Hi. On Mon, 31 Aug 2015 19:40:54 +0100 Lisi Reisz <lisi.re...@gmail.com> wrote:
> On Monday 31 August 2015 19:27:31 Reco wrote: > > Hi. > > > > On Mon, 31 Aug 2015 18:25:09 +0100 > > > > Lisi Reisz <lisi.re...@gmail.com> wrote: > > > On Monday 31 August 2015 16:59:48 Nicolas George wrote: > > > > Le quartidi 14 fructidor, an CCXXIII, Lisi Reisz a écrit : > > > > > For those who have still not discovered, you have to press ^ three > > > > > times in succession inside a second. > > > > > > > > > > https://tlhp.cf/lennart-poettering-su/ > > > > > > > > Are you referring to that snippet: > > > > > > > > # Connected to the local host. Press ^] three times within 1s to exit > > > > session. > > > > > > > > ... or are you referring to other parts of the page that I missed or > > > > parts in the video? > > > > > > > > If you are referring to that snippet, I suspect you are reading it > > > > wrong. > > > > > > > > For once, it is "^]", i.e. Ctrl-], i.e. ASCII 0x1D, aka "group > > > > separator". > > > > > > > > You can notice it is the same as the "escape character" present in most > > > > telnet implementations. > > > > > > > > And my second point is: it is obviously meant for emergency exit, like > > > > tilde-point in SSH. You should need it almost never in normal use, > > > > where you exit either by typing the command "exit" or by sending the > > > > EOF code (usually Ctrl-D), just like su. > > > > > > > > Actually, AFAIK neither sudo nor su support an emergency exit sequence. > > > > If that has not bothered you until now, it should not bother you from > > > > now on either. > > > > > > Then I have misunderstood, which does not surprise me. > > > > > > What is the alternative to su that there is so much fuss about? And I > > > don't care about the session ending function it apparently has. <su> > > > will change me to root and <su $USER> will change me to the user. Is > > > that what people fear will disappear? And what do they fear will be put > > > in its place? (Yes, I understand that so far it is in addition, not > > > instead of, but what is the fuss about? What has Lennart proposed?) > > > > It's really simple. > > > > 1) Boot with init=/bin/sh kernel commandline. > > > > 2) Invoke su - <some_user>. Observe the result. > > > > 3) Invoke "machinectl shell". Observe the result. > > > > 4) Compare results from 2) and 3). > > Thanks for the replies. But I still don't know what is "going" and what > is "coming". 2) should work. 3) should not. Since booting with "init=/bin/sh" is one of the valid ways of thoubleshooting failing OS, replacing "su" with "machinectl shell" effectively limits usefulness of such approach. I'm *not* saying that it will render "boot with init=/bin/sh" completely useless, for the record. But this long road consists of small steps, and some of them have been taken already. > Oh well, I shall no doubt discover. At the moment on my Jessie machine I can > open a terminal, su into root, perform whatever it is, and su out. > > I'll worry about it when/if I can't. That's my approach too (but I have the contingency plan already :). > Is the proposed change only going to have an effect so early in the process? Of course, not. "machinectl login" should provide the user with a new "session" (whatever that term means in newspeak). The whole idea of [1] is that "su" does not do so (whenever it should is another topic), therefore "su" should be replaced. [1] https://github.com/systemd/systemd/issues/825 Reco