Daniel Farina

> So I guess that it is definitely possible to remove some overhead using the daemon mode. > > Also, I think that it is possible to introduce some lightweight...

I'm a big fan of the anti-join, I think it's a pity that that SQL makes it syntactically asymmetric with the other kinds of join (especially since NOT IN is...

Is that from a donation? schedules don't guarantee that a strand won't be run.

I think we can include-by-default. NetAddr can be handled specially as `to_s`, until we fork that thing and start changing some stuff (the `#inspect` rendering is probably the first priority,...

Unrelatedly, I also go back and forth whether the block form I used for the values in `Clog.emit` was a good idea, so, if you are finding it causing friction...

what does `resolvectl` say? Example from my laptop: ``` Global Protocols: LLMNR=resolve -mDNS -DNSOverTLS DNSSEC=no/unsupported resolv.conf mode: stub Link 2 (wlp59s0) Current Scopes: DNS LLMNR/IPv4 LLMNR/IPv6 Protocols: +DefaultRoute LLMNR=resolve -mDNS...

Interesting, so, on the VM, dnsmasq absorbs queries for ipv4 but not for ipv6. May be something to smooth out, to enable local DNS resolutions, I guess the ipv6 dns...

> > can next_label be required? maybe the old retval form has no virtues and should be eliminated. > > If the prog runs and returns some values, it could...

Hi there! I haven't been with Citus/Microsoft for about two years now. But I think the others still are.

Thinking back on this, one reason this persisted is that customers were generally okay with this relaxed guarantee as it comes to recoveries for making copies, and they don't have...