David Gibson

/cc @Luap99 > While injecting mtu=65520 as option is easy it will break consumers that inspect the networks and expect certain options to be (or not be) there... Doing something...

> > Ok.. I'm not totally clear on what the difference is between this and the first option you rejected. > > Basically it comes down to not showing the...

@Luap99 @edsantiago, finally got around to revising this, sorry it took so long. A bunch of the Fedora 41 and rawhide tests are failing. I think it looks like an...

New version pushed. Only change is a rebase + an extra patch attempting to address @edsantiago's comment.

[snip] > So, if no address is given explicitly for `--publish / -p`, pasta binds the port to any address (which looks rather convenient and actually correct to me), and...

> > [snip] > > > So, if no address is given explicitly for `--publish / -p`, pasta binds the port to any address (which looks rather convenient and actually...

Patches to change this behaviour are now merged into pasta upstream and should be in the next release.

@pwige Hi. I'm looking specifically at Scenario 2. As @sbrivio-rh notes that looks like a regression caused by the flow table, which I imlpemented. Unfortunately, I haven't yet been able...

@pwige Oh, sorry, one more 5. > Run a [speedtest](https://www.speedtest.net/) first with multiple connections and then with only a single connection enabled. I'm not entirely sure what you mean here....

> @pwige Oh, sorry, one more 5. > > > Run a [speedtest](https://www.speedtest.net/) first with multiple connections and then with only a single connection enabled. > > I'm not entirely...