Lennart Poettering

> Checking on the "whole" block what does means ? In my PC I have 5 hard disks; the problem against I want to warn is that the EFI firmware...

Nah, I am pretty sure we should not warn about things that are quite likely not a problem at all. In my work I basically daily have more than one...

We could keep the lstening fd open i guess, so that we remain connectable all the time. that would mean ongoing connections might be dropped, but you can intstantly create...

There should be only one listener fd for the oomd stuff. hence I'd just add a call varlink_server_get_listen_fd() or so, and it returns -ENOTUNIQ if more than one listener is...

> > ``` > > > > ``` > > > I reordered functions to be in two groups: message construction and reading. > enter+exit is for reading, and open+cancel...

Regarding 7 + 8. I think it's a bad idea to dump so many params on users just like that. I do acknowledge is that it is useful to have...

Please provide debug logs of this, otherwise this is not actionable. i.e. issue `systemd service-log-level systemd-resolved debug` and then reproduce the issue, and paste the logs this then generated here.

uh, this seems overly broad. What is the usecase here? we synthesize a variety of RRs, and I doubt most of those are anything anyone would ever want to disable....

There's now an env var too: SYSTEMD_NSS_RESOLVE_SYNTHESIZE, see nss-resolve man page. Wouldn't that suffice? See 8ef114c692846b0a801807a087ee65a1c7c6c7c3

So I'd be with a change like this to disable synthesis in resolved server-side, but I think it should be more focussed: only on the local hostname itself (leave the...