The Code To Intel's New Linux Wireless Daemon Is Now Public

[phoronix]

Phoronix: The Code To Intel's New Linux Wireless Daemon Is Now Public

During this year's systemd conference there was talk of A New Wireless Daemon Is In Development To Potentially Replace wpa_supplicant. At that time the code wasn't yet public to this new open-source WiFi daemon developed by Intel, but since then the code has now opened up...

http://www.phoronix.com/scan.php?page=news_item&px=IWD-Git-Intel-Wireless-Daemon

[Sonadow]

Just tried building this. The instructions do not mention how to use iwd, iwmon or iwctl to connect and perform wpa authentication with an AP.

[bug77]

I haven't been paying attention, but what was wrong with wpa_supplicant and why couldn't it be simply enhanced?
I know even the best maintained code eventually becomes cheaper to replace than to maintain (and even more so when following the Unix philosophy), but I still don't know the whys in this case.

[mastercoms]

I haven't been paying attention, but what was wrong with wpa_supplicant and why couldn't it be simply enhanced?
I know even the best maintained code eventually becomes cheaper to replace than to maintain (and even more so when following the Unix philosophy), but I still don't know the whys in this case.

https://www.phoronix.com/scan.php?page=news_item&px=New-Linux-Wireless-Daemon

[phred14]

What are the implications for non-Intel wifi? Is this supposed to work for them as well>

[bug77]

https://www.phoronix.com/scan.php?pa...ireless-Daemon

I've looked there, it only says what the new daemon does, it doesn't talk of any insurmountable shortcomings of wpa_supplicant.

[starshipeleven]

What are the implications for non-Intel wifi? Is this supposed to work for them as well>

This daemon isn't Intel-specific.

Besides, they can't really claim they want to replace wpa_supplicant with a intel-only daemon when like 90% of the wifi chipsets aren't made by Intel anyway.

[starshipeleven]

I've looked there, it only says what the new daemon does, it doesn't talk of any insurmountable shortcomings of wpa_supplicant.

Look at the video.
tl;dr the wpa_supplicant has a shitty API towards userspace requiring "like 60x more code than for an application targeting BlueZ" (citing).