__/ [ Michael B. Trausch ] on Saturday 27 May 2006 08:01 \__
> Chirag Shukla wrote in
> <1148705602.842732.321860@xxxxxxxxxxxxxxxxxxxxxxxxxxxx> on Sat, May 27 2006
>> Before calling it a day, one more thing struck me about Ubuntu that we
>> could make better. One pressing point that gives Linux leverage over
>> Windows is that we need not restart the computer so often. Well, I have
>> had Windows and Linuxes running parallel without dying, so I wont
>> complain about either OSes on that, but I would agree that after
>> updates Windows strongly prompts users to reboot whereas FC/Deb/SuSE
>> does not -- But Ubuntu seems to do that. Ubuntu informed me that I need
>> to reboot my machine. A general user might take it at heart and reboot
>> the machine even if it was not necessary. I dont know how updates and
>> reboots are handled in Mac, but if we could suggest Ubuntu developers
>> to not Window-fy Linux users by strongly suggesting them to reboot,
>> that would be good.
>> What do you say, folks?
>> Chirag Shukla.
> I've only seen this after a kernel upgrade -- is it happening more often
> then that? Typically, upgrades to the system kernel do not happen all that
> often, and especially if there are important root exploits or other
> important patches, then it can be very important to reboot.
This reminds me of last night's news:
( Red Hat Plugs Multiple Linux Kernel Flaws )
The good news, of course, is that all such fixes to the kernel propagate to
other distributors. There are so many eyes and hands on the code (well, only
one patchmaster -- Andy Morton -- does the work though). Compare that with
Apple and Microsoft where the kernel is closed-source. yes, even Darwin.
( Mac OS X Darwin x86 Still Open? )
I was once curious enough to read the kernel patches digest and I saw patches
coming from many directions, including @suse addresses. All in all, kernel
patches that are important are fairly rare. The most recent one I can think
of was X.org, but there was no exploit and only local users can take
advantage of carefully-doctored buffer overflows, IIRC.
> There are certain other things where a restart of the desktop environment
> windowing system might be a good thing, too. For example, there was an
> interesting exploit against X.org that, while I'm not sure if it is patched
> in most distributions yet, I would expect that some sort of restart of the
> X11 system should be prompted after the update is put into the system.
> It's only natural to restart when something is needed to be restarted.
Oh, hadn't read this paragraph before replying to the previous. Not an ideal
posting style, I guess. *smile* I guess I will not proofread, either.
> Well, if there were a way to force programs to reload theirselves while in
> RAM without creating the appearances of a restart, that'd be great, but I
> am not sure if that would be entirely possible.
In Vista, they speak of updates that do not require a reboot. How 'deep' the
update has to be remains an open question to me. There is also the quick
reboot debate, which relies on the chips, if not RAM, which needs
All that said, any reboot is or even logoff is highly undesirable to me as it
takes me about 5-10 minutes to restore a desktop session the way which suits
me, although KDE restores programs from the last session. Actually, as I
come to think of it, although my mail notifier has not worked for weeks and
it requires me to log in again, I still refuse to do it. I think that
frequent logons discourage good working habits in the desktop environment.
It may seem like it's rarely worth the investment, time-wise. Think in terms
of "why should I open all those programs if I need to log in again, later
on?". One thing that I miss in KDE 3.1 (work) and KDE 3.4 (home) is the
ability to have windows and sessions restored in the correct (as in
previously used) virtual desktop. At present, they all get opened in the
> Shared library updates might have similar consequences, depending on the
> reasons for patching/upgrade. In any case, most end-users restart their
> systems often enough due to power outages, portable computers, or because
> they have annoying habits like powering down the PC when they leave it for
> three hours or something, so it's all good.