IWD 1.1 Released For Intel's Linux Wireless Daemon

[phoronix]

Phoronix: IWD 1.1 Released For Intel's Linux Wireless Daemon

Released at the end of October was Intel's IWD 1.0 wireless daemon while hitting the web this week was version 1.1...

http://www.phoronix.com/scan.php?page=news_item&px=Intel-IWD-1.1-Released

[kylew77]

I wonder if this will get ported to the other open source operating systems, particularly FreeBSD and OpenBSD because I see posts on other forms all the time asking for ways to control what wireless network a computer joins from those operating systems by new users. Since Intel wireless NICs seem to be the most supported by the BSDs this would seem like a God sent.

[polarathene]

I wonder if a future update will be able to report what type of wifi connection was established(802.11 N/AC/etc). Bit strange that such info isn't available on Linux? I don't know much about what's involved on a technical level establishing the protocol version to use, but would think that there is some type of negotiation involved which would be able to provide that information?

[R41N3R]

I've switched some systems to iwd and it works great, connection time is much shorter. Only the password dialog looked different on KDE, maybe not fully integrated yet to Network manager.

[Alpha64]

I wonder if a future update will be able to report what type of wifi connection was established(802.11 N/AC/etc). Bit strange that such info isn't available on Linux? I don't know much about what's involved on a technical level establishing the protocol version to use, but would think that there is some type of negotiation involved which would be able to provide that information?

I use:

iw <devname> link

to get this information. It shows basically everything, although it is pretty technical. For example, here is what I see on my system:

# iw wlo1 link
Connected to xx:xx:xx:xx:xx:xx (on wlo1)
        SSID: <removed>
        freq: 5745
        RX: 468637 bytes (628 packets)
        TX: 54225 bytes (434 packets)
        signal: -79 dBm
        tx bitrate: 120.0 MBit/s VHT-MCS 5 40MHz short GI VHT-NSS 1

        bss flags:      short-slot-time
        dtim period:    2
        beacon int:     100

We see that since freq starts with 5, we know we are using the 5GHz band (if it started with 2, one would know it was using the 2.4GHz frequency band). The tx bitrate tells us a lot about the connection. Since we see "VHT-MCS 5", we know the connection is 802.11ac. It also shows the connection is 40MHz wide, and is using short guard interval (short GI).

You can find out more about these details even on the 802.11ac page of Wikipedia, or your favorite wireless reference site.

I share this because I had the same question at one point, and finally found this. Yes, it is very low-level, but all of the relevant details are there...

[starshipeleven]

I wonder if this will get ported to the other open source operating systems, particularly FreeBSD and OpenBSD because I see posts on other forms all the time asking for ways to control what wireless network a computer joins from those operating systems by new users. Since Intel wireless NICs seem to be the most supported by the BSDs this would seem like a God sent.

I really doubt it, IWD is heavily reliant on Linux kernel API, that's the whole point of the project. It does not care about wifi hardware brand.

I suggest looking into networkmgr instead https://github.com/ghostbsd/networkmgr

[timofonic]

I wonder if this will get ported to the other open source operating systems, particularly FreeBSD and OpenBSD because I see posts on other forms all the time asking for ways to control what wireless network a computer joins from those operating systems by new users. Since Intel wireless NICs seem to be the most supported by the BSDs this would seem like a God sent.

Oh yes, please port systemd to BSD too...

It's a sarcasm. Nobody cares really about BSD other than scavengers, aka corporate predators (Sony for PlayStation operating system, content providers using their own sauce for network servers such as Netflix, etc. ).

I'm not against it, if course. I just joke at such infantile possibilities.

I just want wpa_supplicant to die and get totally and functionally replaced by an extremely full featured and extremely robust IWD, merged to systemd in a very coherent and well done way no other project ever did before in the computing history. Please not laugh at me, or do it if you want to

[starshipeleven]

I just want wpa_supplicant to die and get totally and functionally replaced by an extremely full featured and extremely robust IWD, merged to systemd in a very coherent and well done way no other project ever did before in the computing history.

There is no real need to call systemd to the party. IWD is locked to Linux kernel anyway due to its founding principles and design, it cannot and will not be ported.

[Danny3]

I've switched some systems to iwd and it works great, connection time is much shorter. Only the password dialog looked different on KDE, maybe not fully integrated yet to Network manager.

Hi,
Can you please tell me how did you switched?
I'm on Kubuntu 19.10 and I'm curios how easy is to switch and if it improves the stability of my connection.
Thanks!

[starshipeleven]

Hi,
Can you please tell me how did you switched?
I'm on Kubuntu 19.10 and I'm curios how easy is to switch and if it improves the stability of my connection.
Thanks!

This is for Ubuntu 19.10, and assumes you are using NetworkManager https://blobfolio.com/2019/10/replac...n-ubuntu-eoan/

EDIT: the last part where he is editing the wpa_supplicant service file is better done as a local service override, so you can freely update the wpa_supplicant package without undoing your little hack to neuter it.

https://askubuntu.com/questions/659267/how-do-i-override-or-configure-systemd-services