[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: sysvinit is still here, and here to stay for jessie (was Re: systemd is here to stay, get over it now)

Hash: SHA512

On 07/03/2014 11:53 PM, Russ Allbery wrote:

> The Wanderer <wanderer@fastmail.fm> writes:
>> I, for one, would be highly displeased if a routine dist-upgrade to
>> testing required me to reboot to avoid having things break.
>> I generally dist-upgrade my primary computer to testing about once
>> a week, give or take, but I don't reboot it more often than once a
>> month - more commonly three to six months, and I'd prefer that to
>> be longer if possible. (And often when I do reboot, it's due to a
>> power outage that overwhelms my UPS.)
>> Yes, this means that I don't get the benefit of some of the
>> upgraded packages (e.g. new kernels) until I do reboot - but
>> nothing breaks, either. Given the inconvenience of needing to shut
>> down everything I'm doing (including dozens of xterms, many with
>> running programs) to reboot, and manually bring up what parts of it
>> I can afterwards, I'm entirely willing to live without those
>> benefits in the interim.
> Speaking as someone who runs unstable on his laptop, power management
> has not worked across a dist-upgrade many times in unstable during
> both the current development cycle and the wheezy development cycle.
> Usually it's just that suspend functions don't work without a reboot
> (which in many cases, such as a new Linux kernel version, makes
> obvious sense, but I've had it not work a bunch of times when there
> wasn't a new kernel version), but I've had the power button in the
> Xfce bar not work at all in the past. So this is already the current
> state of play in Debian based on my personal experience, and has been
> for quite some time.

Suspend - or, rather, resume from suspend - hasn't worked on this
(desktop) computer ever, that I've seen. Given the symptoms, I suspect
the problem lies in fglrx, which is why I haven't bothered reporting it.

On my laptop (which gets a similar upgrade pattern), by contrast, the
only times I've ever seen suspend fail have been from an uninterruptable
running task - most commonly updatedb trying to access an inaccessible
NFS share. For whatever little or nothing that may be worth.

> If you haven't noticed, I suspect that you don't have as high of a 
> sensitivity to things breaking than you might think.  Or, at least,
> things you care about have not broken.

That's entirely possible, though I think the latter is a bit more likely
than the former - particularly because I use rather fewer things than
many other people, and don't use most fancy GUI elements. (For example,
I don't have a graphical "power button" at all; I shut down by exiting
my window manager, logging out of the console where I had originally run
startx, logging in as root, and running 'shutdown -h'.)

I'll also note that I specifically track testing, not unstable. I used
to track unstable, but I had far too many problems (not all of which I
could find a practical way to fix without a from-scratch reinstall) as a
result, quite possibly including some of the sort of problems you're
talking about.

- --
   The Wanderer

Secrecy is the beginning of tyranny.

A government exists to serve its citizens, not to control them.
Version: GnuPG v1
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/


Reply to: