I completely agree with the opt-in approach. I can take a look at the
codebase to see how config options are handled, and try to patch that
Regarding git send-email, I initially thought that doing so would not
let me add context or explanation about the patch, but now I see that
it is not the case (w/ --compose). I'll follow this approach for the
On Fri, 2020-07-17 at 11:03 -0500, Denis Kenzior wrote:
On 7/17/20 9:58 AM, Roberto "Roobre" Santalla wrote:
> This is my first time submitting a patch/feature request over
> email, I
> hope I do not mess up :)
We prefer git send-email, similar to how kernel patches are
HACKING document for more details.
> First of all, thank you for your work on iwd, I have recently
> to use it on my laptop and seems to be working very well.
> I have, however, noticed that when using the built-in IP
> management, my
> hostname was not being sent to the DHCP server. This is unfortunate
> since it breaks my home network, which uses dnsmasq to build dns
> records on-the-fly for hosts.
Right. We chose not to send the hostname in order to comply with RFC
For networks where this is required (like yours) we should probably
'opt-in' method. Perhaps the easy way to start is by adding another
the network IPv4 settings. Maybe a boolean setting named
the 'DHCPv4' group.
> I have written and extremely simple, PoC patch which calls
> and pass the returned string to the ell dhcp client.
> What are your thoughts about this feature?
This looks fine, but needs to check the 'opt-in' setting as discussed
> - ℝ
> 👤 Roberto Santalla Fernández
> 🔑 https://keybase.io/roobre/key.asc
> 🔑 0E19 86A3 593E 6226
> 🌍 https://roobre.es
👤 Roberto Santalla Fernández
🔑 0E19 86A3 593E 6226