, iOS's Airdrop transparently inter-operates with it.

AWDL is such an amazing technology, it's understandable that Apple wants to keep it only for their devices as it gives them a noticeable advantage for quick stuff sharing.

> which is a proprietary peer-to-peer layer that runs alongside your existing WiFi connection without dropping it. It uses a time-sliced channel-hopping mechanism so the radio can serve both infrastructure WiFi and the direct peer link simultaneously.

Maybe a network nerd can chime in - is this implementation so difficult that it's unrealistic we'll see an OSS version?

  • I think the thing that makes an OSS implementation more difficult than iOS/macOS is the friction involved.

    Say you've got an android phone, windows PC, and a linux box, and you want to be able to quickly drop files from each one. unless we get some kind of cooperation across all three platforms at the OS level, you'd at minimum need to install some kind of client into each system - when the nicest feature of airdrop is that it's baked into all of Apple's OSs, in my opinion. even if it worked exactly the same way, but had to be installed, I think it would see less use - and there's no real way for a single OSS project to do that across multiple OS platforms, to my knowledge

  • The physical layer part really isn't complicated, and most Wi-Fi chipsets have supported something like it for over a decade now.

    What's tricky is to specify and get everybody to implement the layers on top of it, including device discovery (frequently offloaded to Bluetooth for efficiency reasons), user identification (Apple runs a PKI for this) etc.

  • Not an expert on mobile development but I doubt an android app has the low-level access needed to the wifi stack to do this.

There is an open standard for that which is included in Apple devices since the iPhone 15. google implements it since the pixel 3. It’s called NAN. There are no WiFi cards available for consumers to buy which expose that as part of their firmware sadly. But wpa_supplicant has implemented part of the standard.

This is misinformation, including most of the comments here, the majority of phones from 2014 support Wi-Fi Direct, and simultaneous group and station mode (2 BSS, yes even different channels). Even most Wi-Fi chips generally not just smartphones for a very long time. They stay connected to your home network.

When Quickshare drops your Wi-Fi connection, its not Direct anymore, that's just soft AP from an error, and if that doesn't work, it fallback to Bluetooth. Bluetooth is used for provisioning as well.

The only reason why many apps don't use it is because of buggy implementation, some phones require a full restart after using Wi-Fi Direct to fix connectivity issues, even Motorola's own product line with Smart Connect use it only with certain models, despite having Wi-Fi direct due to poor implementation (can be forced). They even have a white list of supported adapter for the Windows app since direct is used as well, can be unofficially force enabled for Mediatek based adapters (rare on some laptops).

Back in 2016 things were much stable on Android phones with Wi-Fi Direct, even with old Blackberry, there were many apps including file managers that used it before it was essentially dropped, even for onboarding/provisioning apps like HP printers...

Apple's Airdrop success is about gaining traction, in the era of Wi-Fi Direct or other methods, most people were not aware of such features, as it required an app to be installed, they used email/messaging, even when Airdrop was first introduced and preinstalled, it took years for the average person to use it.

> It uses a time-sliced channel-hopping mechanism so the radio can serve both infrastructure WiFi and the direct peer link simultaneously.

This is really nothing special as 802.11 implementations go, as it's pretty easy to do as long as you can control the physical channel for at least one side.

Many Windows, Linux, and Android devices have been supporthing this for years. It's usually called something like "simultaneous AP/STA mode".

also they use mDNS, which many programming languages, such as go, got it in their net library

  • No they don’t? There are several 3p libs for it but not in std. Unless I’m blind and didn’t get the memo.

    • No you didn't miss. I mixed the concept, UDP broadcast similar to mDNS, yes part of net library. actual mDNS like visible in Finder and actual RFC implementation not part of standard library. What I was trying to say was easy to make "mdns-ish" with just standard lib, rush typed it and ended up like that.